nba.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# Predicting Wins/Losses"
      ],
      "metadata": {
        "id": "7C-jL9Hhoqk_"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "This sections focuses on predicting wins and losses for NBA teams. Our final goal was to predict spreads, but we wanted to see how accurate we could predict wins or losses, with a goal of >60% to be significantly higher than that of a coin flip. "
      ],
      "metadata": {
        "id": "AwbIl-VEBOPh"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NRqMHF9XIA7U"
      },
      "source": [
        "**Data Preprocessing / Feature Engineering**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5vYU1bQQIid1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5a358292-c016-4825-bbfa-713445937f7d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[ 401540753\n",
            "401540757 ]\n"
          ]
        }
      ],
      "source": [
        "import requests\n",
        "import datetime\n",
        "import pandas as pd\n",
        "from bs4 import BeautifulSoup\n",
        "\n",
        "# Coded by Austin Fett\n",
        "month = {\n",
        "    \"January\": 1,\n",
        "    \"February\": 2,\n",
        "    \"March\": 3,\n",
        "    \"April\": 4,\n",
        "    \"May\": 5,\n",
        "    \"June\": 6,\n",
        "    \"July\": 7,\n",
        "    \"August\": 8,\n",
        "    \"September\": 9,\n",
        "    \"October\": 10,\n",
        "    \"November\": 11,\n",
        "    \"December\": 12\n",
        "}\n",
        "\n",
        "refTable = {\n",
        "    \"NY\" : \"NYK\",\n",
        "    \"UTAH\" : \"UTA\",\n",
        "    \"NO\" : \"NOP\",\n",
        "    \"SA\" : \"SAS\",\n",
        "    \"GS\" : \"GSW\",\n",
        "    \"WSH\" : \"WAS\"\n",
        "}\n",
        "\n",
        "# Gets games from certain range of espn gameids\n",
        "def getGames():\n",
        "  games = []\n",
        "  checker = False\n",
        "\n",
        "  # espn gameid range for 2023 season for all games including 2023-2-27\n",
        "  # 401468016, 401468029\n",
        "  # 401468170, 401469077\n",
        "  # espn gameid range for all games in 2022 season\n",
        "  # 401358773, 401358785\n",
        "  # 401359833, 401361049\n",
        "  # espn gameid range for all games in 2021 season\n",
        "  # 401266805, 401266806\n",
        "  # 401267150, 401267718\n",
        "  # 401307309, 401307891\n",
        "\n",
        "  for i in range(401540753, 401540780+1):\n",
        "    URL = \"https://www.espn.com/nba/boxscore/_/gameId/\" + str(i)\n",
        "    page = requests.get(URL)\n",
        "    soup = BeautifulSoup(page.content, \"html.parser\")\n",
        "\n",
        "    g = getOdds(soup)\n",
        "    \n",
        "    # Prints out gameid ranges which are useful\n",
        "    if g != None:\n",
        "      if not checker:\n",
        "        print('[', i)\n",
        "      games.append(g)\n",
        "      checker = True\n",
        "    elif checker:\n",
        "      print(i - 1, ']')\n",
        "      checker = False\n",
        "    \n",
        "  return games\n",
        "\n",
        "# Gets odds of specific game given soup content of espn page\n",
        "def getOdds(soup):\n",
        "  results = soup.find(class_=\"n8 GameInfo__Meta\")\n",
        "\n",
        "  # Return none if page is empty\n",
        "  if results is None:\n",
        "    return None\n",
        "  \n",
        "  # Gets date of game\n",
        "  date_list = results.text.split(\" \")\n",
        "  date = datetime.datetime(int(date_list[4][:4]), int(month[date_list[2]]), int(date_list[3][:-1]))\n",
        "  date_string = date.strftime(\"%Y-%m-%d\")\n",
        "\n",
        "  # Gets name of teams that played\n",
        "  results = soup.find_all(class_=\"ScoreCell__Truncate Gamestrip__Truncate h4 clr-gray-01\")\n",
        "  away = results[0].find('a', href=True)['href'].split(\"/\")[-2].upper()\n",
        "  home = results[1].find('a', href=True)['href'].split(\"/\")[-2].upper()\n",
        "\n",
        "  # Changes team initials if needed for conversion\n",
        "  a = refTable.get(away)\n",
        "  if a is None:\n",
        "    a = away\n",
        "  h = refTable.get(home)\n",
        "  if h is None:\n",
        "    h = home\n",
        "\n",
        "  results = soup.find(class_=\"betting-details-with-logo\")\n",
        "\n",
        "  # Return none if game was not played\n",
        "  if results is not None:\n",
        "    odds = results.find_all('div')\n",
        "    line = odds[0].text.split(' ')\n",
        "    overunder = odds[1].text.split(' ')\n",
        "  else:\n",
        "    return None\n",
        "\n",
        "  # Gets line for game, 0 point spread if even\n",
        "  if len(line) < 3:\n",
        "    l = [home, \"0\"]\n",
        "  else:\n",
        "    l = line[1:]\n",
        "\n",
        "  # Gets favorite of game\n",
        "  fav = refTable.get(l[0])\n",
        "  if fav is None:\n",
        "    fav = l[0]\n",
        "  \n",
        "  line_string = fav + \" \" + l[1]\n",
        "\n",
        "  # Formats overunder\n",
        "  if float(overunder[1]) % 1 == 0:\n",
        "    o = int(overunder[1])\n",
        "  else:\n",
        "    o = float(overunder[1])\n",
        "\n",
        "  return ([date_string, a, h, fav, l[1], o])\n",
        "\n",
        "# Gets game lines and transforms into a csv file\n",
        "def getOddsCSV():\n",
        "  games = getGames()\n",
        "  games_lines_df = pd.DataFrame(games)\n",
        "  games_lines_df = games_lines_df.rename(columns={0: 'Date', 1: 'Away', 2: 'Home', 3: 'Favorite', 4: 'Line', 5: 'Over/Under'})\n",
        "  games_lines_df.to_csv('2021_Season_Odds.csv')\n",
        "\n",
        "getOddsCSV()"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "These next two sections are pulling in the season average data to be used to predict wins and losses. The averages are then combined with the vegas lines pulled from ESPN. "
      ],
      "metadata": {
        "id": "zwoNiJ1pCAzZ"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "byi8lN4STC26"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "\n",
        "## Pulls box scores from PBP stats CSV files. The averages for each game throughout\n",
        "## the season up to that game are then calculated and added into the dataframe at that date.\n",
        "## This data will be used as testing data in our later models.\n",
        "## Pulls box scores from PBP stats CSV files. The averages for each game throughout\n",
        "## the season up to that game are then calculated and added into the dataframe at that date.\n",
        "## This data will be used as testing data in our later models.\n",
        "## Coded by Austin Fett\n",
        "def get_box_score_average_data():\n",
        "\n",
        "  # For future use, we will be pulling in all teams data\n",
        "  team_list = [\"Atl\", \"Bkn\", \"Bos\", \"Cha\", \"Chi\", \"Cle\", \"Dal\", \"Den\", \"Det\", \"Gsw\", \"Hou\", \"Ind\", \"Lac\", \n",
        "              \"Lal\", \"Mem\", \"Mia\", \"Mil\", \"Min\", \"Nop\", \"Nyk\", \"Okc\", \"Orl\", \"Phi\", \"Phx\", \"Por\", \"Sac\", \n",
        "              \"Sas\", \"Tor\", \"Uta\", \"Was\"]\n",
        "  \n",
        "  # Currently dealing with the Atlanta Hawks data over the last many seasons. \n",
        "  team = \"Atl\"\n",
        "\n",
        "  # CSV files are currently stored in a github that allows us to easily access the CSV files from coLab\n",
        "  team_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/' + team + '_Box_Score.csv'\n",
        "  opp_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/' + team + '_Opp_Box_Score.csv'\n",
        "\n",
        "  # Read the CSV files\n",
        "  team_data = pd.read_csv(team_file)\n",
        "  opp_data = pd.read_csv(opp_file)\n",
        "\n",
        "  # This is going to be more important in the future, but this csv files contains all of the season odds\n",
        "  # Up to this point in the sesason. \n",
        "  game_lines_2023_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2023_Season_Odds.csv'\n",
        "  game_lines_2023_data = pd.read_csv(game_lines_2023_file)\n",
        "\n",
        "  # Features that will be looked at in models\n",
        "  stat_list = [\"OffPoss\", \"Points\", \"FG2M\", \"FG2A\", \"Fg2Pct\", \"FG3M\", \"FG3A\", \"Fg3Pct\", \"NonHeaveFg3Pct\", \n",
        "              \"FtPoints\", \"PtsAssisted2s\", \"PtsUnassisted2s\", \"PtsAssisted3s\", \"PtsUnassisted3s\",\n",
        "              \"Assisted2sPct\", \"NonPutbacksAssisted2sPct\", \"Assisted3sPct\", \"FG3APct\", \"ShotQualityAvg\", \"EfgPct\",\n",
        "              \"TsPct\", \"PtsPutbacks\", \"Fg2aBlocked\", \"FG2APctBlocked\", \"Fg3aBlocked\", \"FG3APctBlocked\"]\n",
        "\n",
        "  # Initializing list of dicts\n",
        "  rows = [{\"Date\": ''}, {\"Spread_Result\": ''}, {\"O/U_Result\": ''}]\n",
        "  average_rows = [{\"Date\": ''}, {\"Spread_Result\": ''}, {\"O/U_Result\": ''}]\n",
        "  averageL5_rows = [{\"Date\": ''}, {\"Spread_Result\": ''}, {\"O/U_Result\": ''}]\n",
        "\n",
        "  # Initialize dataframes that will be used to hold features (statistical information)\n",
        "  for i in team_data.index:\n",
        "    d1 = {}\n",
        "    d2 = {}\n",
        "    d3 = {}\n",
        "\n",
        "    for stat in stat_list:\n",
        "      d1[stat] = ''\n",
        "      d2[stat] = ''\n",
        "      d3[stat] = ''\n",
        "\n",
        "    for stat in stat_list:\n",
        "      d1['Opp_' + stat] = ''\n",
        "      d2['Opp_' + stat] = ''\n",
        "      d3['Opp_' + stat] = ''\n",
        "\n",
        "    rows.append(d1)\n",
        "    average_rows.append(d2)\n",
        "    averageL5_rows.append(d3)\n",
        "\n",
        "  result_found = []\n",
        "\n",
        "  for i in team_data.index:\n",
        "    result_found.append(False)\n",
        "\n",
        "  # Go through csv and populate stat values\n",
        "  index = 0\n",
        "\n",
        "  # For each feature that will be used, we will loop through the csv files and grab the necessary\n",
        "  # statistical information\n",
        "  for stat in stat_list:\n",
        "    sum_team = 0\n",
        "    sum_opp = 0\n",
        "\n",
        "    # Loop through team data\n",
        "    for i in team_data.index:\n",
        "      # Ensures to only add certain data once\n",
        "      if not result_found[i]:\n",
        "        # Loop through game lines to find correct line\n",
        "        for j in range(len(game_lines_2023_data.index)):\n",
        "          if game_lines_2023_data[\"Date\"][j] == team_data[\"Date\"][i]:\n",
        "            away = game_lines_2023_data[\"Away\"][j]\n",
        "            home = game_lines_2023_data[\"Home\"][j]\n",
        "            \n",
        "            if away == team.upper():\n",
        "              rows[i][\"Home\"] = home\n",
        "              rows[i][\"Away\"] = away\n",
        "              index = j\n",
        "              break\n",
        "            elif home == team.upper():\n",
        "              rows[i][\"Home\"] = home\n",
        "              rows[i][\"Away\"] = away\n",
        "              index = j\n",
        "              break\n",
        "        \n",
        "        # Saves date of game\n",
        "        rows[i][\"Date\"] = game_lines_2023_data[\"Date\"][index]\n",
        "        average_rows[i][\"Date\"] = game_lines_2023_data[\"Date\"][index]\n",
        "        averageL5_rows[i][\"Date\"] = game_lines_2023_data[\"Date\"][index]\n",
        "\n",
        "        # Adds 1 for win, 0 for loss\n",
        "        if (team_data[\"Points\"][i] - opp_data[\"Points\"][i] > 0):\n",
        "          average_rows[i][\"won\"] = 1\n",
        "          averageL5_rows[i][\"won\"] = 1\n",
        "        else:\n",
        "          average_rows[i][\"won\"] = 0\n",
        "          averageL5_rows[i][\"won\"] = 0\n",
        "        \n",
        "        rows[i][\"PointDifference\"] = opp_data[\"Points\"][i] - team_data[\"Points\"][i]\n",
        "        average_rows[i][\"PointDifference\"] = opp_data[\"Points\"][i] - team_data[\"Points\"][i]\n",
        "\n",
        "        # Adds spread result of difference between favorite covering and the spread\n",
        "        if game_lines_2023_data[\"Favorite\"][index] == team.upper():\n",
        "          if (team_data[\"Points\"][i] - opp_data[\"Points\"][i]) != -float(game_lines_2023_data[\"Line\"][index]):\n",
        "            rows[i][\"Spread_Result\"] = float(team_data[\"Points\"][i]) - float(opp_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "            average_rows[i][\"Spread_Result\"] = float(team_data[\"Points\"][i]) - float(opp_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "            averageL5_rows[i][\"Spread_Result\"] = float(team_data[\"Points\"][i]) - float(opp_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "          else:\n",
        "            rows[i][\"Spread_Result\"] = 0\n",
        "            average_rows[i][\"Spread_Result\"] = 0\n",
        "            averageL5_rows[i][\"Spread_Result\"] = 0\n",
        "        else:\n",
        "          if (opp_data[\"Points\"][i] - team_data[\"Points\"][i]) != -float(game_lines_2023_data[\"Line\"][index]):\n",
        "            rows[i][\"Spread_Result\"] = float(opp_data[\"Points\"][i]) - float(team_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "            average_rows[i][\"Spread_Result\"] = float(opp_data[\"Points\"][i]) - float(team_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "            averageL5_rows[i][\"Spread_Result\"] = float(opp_data[\"Points\"][i]) - float(team_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "          else:\n",
        "            rows[i][\"Spread_Result\"] = 0\n",
        "            average_rows[i][\"Spread_Result\"] = 0\n",
        "            averageL5_rows[i][\"Spread_Result\"] = 0\n",
        "\n",
        "        # Adds over/under result of difference between total score and the line\n",
        "        if (team_data[\"Points\"][i] + opp_data[\"Points\"][i]) != float(game_lines_2023_data[\"Over/Under\"][index]):\n",
        "          rows[i][\"O/U_Result\"] = float(team_data[\"Points\"][i]) + float(opp_data[\"Points\"][i]) - float(game_lines_2023_data[\"Over/Under\"][index])\n",
        "          average_rows[i][\"O/U_Result\"] = float(team_data[\"Points\"][i]) + float(opp_data[\"Points\"][i]) - float(game_lines_2023_data[\"Over/Under\"][index])\n",
        "          averageL5_rows[i][\"O/U_Result\"] = float(team_data[\"Points\"][i]) + float(opp_data[\"Points\"][i]) - float(game_lines_2023_data[\"Over/Under\"][index])\n",
        "        else:\n",
        "          rows[i][\"O/U_Result\"] = 0\n",
        "          average_rows[i][\"O/U_Result\"] = 0\n",
        "          averageL5_rows[i][\"O/U_Result\"] = 0\n",
        "\n",
        "        result_found[i] = True\n",
        "\n",
        "      sum_team += team_data[stat][i]\n",
        "      sum_opp += opp_data[stat][i]\n",
        "\n",
        "      # Adds the stats\n",
        "      rows[i][stat] = team_data[stat][i]\n",
        "      rows[i]['Opp_' + stat] = opp_data[stat][i]\n",
        "\n",
        "      # Averages the stats of all games up to now, adding it for the next game\n",
        "      if i != len(team_data.index):\n",
        "        average_rows[i + 1][stat] = sum_team / (i + 1)\n",
        "        average_rows[i + 1]['Opp_' + stat] = sum_opp / (i + 1)\n",
        "\n",
        "      # Averages the stats of the last 5 games\n",
        "      sum_stat = 0\n",
        "      sum_opp_stat = 0\n",
        "      count = 0\n",
        "      for j in range(1, 6):\n",
        "        if i - j >= 0:\n",
        "          sum_stat += team_data[stat][i - j]\n",
        "          sum_opp_stat += opp_data[stat][i - j]\n",
        "          count += 1\n",
        "\n",
        "      if count != 0:\n",
        "        averageL5_rows[i][stat] = sum_stat / count\n",
        "        averageL5_rows[i]['Opp_' + stat] = sum_opp_stat / count\n",
        "\n",
        "  Df = pd.DataFrame(rows[:-1])\n",
        "  averageDf = pd.DataFrame(average_rows[:-1])\n",
        "  averageL5Df = pd.DataFrame(averageL5_rows[:-1])\n",
        "  averageDf = averageDf.dropna()\n",
        "  wins = averageDf[['won']]\n",
        "  averageDf = averageDf.drop('won', axis = 1)\n",
        "  averageDf = averageDf.shift(periods = 1)\n",
        "  averageDf['won'] = wins\n",
        "  averageDf = averageDf.drop(1)\n",
        "  return averageDf"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "UGPy7CRT6nfR"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "\n",
        "## Pulls box scores from PBP stats CSV files. The averages for each game throughout\n",
        "## the season up to that game are then calculated and added into the dataframe at that date.\n",
        "## This data will be used as testing data in our later models.\n",
        "## Coded by Austin Fett\n",
        "def get_box_score_average_data_1():\n",
        "\n",
        "  # For future use, we will be pulling in all teams data\n",
        "  team_list = [\"Atl\", \"Bkn\", \"Bos\", \"Cha\", \"Chi\", \"Cle\", \"Dal\", \"Den\", \"Det\", \"Gsw\", \"Hou\", \"Ind\", \"Lac\", \n",
        "              \"Lal\", \"Mem\", \"Mia\", \"Mil\", \"Min\", \"Nop\", \"Nyk\", \"Okc\", \"Orl\", \"Phi\", \"Phx\", \"Por\", \"Sac\", \n",
        "              \"Sas\", \"Tor\", \"Uta\", \"Was\"]\n",
        "\n",
        "  ind = 0\n",
        "  Df = []\n",
        "  averageDf = []\n",
        "  averageL5Df = []\n",
        "  for team in team_list:\n",
        "    # Currently dealing with the Atlanta Hawks data over the last many seasons. \n",
        "\n",
        "    # CSV files are currently stored in a github that allows us to easily access the CSV files from coLab\n",
        "    team_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/' + team + '_Box_Score_2021_2023.csv'\n",
        "    opp_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/' + team + '_Opp_Box_Score_2021_2023.csv'\n",
        "\n",
        "\n",
        "    # Read the CSV files\n",
        "    team_data = pd.read_csv(team_file)\n",
        "    opp_data = pd.read_csv(opp_file)\n",
        "\n",
        "    # This is going to be more important in the future, but this csv files contains all of the season odds\n",
        "    # Up to this point in the sesason. \n",
        "    game_lines_2023_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2021_2023_Season_Odds.csv'\n",
        "    game_lines_2023_data = pd.read_csv(game_lines_2023_file)\n",
        "\n",
        "    # Features that will be looked at in models\n",
        "    stat_list = [\"OffPoss\", \"Points\", \"FG2M\", \"FG2A\", \"Fg2Pct\", \"FG3M\", \"FG3A\", \"Fg3Pct\", \"NonHeaveFg3Pct\", \n",
        "                \"FtPoints\", \"PtsAssisted2s\", \"PtsUnassisted2s\", \"PtsAssisted3s\", \"PtsUnassisted3s\",\n",
        "                \"Assisted2sPct\", \"NonPutbacksAssisted2sPct\", \"Assisted3sPct\", \"FG3APct\", \"ShotQualityAvg\", \"EfgPct\",\n",
        "                \"TsPct\", \"PtsPutbacks\", \"Fg2aBlocked\", \"FG2APctBlocked\", \"Fg3aBlocked\", \"FG3APctBlocked\"]\n",
        "\n",
        "    # Initializing list of dicts\n",
        "    rows = [{\"Date\": ''}, {\"Spread_Result\": ''}, {\"O/U_Result\": ''}]\n",
        "    average_rows = [{\"Date\": ''}, {\"Spread_Result\": ''}, {\"O/U_Result\": ''}]\n",
        "    averageL5_rows = [{\"Date\": ''}, {\"Spread_Result\": ''}, {\"O/U_Result\": ''}]\n",
        "\n",
        "    # Initialize dataframes that will be used to hold features (statistical information)\n",
        "    for i in team_data.index:\n",
        "      d1 = {}\n",
        "      d2 = {}\n",
        "      d3 = {}\n",
        "\n",
        "      for stat in stat_list:\n",
        "        d1[stat] = ''\n",
        "        d2[stat] = ''\n",
        "        d3[stat] = ''\n",
        "\n",
        "      for stat in stat_list:\n",
        "        d1['Opp_' + stat] = ''\n",
        "        d2['Opp_' + stat] = ''\n",
        "        d3['Opp_' + stat] = ''\n",
        "\n",
        "      rows.append(d1)\n",
        "      average_rows.append(d2)\n",
        "      averageL5_rows.append(d3)\n",
        "\n",
        "    result_found = []\n",
        "\n",
        "    for i in team_data.index:\n",
        "      result_found.append(False)\n",
        "\n",
        "    # Go through csv and populate stat values\n",
        "    index = 0\n",
        "\n",
        "    # For each feature that will be used, we will loop through the csv files and grab the necessary\n",
        "    # statistical information\n",
        "    for stat in stat_list:\n",
        "      sum_team = 0\n",
        "      sum_opp = 0\n",
        "\n",
        "      # Loop through team data\n",
        "      for i in team_data.index:\n",
        "        # Ensures to only add certain data once\n",
        "        if not result_found[i]:\n",
        "          # Loop through game lines to find correct line\n",
        "          for j in range(len(game_lines_2023_data.index)):\n",
        "            if game_lines_2023_data[\"Date\"][j] == team_data[\"Date\"][i]:\n",
        "              away = game_lines_2023_data[\"Away\"][j]\n",
        "              home = game_lines_2023_data[\"Home\"][j]\n",
        "              \n",
        "              if away == team.upper():\n",
        "                rows[i][\"Home\"] = 0\n",
        "                average_rows[i][\"Home\"] = 0\n",
        "                averageL5_rows[i][\"Home\"] = 0\n",
        "                index = j\n",
        "                break\n",
        "              elif home == team.upper():\n",
        "                rows[i][\"Home\"] = 1\n",
        "                average_rows[i][\"Home\"] = 1\n",
        "                averageL5_rows[i][\"Home\"] = 1\n",
        "                index = j\n",
        "                break\n",
        "          \n",
        "          # Saves date of game\n",
        "          rows[i][\"Date\"] = game_lines_2023_data[\"Date\"][index]\n",
        "          average_rows[i][\"Date\"] = game_lines_2023_data[\"Date\"][index]\n",
        "          averageL5_rows[i][\"Date\"] = game_lines_2023_data[\"Date\"][index]\n",
        "\n",
        "          # Adds 1 for win, 0 for loss\n",
        "          if (team_data[\"Points\"][i] - opp_data[\"Points\"][i] > 0):\n",
        "            rows[i][\"won\"] = 1\n",
        "            average_rows[i][\"won\"] = 1\n",
        "            averageL5_rows[i][\"won\"] = 1\n",
        "          else:\n",
        "            rows[i][\"won\"] = 0\n",
        "            average_rows[i][\"won\"] = 0\n",
        "            averageL5_rows[i][\"won\"] = 0\n",
        "        \n",
        "          rows[i][\"PointDifference\"] = opp_data[\"Points\"][i] - team_data[\"Points\"][i]\n",
        "          average_rows[i][\"PointDifference\"] = opp_data[\"Points\"][i] - team_data[\"Points\"][i]\n",
        "          averageL5_rows[i][\"PointDifference\"] = opp_data[\"Points\"][i] - team_data[\"Points\"][i]\n",
        "\n",
        "          # Adds spread result of difference between favorite covering and the spread\n",
        "          if game_lines_2023_data[\"Favorite\"][index] == team.upper():\n",
        "            if (team_data[\"Points\"][i] - opp_data[\"Points\"][i]) != -float(game_lines_2023_data[\"Line\"][index]):\n",
        "              rows[i][\"Spread_Result\"] = float(team_data[\"Points\"][i]) - float(opp_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "              average_rows[i][\"Spread_Result\"] = float(team_data[\"Points\"][i]) - float(opp_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "              averageL5_rows[i][\"Spread_Result\"] = float(team_data[\"Points\"][i]) - float(opp_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "            else:\n",
        "              rows[i][\"Spread_Result\"] = 0\n",
        "              average_rows[i][\"Spread_Result\"] = 0\n",
        "              averageL5_rows[i][\"Spread_Result\"] = 0\n",
        "          else:\n",
        "            if (opp_data[\"Points\"][i] - team_data[\"Points\"][i]) != -float(game_lines_2023_data[\"Line\"][index]):\n",
        "              rows[i][\"Spread_Result\"] = float(opp_data[\"Points\"][i]) - float(team_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "              average_rows[i][\"Spread_Result\"] = float(opp_data[\"Points\"][i]) - float(team_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "              averageL5_rows[i][\"Spread_Result\"] = float(opp_data[\"Points\"][i]) - float(team_data[\"Points\"][i]) + float(game_lines_2023_data[\"Line\"][index])\n",
        "            else:\n",
        "              rows[i][\"Spread_Result\"] = 0\n",
        "              average_rows[i][\"Spread_Result\"] = 0\n",
        "              averageL5_rows[i][\"Spread_Result\"] = 0\n",
        "\n",
        "          # Adds over/under result of difference between total score and the line\n",
        "          if (team_data[\"Points\"][i] + opp_data[\"Points\"][i]) != float(game_lines_2023_data[\"Over/Under\"][index]):\n",
        "            rows[i][\"O/U_Result\"] = float(team_data[\"Points\"][i]) + float(opp_data[\"Points\"][i]) - float(game_lines_2023_data[\"Over/Under\"][index])\n",
        "            average_rows[i][\"O/U_Result\"] = float(team_data[\"Points\"][i]) + float(opp_data[\"Points\"][i]) - float(game_lines_2023_data[\"Over/Under\"][index])\n",
        "            averageL5_rows[i][\"O/U_Result\"] = float(team_data[\"Points\"][i]) + float(opp_data[\"Points\"][i]) - float(game_lines_2023_data[\"Over/Under\"][index])\n",
        "          else:\n",
        "            rows[i][\"O/U_Result\"] = 0\n",
        "            average_rows[i][\"O/U_Result\"] = 0\n",
        "            averageL5_rows[i][\"O/U_Result\"] = 0\n",
        "\n",
        "          result_found[i] = True\n",
        "\n",
        "        sum_team += team_data[stat][i]\n",
        "        sum_opp += opp_data[stat][i]\n",
        "\n",
        "        # Adds the stats\n",
        "        rows[i][stat] = team_data[stat][i]\n",
        "        rows[i]['Opp_' + stat] = opp_data[stat][i]\n",
        "\n",
        "        # Averages the stats of all games up to now, adding it for the next game\n",
        "        if i != len(team_data.index):\n",
        "          average_rows[i + 1][stat] = sum_team / (i + 1)\n",
        "          average_rows[i + 1]['Opp_' + stat] = sum_opp / (i + 1)\n",
        "\n",
        "        # Averages the stats of the last 5 games\n",
        "        sum_stat = 0\n",
        "        sum_opp_stat = 0\n",
        "        count = 0\n",
        "        for j in range(1, 6):\n",
        "          if i - j >= 0:\n",
        "            sum_stat += team_data[stat][i - j]\n",
        "            sum_opp_stat += opp_data[stat][i - j]\n",
        "            count += 1\n",
        "\n",
        "        if count != 0:\n",
        "          averageL5_rows[i][stat] = sum_stat / count\n",
        "          averageL5_rows[i]['Opp_' + stat] = sum_opp_stat / count\n",
        "    \n",
        "    \n",
        "    tempDf = pd.DataFrame(rows[:-1])\n",
        "    Df.append(tempDf)\n",
        "    tempAverageDf = pd.DataFrame(average_rows[:-1])\n",
        "    tempAverageL5Df = pd.DataFrame(averageL5_rows[:-1])\n",
        "    tempAverageDf = tempAverageDf.dropna()\n",
        "    # wins = tempAverageDf[['won']]\n",
        "    # home = tempAverageDf[['Home']]\n",
        "    # tempAverageDf = tempAverageDf.drop('won', axis = 1)\n",
        "    # tempAverageDf = tempAverageDf.drop('Home', axis = 1)\n",
        "    # tempAverageDf = tempAverageDf.shift(periods = 1)\n",
        "    # tempAverageDf['won'] = wins\n",
        "    # tempAverageDf['Home'] = home\n",
        "    \n",
        "    # tempAverageDf = tempAverageDf.drop(1)\n",
        "    averageDf.append(tempAverageDf)\n",
        "    averageL5Df.append(tempAverageL5Df)\n",
        "\n",
        "    ind = ind + 1\n",
        "\n",
        "  return averageDf, Df, averageL5Df"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_pvYQD0BUN8y"
      },
      "outputs": [],
      "source": [
        "averageData, boxScoreData, l5Data = get_box_score_average_data_1()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7a5uJ9g9Q_mt",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6d3cc994-a28e-4909-f561-ac4f9d368785"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "6394"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ],
      "source": [
        "##This combines the boxscoredata to be one single dataframe and drops the nulls, we do\n",
        "##this for all of the dataframes returned \n",
        "combinedBox = pd.concat(boxScoreData)\n",
        "combinedBox = combinedBox.dropna()\n",
        "\n",
        "averageDataCombined = pd.concat(averageData)\n",
        "averageDataCombined = averageDataCombined.dropna()\n",
        "\n",
        "lastFive = pd.concat(l5Data)\n",
        "lastFive = lastFive.dropna()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "lywWKL6V_SHy"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "## This function is very similar to the above function grabbing average data, it is\n",
        "## just returning a different dataframe without the data and using different files.\n",
        "## These two function will be combined, but due to people working seperate ideas at the same\n",
        "## time, we decided to leave both in at the moment so we didn't ruin anyone's current work.\n",
        "\n",
        "#Coded by Austin Fett\n",
        "def get_box_score_train_data():\n",
        "\n",
        "  team_list = [\"Atl\", \"Bkn\", \"Bos\", \"Cha\", \"Chi\", \"Cle\", \"Dal\", \"Den\", \"Det\", \"Gsw\", \"Hou\", \"Ind\", \"Lac\", \n",
        "             \"Lal\", \"Mem\", \"Mia\", \"Mil\", \"Min\", \"Nop\", \"Nyk\", \"Okc\", \"Orl\", \"Phi\", \"Phx\", \"Por\", \"Sac\", \n",
        "             \"Sas\", \"Tor\", \"Uta\", \"Was\"]\n",
        "\n",
        "  team = \"ATL\"\n",
        "\n",
        "  team_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/' + team + '_Box_Score_4_Years.csv'\n",
        "  opp_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/' + team + '_Box_Score_Opp_4_Years.csv'\n",
        "  team_data = pd.read_csv(team_file)\n",
        "  opp_data = pd.read_csv(opp_file)\n",
        "\n",
        "  stat_list = [\"OffPoss\", \"Points\", \"FG2M\", \"FG2A\", \"Fg2Pct\", \"FG3M\", \"FG3A\", \"Fg3Pct\", \"NonHeaveFg3Pct\", \n",
        "            \"FtPoints\", \"PtsAssisted2s\", \"PtsUnassisted2s\", \"PtsAssisted3s\", \"PtsUnassisted3s\",\n",
        "            \"Assisted2sPct\", \"NonPutbacksAssisted2sPct\", \"Assisted3sPct\", \"FG3APct\", \"ShotQualityAvg\", \"EfgPct\",\n",
        "            \"TsPct\", \"PtsPutbacks\", \"Fg2aBlocked\", \"FG2APctBlocked\", \"Fg3aBlocked\", \"FG3APctBlocked\"]\n",
        "\n",
        "  # Initializing list of dicts\n",
        "  rows = [{\"Date\": ''}]\n",
        "  for i in team_data.index:\n",
        "    d1 = {}\n",
        "    d2 = {}\n",
        "\n",
        "    for stat in stat_list:\n",
        "      d1[stat] = ''\n",
        "      d2[stat] = ''\n",
        "\n",
        "    for stat in stat_list:\n",
        "      d1['Opp_' + stat] = ''\n",
        "      d2['Opp_' + stat] = ''\n",
        "\n",
        "    rows.append(d1)\n",
        "\n",
        "  result_found = []\n",
        "  for i in team_data.index:\n",
        "    result_found.append(False)\n",
        "\n",
        "  # Go through csv and populate stat values\n",
        "  index = 0\n",
        "  for stat in stat_list:\n",
        "    sum_team = 0\n",
        "    sum_opp = 0\n",
        "\n",
        "    for i in team_data.index:    \n",
        "      if not result_found[i]:\n",
        "        \n",
        "        rows[i][\"Date\"] = opp_data[\"Date\"][index]\n",
        "      \n",
        "        if (team_data[\"Points\"][i] - opp_data[\"Points\"][i] > 0):\n",
        "              rows[i][\"won\"] = 1\n",
        "        else:\n",
        "          rows[i][\"won\"] = 0\n",
        "\n",
        "        rows[i][\"PointDifference\"] = opp_data[\"Points\"][i] - team_data[\"Points\"][i]\n",
        "          \n",
        "        result_found[i] = True\n",
        "\n",
        "      sum_team += team_data[stat][i]\n",
        "      sum_opp += opp_data[stat][i]\n",
        "      \n",
        "      rows[i][stat] = team_data[stat][i]\n",
        "      rows[i]['Opp_' + stat] = opp_data[stat][i]\n",
        "      rows[i][stat + '_diff'] = team_data[stat][i] - opp_data[stat][i]\n",
        "      \n",
        "  Df = pd.DataFrame(rows[:-1])\n",
        "  Df = Df.dropna()\n",
        "  Df = Df[['Date', 'won', 'PointDifference', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "              'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'FG3APct', 'ShotQualityAvg',\n",
        "              'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "              'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_FG3APct', 'Opp_ShotQualityAvg',\n",
        "              'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "  return Df"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5rg8sAdE_VpQ"
      },
      "outputs": [],
      "source": [
        "## From the notes in assignment 2\n",
        "def split_train_test(data, test_ratio):\n",
        "    shuffled_indices = np.random.permutation(len(data))\n",
        "    test_set_size = int(len(data) * test_ratio)\n",
        "    test_indices = shuffled_indices[:test_set_size]\n",
        "    train_indices = shuffled_indices[test_set_size:]\n",
        "    return data.iloc[train_indices], data.iloc[test_indices]\n",
        "\n",
        "train, testData = split_train_test(averageDataCombined, .2)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qiyzqt54IMfw"
      },
      "source": [
        "**Split Train/Test Data**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "akKKdvxtBS9d"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import accuracy_score\n",
        "from keras.layers import Dense\n",
        "from keras.layers import LSTM\n",
        "\n",
        "# This function splits the data in train and test data. The training data consists\n",
        "# of data over the last 4 nba seasons, not including the current season. This data\n",
        "# is solely box scores, with a column indicating whether the team won or not (1 or 0).\n",
        "# The test set is the averages throughout the games played until that point in the set.\n",
        "# There is also a column indicating whether the team won or not at that point in the season\n",
        "# with those averages (1 or 0).\n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "def classification_train_test_data():\n",
        "  \n",
        "  # Pull box score data from the last 4 seasons\n",
        "  Df = combinedBox\n",
        "  # Pull average box score data from the current season \n",
        "  averageDf = get_box_score_average_data()\n",
        "\n",
        "  #Features used in the training data set and test set are the same\n",
        "  x_train = Df[['Home', 'Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'FG3APct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_FG3APct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "\n",
        "  #Label is 1 or 0 whether the team won (1) or lost (0)\n",
        "  y_train = Df[['won']]\n",
        "\n",
        "  x_test = testData[['Fg2Pct', 'Home', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'FG3APct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_FG3APct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "\n",
        "  y_test = testData[['won']]\n",
        "\n",
        "  return x_train, y_train, x_test, y_test\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "DM9kXWL_79q-"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import accuracy_score\n",
        "from keras.layers import Dense\n",
        "from keras.layers import LSTM\n",
        "from sklearn import preprocessing\n",
        "\n",
        "# This function splits the data in train and test data. The training data consists\n",
        "# of data over the last 4 nba seasons, not including the current season. This data\n",
        "# is solely box scores, with a column indicating whether the team won or not (1 or 0).\n",
        "# The test set is the averages throughout the games played until that point in the set.\n",
        "# There is also a column indicating whether the team won or not at that point in the season\n",
        "# with those averages (1 or 0).\n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "def classification_train_test_data_param(df):\n",
        "  \n",
        "  train, testData = split_train_test(df, .2)\n",
        "\n",
        "  #Features used in the training data set and test set are the same\n",
        "  x_train = train[['Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "              'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "              'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked']]\n",
        "\n",
        "  x_train = x_train.values\n",
        "  x_train = preprocessing.MinMaxScaler().fit_transform(x_train)\n",
        "  x_train = pd.DataFrame(x_train)\n",
        "\n",
        "  #Label is 1 or 0 whether the team won (1) or lost (0)\n",
        "  y_train = train[['won']]\n",
        "\n",
        "  x_test = testData[['Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "              'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "              'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked']]\n",
        "\n",
        "  x_test = x_test.values\n",
        "  x_test = preprocessing.MinMaxScaler().fit_transform(x_test)\n",
        "  x_test = pd.DataFrame(x_test)\n",
        "  \n",
        "  y_test = testData[['won']]\n",
        "\n",
        "  return x_train, y_train, x_test, y_test\n",
        "\n",
        "def classification_train_test_data():\n",
        "  \n",
        "  # Pull box score data from the last 4 seasons\n",
        "  Df = combinedBox[:5115]\n",
        "  # Pull average box score data from the current season \n",
        "  averageDf = averageDataCombined[5115:]\n",
        "\n",
        "\n",
        "  #Features used in the training data set and test set are the same\n",
        "  x_train = Df[['Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'FG3APct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_FG3APct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "  \n",
        "  x_train = x_train.values\n",
        "  x_train = preprocessing.MinMaxScaler().fit_transform(x_train)\n",
        "  x_train = pd.DataFrame(x_train)\n",
        "  #Label is 1 or 0 whether the team won (1) or lost (0)\n",
        "  y_train = Df[['won']]\n",
        "\n",
        "  x_test = averageDf[['Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'FG3APct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_FG3APct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "  \n",
        "  x_test = x_test.values\n",
        "  x_test = preprocessing.MinMaxScaler().fit_transform(x_test)\n",
        "  x_test = pd.DataFrame(x_test)\n",
        "\n",
        "  y_test = averageDf[['won']]\n",
        "\n",
        "  return x_train, y_train, x_test, y_test\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PlmKa2tF_paY"
      },
      "outputs": [],
      "source": [
        "## This function splits the data in train and test data. The training data consists\n",
        "## of data over the last 4 nba seasons, not including the current season. This data\n",
        "## is solely box scores, with a column indicating how much the team won or lost by (PointDifference).\n",
        "## The test set is the averages throughout the games played until that point in the set.\n",
        "## There is also a column indicating by how much the team won or lost by with those averages\n",
        "## at that point in the season. \n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "def regression_train_test_data():\n",
        "  \n",
        "  # Pull box score data from the last 4 seasons\n",
        "  Df = combinedBox[:5115]\n",
        "  # Pull average box score data from the current season \n",
        "  averageDf = averageDataCombined[5115:]\n",
        "\n",
        "  # Initialize train set\n",
        "  train_set = Df\n",
        "\n",
        "  # Drop Label and Date in the x train set. Won will also be dropped as that is something\n",
        "  # that won't be able to be used when predicting future games\n",
        "  x_train = Df[['Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'FG3APct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_FG3APct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "\n",
        "  x_train = x_train.values\n",
        "  x_train = preprocessing.MinMaxScaler().fit_transform(x_train)\n",
        "  x_train = pd.DataFrame(x_train)\n",
        "  # PointDifference will be used as the label\n",
        "  y_train = train_set[['PointDifference']]\n",
        "\n",
        "  #Initialize the test_set with these featuers\n",
        "  test_set = averageDf[['Home', 'Date', 'won', 'PointDifference', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'FG3APct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_FG3APct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "\n",
        "  #Drop date, won, and PointDifference for same reaseons as above\n",
        "  x_test = test_set.drop(columns = ['Date', 'won', 'PointDifference'])\n",
        "\n",
        "  x_test = x_test.values\n",
        "  x_test = preprocessing.MinMaxScaler().fit_transform(x_test)\n",
        "  x_test = pd.DataFrame(x_test)\n",
        "  #PointDifference will be label\n",
        "  y_test = test_set[['PointDifference']]\n",
        "\n",
        "  return x_train, y_train, x_test, y_test\n",
        "\n",
        "\n",
        "def regression_train_test_data_param(df):\n",
        "  \n",
        "  train_set, test_set = split_train_test(df, .2)\n",
        "\n",
        "  # Drop Label and Date in the x train set. Won will also be dropped as that is something\n",
        "  # that won't be able to be used when predicting future games\n",
        "  x_train = train_set[['Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "  x_train = x_train.values\n",
        "  x_train = preprocessing.MinMaxScaler().fit_transform(x_train)\n",
        "  x_train = pd.DataFrame(x_train)\n",
        "  # PointDifference will be used as the label\n",
        "  y_train = train_set[['PointDifference']]\n",
        "\n",
        "\n",
        "  #Drop date, won, and PointDifference for same reaseons as above\n",
        "  x_test = test_set[['Home', 'Fg2Pct', 'Fg3Pct', 'NonHeaveFg3Pct',\n",
        "             'FtPoints', 'Assisted2sPct', 'NonPutbacksAssisted2sPct', 'Assisted3sPct', 'ShotQualityAvg',\n",
        "             'EfgPct', 'TsPct', 'PtsPutbacks', 'FG2APctBlocked', 'FG3APctBlocked', 'Opp_Fg2Pct', 'Opp_Fg3Pct', 'Opp_NonHeaveFg3Pct',\n",
        "             'Opp_FtPoints', 'Opp_Assisted2sPct', 'Opp_NonPutbacksAssisted2sPct', 'Opp_Assisted3sPct', 'Opp_ShotQualityAvg',\n",
        "             'Opp_EfgPct', 'Opp_TsPct', 'Opp_PtsPutbacks', 'Opp_FG2APctBlocked', 'Opp_FG3APctBlocked']]\n",
        "\n",
        "  x_test = x_test.values\n",
        "  x_test = preprocessing.MinMaxScaler().fit_transform(x_test)\n",
        "  x_test = pd.DataFrame(x_test)\n",
        "  #PointDifference will be label\n",
        "  y_test = test_set[['PointDifference']]\n",
        "\n",
        "  return x_train, y_train, x_test, y_test\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "This is the beginning of training the models. The models will consist of logistic regression, random forest, linear regression, and a neural network. "
      ],
      "metadata": {
        "id": "of5fAlfvQB4p"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bUHOs49IIUr-"
      },
      "source": [
        "**Logistic Regression**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Cwf68Iiv_Yse",
        "outputId": "94da31d8-7d20-4748-e94e-10ba8422a374"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.5909715407262022, 0.5467399842890809)"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "## This function fits a logistic regression on the train and test sets that \n",
        "## are initialized in the classification data function\n",
        "## This is using data from PBP stats\n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "\n",
        "#Runs logistic regression on box score data\n",
        "def get_logistic_accuracy():\n",
        "\n",
        "  # Grab train and test sets\n",
        "  x_train, y_train, x_test, y_test = classification_train_test_data()\n",
        "\n",
        "  x, y , q, z = classification_train_test_data_param(lastFive)\n",
        "  #Initialize and fit simple logistic regression\n",
        "  logis = LogisticRegression(random_state = 0, max_iter=2000)\n",
        "  logis.fit(x_train, y_train.values.ravel())\n",
        "\n",
        "  # Predict on training data\n",
        "  train_pred = logis.predict(x_train)\n",
        "\n",
        "  # Predict on testing data \n",
        "  y_pred = logis.predict(x_test)\n",
        "\n",
        "  # Print accuracy score\n",
        "  # print(\"Logistic Regression Train Accuracy: \" + str(accuracy_score(y_train, train_pred)))\n",
        "  # print(\"Logistic Regression Test Accuracy: \" + str(accuracy_score(y_test, y_pred)))\n",
        "  return accuracy_score(y_train, train_pred), accuracy_score(y_test, y_pred)\n",
        "\n",
        "#Runs logistic regression on the averages of the last five days\n",
        "def get_logistic_accuracy_five():\n",
        "\n",
        "  # Grab train and test sets\n",
        "  x_train, y_train, x_test, y_test = classification_train_test_data_param(lastFive)\n",
        "\n",
        "  #Initialize and fit simple logistic regression\n",
        "  logis = LogisticRegression(random_state = 0, max_iter=2000)\n",
        "  logis.fit(x_train, y_train.values.ravel())\n",
        "\n",
        "  # Predict on training data\n",
        "  train_pred = logis.predict(x_train)\n",
        "\n",
        "  # # Predict on testing data \n",
        "  y_pred = logis.predict(x_test)\n",
        "\n",
        "  # Print accuracy score\n",
        "  # print(\"Logistic Regression Train Accuracy: \" + str(accuracy_score(y_train, train_pred)))\n",
        "  # print(\"Logistic Regression Test Accuracy: \" + str(accuracy_score(y_test, y_pred)))\n",
        "  return accuracy_score(y_train, train_pred), accuracy_score(y_test, y_pred)\n",
        "\n",
        "#Gets logistic regression accuracy on season averages\n",
        "def get_logistic_accuracy_average():\n",
        "  # Grab train and test sets\n",
        "  x_train, y_train, x_test, y_test = classification_train_test_data_param(averageDataCombined)\n",
        "\n",
        "  #Initialize and fit simple logistic regression\n",
        "  logis = LogisticRegression(random_state = 0, max_iter=2000)\n",
        "  logis.fit(x_train, y_train.values.ravel())\n",
        "\n",
        "  # Predict on training data\n",
        "  train_pred = logis.predict(x_train)\n",
        "\n",
        "  # # Predict on testing data \n",
        "  y_pred = logis.predict(x_test)\n",
        "\n",
        "  # Print accuracy score\n",
        "  # print(\"Logistic Regression Train Accuracy: \" + str(accuracy_score(y_train, train_pred)))\n",
        "  # print(\"Logistic Regression Test Accuracy: \" + str(accuracy_score(y_test, y_pred)))\n",
        "  return accuracy_score(y_train, train_pred), accuracy_score(y_test, y_pred)\n",
        "\n",
        "get_logistic_accuracy_average()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Fi306vv6IXv3"
      },
      "outputs": [],
      "source": [
        "one = []\n",
        "two = []\n",
        "three = []\n",
        "for i in range(50):\n",
        "  one.append(get_logistic_accuracy())\n",
        "  two.append(get_logistic_accuracy_average())\n",
        "  three.append(get_logistic_accuracy_five())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VvN3CS6dJIDB",
        "outputId": "753ab305-3866-4ddd-96f1-89d267a5cc34"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(0.8654936461388083, 0.4932162809257787)\n",
            "(0.5897860647693816, 0.5675569520816967)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.5795328753680079, 0.5652788688138258)"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ],
      "source": [
        "answer = sum(i[0] for i in one)/len(one), sum(i[1] for i in one)/len(one)\n",
        "answerTwo = sum(i[0] for i in two) / len(two), sum(i[1] for i in two) /len(two)\n",
        "answerThree = sum(i[0] for i in three) / len(three), sum(i[1] for i in three)/len(three)\n",
        "\n",
        "print(answer)\n",
        "print(answerTwo)\n",
        "print(answerThree)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CUe2jr30IY2K"
      },
      "source": [
        "**Linear Regression**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bik9ogKL_9aT",
        "outputId": "492cff15-d902-4808-990c-4e4d244fd836"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Linear Regression Train Data RMSE: 14.450811322503059\n",
            "Linear Regression Test Data RMSE: 14.38964833850689\n",
            "Linear Regression Train Data RMSE: 14.397505937755758\n",
            "1273\n",
            "[[-3.9564481 ]\n",
            " [ 3.87659207]\n",
            " [-8.6278631 ]\n",
            " ...\n",
            " [-5.77965127]\n",
            " [ 1.07494803]\n",
            " [ 6.04066035]]\n",
            "     PointDifference\n",
            "147            -23.0\n",
            "30               6.0\n",
            "4               10.0\n",
            "62              21.0\n",
            "128            -30.0\n",
            "..               ...\n",
            "15              19.0\n",
            "35              11.0\n",
            "42              15.0\n",
            "95              22.0\n",
            "50              19.0\n",
            "\n",
            "[1273 rows x 1 columns]\n",
            "Linear Regression Test Data RMSE: 14.27358774889765\n"
          ]
        }
      ],
      "source": [
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.metrics import mean_squared_error\n",
        "# Will run a linear regression on the test and training sets from the \"regression\"\n",
        "# sets from the pbp stats data\n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "#Runs linear regresion on box scores\n",
        "def run_linear_regression():\n",
        "  # Initialize train and test sets\n",
        "  x_train, y_train, x_test, y_test = regression_train_test_data()\n",
        "\n",
        "  #initialize and fit linear regression on training data\n",
        "  lin_reg = LinearRegression()\n",
        "  lin_reg.fit(x_train, y_train)\n",
        "\n",
        "  #Predict with training set\n",
        "  trainPrediction = lin_reg.predict(x_train)\n",
        "  train_lin_mse = mean_squared_error(y_train, trainPrediction)\n",
        "  train_lin_rmse = np.sqrt(train_lin_mse)\n",
        "  print(\"Linear Regression Train Data RMSE: \" + str(train_lin_rmse))\n",
        "\n",
        "  #Predict with test data\n",
        "  prediction = lin_reg.predict(x_test)\n",
        "  # Find Mean squared error for prediction\n",
        "  lin_mse = mean_squared_error(y_test, prediction)\n",
        "  # Find RMSE\n",
        "  lin_rmse = np.sqrt(lin_mse)\n",
        "  print(\"Linear Regression Test Data RMSE: \" + str(lin_rmse))\n",
        "\n",
        "#Runs linear regression on last five game averages\n",
        "def linear_regression_l5():\n",
        "# Initialize train and test sets\n",
        "  x_train, y_train, x_test, y_test = regression_train_test_data_param(lastFive)\n",
        "\n",
        "  #initialize and fit linear regression on training data\n",
        "  lin_reg = LinearRegression()\n",
        "  lin_reg.fit(x_train, y_train)\n",
        "\n",
        "  #Predict with training set\n",
        "  trainPrediction = lin_reg.predict(x_train)\n",
        "  train_lin_mse = mean_squared_error(y_train, trainPrediction)\n",
        "  train_lin_rmse = np.sqrt(train_lin_mse)\n",
        "  print(\"Linear Regression Train Data RMSE: \" + str(train_lin_rmse))\n",
        "\n",
        "  #Predict with test data\n",
        "  prediction = lin_reg.predict(x_test)\n",
        "  # Find Mean squared error for prediction\n",
        "  lin_mse = mean_squared_error(y_test, prediction)\n",
        "  # Find RMSE\n",
        "  lin_rmse = np.sqrt(lin_mse)\n",
        "  print(\"Linear Regression Test Data RMSE: \" + str(lin_rmse))\n",
        "\n",
        "linear_regression_l5()\n",
        "\n",
        "#runs linear regression on season averages\n",
        "def linear_regression_average():\n",
        "# Initialize train and test sets\n",
        "  x_train, y_train, x_test, y_test = regression_train_test_data_param(averageDataCombined)\n",
        "\n",
        "  #initialize and fit linear regression on training data\n",
        "  lin_reg = LinearRegression()\n",
        "  lin_reg.fit(x_train, y_train)\n",
        "\n",
        "  #Predict with training set\n",
        "  trainPrediction = lin_reg.predict(x_train)\n",
        "  train_lin_mse = mean_squared_error(y_train, trainPrediction)\n",
        "  train_lin_rmse = np.sqrt(train_lin_mse)\n",
        "  print(\"Linear Regression Train Data RMSE: \" + str(train_lin_rmse))\n",
        "\n",
        "  #Predict with test data\n",
        "  prediction = lin_reg.predict(x_test)\n",
        "  # Find Mean squared error for prediction\n",
        "  lin_mse = mean_squared_error(y_test, prediction)\n",
        "  print(len(y_test))\n",
        "  print(prediction)\n",
        "  print(y_test)\n",
        "  # Find RMSE\n",
        "  lin_rmse = np.sqrt(lin_mse)\n",
        "  print(\"Linear Regression Test Data RMSE: \" + str(lin_rmse))\n",
        "\n",
        "linear_regression_average()\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KLzFeSm2IbfT"
      },
      "source": [
        "**Random Forest**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FY5QEfv2AA6o"
      },
      "outputs": [],
      "source": [
        "from sklearn.ensemble import RandomForestRegressor\n",
        "\n",
        "# Will run a random forest regressor from the assignment 2 notes using the \n",
        "# \"regression\" training and testing data on PBP data\n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "\n",
        "#runs forest regressor on box score data\n",
        "def run_forest_regressor():\n",
        "  #Initialize train and test data\n",
        "  x_train, y_train, x_test, y_test = regression_train_test_data()\n",
        "  #Initialize random forest regressor and fit on training data\n",
        "  forestReg = RandomForestRegressor()\n",
        "  forestReg.fit(x_train, y_train.values.ravel())\n",
        "\n",
        "  #Predict using training data\n",
        "  predict = forestReg.predict(x_train)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  tree_mse = mean_squared_error(y_train, predict)\n",
        "  tree_rmse = np.sqrt(tree_mse)\n",
        "\n",
        "  #Predict using testing data\n",
        "  test_predict = forestReg.predict(x_test)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  test_tree_mse = mean_squared_error(y_test, test_predict)\n",
        "  test_tree_rmse = np.sqrt(test_tree_mse)\n",
        "  print(\"Random Forest Train Data RMSE: \" + str(tree_rmse))\n",
        "  print(\"Random Forest Test Data RMSE: \" + str(test_tree_rmse))\n",
        "\n",
        "#Runs forest regressor on seasonal averages\n",
        "def run_forest_regressor_average():\n",
        "  #Initialize train and test data\n",
        "  x_train, y_train, x_test, y_test = regression_train_test_data_param(averageDataCombined)\n",
        "  #Initialize random forest regressor and fit on training data\n",
        "  forestReg = RandomForestRegressor()\n",
        "  forestReg.fit(x_train, y_train.values.ravel())\n",
        "\n",
        "  #Predict using training data\n",
        "  predict = forestReg.predict(x_train)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  tree_mse = mean_squared_error(y_train, predict)\n",
        "  tree_rmse = np.sqrt(tree_mse)\n",
        "\n",
        "  #Predict using testing data\n",
        "  test_predict = forestReg.predict(x_test)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  test_tree_mse = mean_squared_error(y_test, test_predict)\n",
        "  test_tree_rmse = np.sqrt(test_tree_mse)\n",
        "  print(\"Random Forest Train Data RMSE: \" + str(tree_rmse))\n",
        "  print(\"Random Forest Test Data RMSE: \" + str(test_tree_rmse))\n",
        "\n",
        "#Runs forest regressor on last five games\n",
        "def run_forest_regressor_l5():\n",
        "  #Initialize train and test data\n",
        "  x_train, y_train, x_test, y_test = regression_train_test_data_param(lastFive)\n",
        "  #Initialize random forest regressor and fit on training data\n",
        "  forestReg = RandomForestRegressor()\n",
        "  forestReg.fit(x_train, y_train.values.ravel())\n",
        "\n",
        "  #Predict using training data\n",
        "  predict = forestReg.predict(x_train)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  tree_mse = mean_squared_error(y_train, predict)\n",
        "  tree_rmse = np.sqrt(tree_mse)\n",
        "\n",
        "  #Predict using testing data\n",
        "  test_predict = forestReg.predict(x_test)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  test_tree_mse = mean_squared_error(y_test, test_predict)\n",
        "  test_tree_rmse = np.sqrt(test_tree_mse)\n",
        "  print(\"Random Forest Train Data RMSE: \" + str(tree_rmse))\n",
        "  print(\"Random Forest Test Data RMSE: \" + str(test_tree_rmse))\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C_driXlu0JTi",
        "outputId": "00d311ce-a9de-405c-a7b6-e53cc4f83746"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Random Forest Train Data RMSE: 2.93575070457515\n",
            "Random Forest Test Data RMSE: 19.06495367367593\n"
          ]
        }
      ],
      "source": [
        "run_forest_regressor()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1zzDb9qJLIZp"
      },
      "source": [
        "**Show results from each of the above models**\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0Tm4AVq-KF_Z",
        "outputId": "d387942a-ca09-48c8-8b62-9b8b8458fc0d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Linear Regression Train Data RMSE: 7.009422856567372\n",
            "Linear Regression Test Data RMSE: 27.941754263276735\n",
            "Random Forest Train Data RMSE: 2.951075841401044\n",
            "Random Forest Test Data RMSE: 19.396125930641883\n"
          ]
        }
      ],
      "source": [
        "get_logistic_accuracy()\n",
        "run_linear_regression()\n",
        "run_forest_regressor()"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Sentiment Analysis\n"
      ],
      "metadata": {
        "id": "UFs7XVmJFP2Y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Coded by Rongtian (Ron) Yue\n",
        "import nltk\n",
        "import os\n",
        "import re\n",
        "from tqdm import tqdm\n",
        "import pandas as pd\n",
        "\n",
        "nltk.download('vader_lexicon')\n",
        "from nltk.sentiment.vader import SentimentIntensityAnalyzer\n",
        "\n",
        "# Create instance of sentiment analyzer\n",
        "sia = SentimentIntensityAnalyzer()"
      ],
      "metadata": {
        "id": "x8eWSA5ACkQu",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8a0cbaab-4090-40ec-8314-bc29187c2bb4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[nltk_data] Downloading package vader_lexicon to /root/nltk_data...\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Download data\n",
        "#import requests\n",
        "#request = requests.get(\"https://drive.google.com/uc?export=download&id=1eL5lOaUmGxFX4ZUVGW_M6-a_RqlhgHEW\")\n",
        "\n",
        "'''\n",
        "with open(\"data.zip\", \"wb\") as file:\n",
        "    file.write(request.content)\n",
        "\n",
        "# Unzip data\n",
        "import zipfile\n",
        "with zipfile.ZipFile('data.zip') as zip:\n",
        "    zip.extractall('data')\n",
        "    '''"
      ],
      "metadata": {
        "id": "da0Zcj7uCo_N",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "390d4544-6d66-422f-b1d8-bee74879f748"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'\\nwith open(\"data.zip\", \"wb\") as file:\\n    file.write(request.content)\\n\\n# Unzip data\\nimport zipfile\\nwith zipfile.ZipFile(\\'data.zip\\') as zip:\\n    zip.extractall(\\'data\\')\\n    '"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Turn CSV into dataframe\n",
        "df = pd.read_csv(\"https://drive.google.com/uc?export=download&id=1eL5lOaUmGxFX4ZUVGW_M6-a_RqlhgHEW\") #celtics hawks april 27 2023\n",
        "\n",
        "#print head of dataframe\n",
        "df.head()"
      ],
      "metadata": {
        "id": "NrDhsAeiC3n3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "295de316-7e5d-48cc-881a-aa50c3abcde9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "             AuthorID                   Author                 Date  \\\n",
              "0  707034809561776169  WestCoastFUNko831🍀#3726  04/27/2023 12:00 AM   \n",
              "1  935685933896962079                Capo#3888  04/27/2023 12:01 AM   \n",
              "2  289438506563010561              Sarnai#1867  04/27/2023 12:01 AM   \n",
              "3  935685933896962079                Capo#3888  04/27/2023 12:01 AM   \n",
              "4  724695205957206197  DefinitelyNotYosh.#3366  04/27/2023 12:01 AM   \n",
              "\n",
              "                                             Content Attachments Reactions  \n",
              "0                   Wow is over green making 3s lmao         NaN       NaN  \n",
              "1  Yeah if warriors advance over the kings then w...         NaN       NaN  \n",
              "2  Goes without saying but if the Kings lose toni...         NaN       NaN  \n",
              "3  If we gotta watch more warriors basketball at ...         NaN       NaN  \n",
              "4                      JIMMY BUTLER’S DOING IT AGAIN         NaN       NaN  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-fec2a200-caaa-4dee-9451-b77bb1505c8b\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>AuthorID</th>\n",
              "      <th>Author</th>\n",
              "      <th>Date</th>\n",
              "      <th>Content</th>\n",
              "      <th>Attachments</th>\n",
              "      <th>Reactions</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>707034809561776169</td>\n",
              "      <td>WestCoastFUNko831🍀#3726</td>\n",
              "      <td>04/27/2023 12:00 AM</td>\n",
              "      <td>Wow is over green making 3s lmao</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>935685933896962079</td>\n",
              "      <td>Capo#3888</td>\n",
              "      <td>04/27/2023 12:01 AM</td>\n",
              "      <td>Yeah if warriors advance over the kings then w...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>289438506563010561</td>\n",
              "      <td>Sarnai#1867</td>\n",
              "      <td>04/27/2023 12:01 AM</td>\n",
              "      <td>Goes without saying but if the Kings lose toni...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>935685933896962079</td>\n",
              "      <td>Capo#3888</td>\n",
              "      <td>04/27/2023 12:01 AM</td>\n",
              "      <td>If we gotta watch more warriors basketball at ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>724695205957206197</td>\n",
              "      <td>DefinitelyNotYosh.#3366</td>\n",
              "      <td>04/27/2023 12:01 AM</td>\n",
              "      <td>JIMMY BUTLER’S DOING IT AGAIN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-fec2a200-caaa-4dee-9451-b77bb1505c8b')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-fec2a200-caaa-4dee-9451-b77bb1505c8b button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-fec2a200-caaa-4dee-9451-b77bb1505c8b');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Remove unnecessary elements from dataframe\n",
        "df.drop(['AuthorID'], inplace=True, axis=1)\n",
        "df.drop(['Author'], inplace=True, axis=1)\n",
        "df.drop(['Date'], inplace=True, axis=1)\n",
        "df.drop(['Attachments'], inplace=True, axis=1)\n",
        "df.drop(['Reactions'], inplace=True, axis=1)\n",
        "df.head()"
      ],
      "metadata": {
        "id": "1KMLhif1C6Rq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "eb93e8dd-a41f-4183-99f4-fdbf54d1af08"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                             Content\n",
              "0                   Wow is over green making 3s lmao\n",
              "1  Yeah if warriors advance over the kings then w...\n",
              "2  Goes without saying but if the Kings lose toni...\n",
              "3  If we gotta watch more warriors basketball at ...\n",
              "4                      JIMMY BUTLER’S DOING IT AGAIN"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b4cf130d-6e2a-4837-b37a-98145b9ac920\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Content</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Wow is over green making 3s lmao</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Yeah if warriors advance over the kings then w...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Goes without saying but if the Kings lose toni...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>If we gotta watch more warriors basketball at ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>JIMMY BUTLER’S DOING IT AGAIN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b4cf130d-6e2a-4837-b37a-98145b9ac920')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-b4cf130d-6e2a-4837-b37a-98145b9ac920 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-b4cf130d-6e2a-4837-b37a-98145b9ac920');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# The first 5 looked good but inspecting more messages \n",
        "# shows us that there are some boxes with empty (NaN)\n",
        "# and also we see discord bot commands (!scores)\n",
        "df[5:11]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "m_JYjmDzDBdO",
        "outputId": "9f44c953-c3bd-41c6-f415-9ba899c164e5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                              Content\n",
              "5   Y'all act like the Warriors are fuckin superhe...\n",
              "6                         Jimmy butler is a superhero\n",
              "7                                     These refs bruh\n",
              "8                                             !scores\n",
              "9                                                 NaN\n",
              "10                                                Wow"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1ba2b66a-35bf-41c8-947c-8981cf92f72b\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Content</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Y'all act like the Warriors are fuckin superhe...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Jimmy butler is a superhero</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>These refs bruh</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>!scores</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>Wow</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1ba2b66a-35bf-41c8-947c-8981cf92f72b')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-1ba2b66a-35bf-41c8-947c-8981cf92f72b button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-1ba2b66a-35bf-41c8-947c-8981cf92f72b');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Drop NaN and reset index\n",
        "df.dropna(how='any', inplace=True)\n",
        "\n",
        "# Drop discord bot commands\n",
        "df = df[~df['Content'].astype(str).str.startswith('!')]\n",
        "\n",
        "# Drop Discord mentions\n",
        "df = df[~df['Content'].astype(str).str.startswith('@')]\n",
        "\n",
        "# Drop links\n",
        "df = df[~df['Content'].astype(str).str.startswith('http')]\n",
        "\n",
        "# Drop Discord emotes\n",
        "df = df[~df['Content'].astype(str).str.startswith(':')]\n",
        "\n",
        "# Drop small sentences\n",
        "df = df[df['Content'].map(len) > 10]\n",
        "\n",
        "# Reset Index numbers after drops\n",
        "df = df.reset_index()\n",
        "df.drop(['index'], inplace=True, axis=1)\n",
        "df[5:11]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "_8n3iW3GDCo3",
        "outputId": "e7a04bd1-7b45-4838-ca12-0fe944266352"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                              Content\n",
              "5   Y'all act like the Warriors are fuckin superhe...\n",
              "6                         Jimmy butler is a superhero\n",
              "7                                     These refs bruh\n",
              "8                Draymond missed the kick to the nuts\n",
              "9                            Jimmy buckets tied it up\n",
              "10                       Wish I could watch that game"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1daa446f-dfea-4cb5-b17d-8c13197354b8\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Content</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Y'all act like the Warriors are fuckin superhe...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Jimmy butler is a superhero</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>These refs bruh</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Draymond missed the kick to the nuts</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>Jimmy buckets tied it up</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>Wish I could watch that game</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1daa446f-dfea-4cb5-b17d-8c13197354b8')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-1daa446f-dfea-4cb5-b17d-8c13197354b8 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-1daa446f-dfea-4cb5-b17d-8c13197354b8');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.apply(lambda x: x.str.lower().str.strip() if isinstance(x, object) else x)\n",
        "df['scores'] = df['Content'].apply(lambda review: sia.polarity_scores(review))\n",
        "df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "mm_UvrgMDD8t",
        "outputId": "b34a4c89-72ac-4e03-d2c1-1f2dff134970"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                             Content  \\\n",
              "0                   Wow is over green making 3s lmao   \n",
              "1  Yeah if warriors advance over the kings then w...   \n",
              "2  Goes without saying but if the Kings lose toni...   \n",
              "3  If we gotta watch more warriors basketball at ...   \n",
              "4                      JIMMY BUTLER’S DOING IT AGAIN   \n",
              "\n",
              "                                              scores  \n",
              "0  {'neg': 0.0, 'neu': 0.394, 'pos': 0.606, 'comp...  \n",
              "1  {'neg': 0.0, 'neu': 0.872, 'pos': 0.128, 'comp...  \n",
              "2  {'neg': 0.163, 'neu': 0.689, 'pos': 0.148, 'co...  \n",
              "3  {'neg': 0.0, 'neu': 1.0, 'pos': 0.0, 'compound...  \n",
              "4  {'neg': 0.0, 'neu': 1.0, 'pos': 0.0, 'compound...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-e7cd3ac8-b5f5-4611-b219-5aa0f9b03352\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Content</th>\n",
              "      <th>scores</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Wow is over green making 3s lmao</td>\n",
              "      <td>{'neg': 0.0, 'neu': 0.394, 'pos': 0.606, 'comp...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Yeah if warriors advance over the kings then w...</td>\n",
              "      <td>{'neg': 0.0, 'neu': 0.872, 'pos': 0.128, 'comp...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Goes without saying but if the Kings lose toni...</td>\n",
              "      <td>{'neg': 0.163, 'neu': 0.689, 'pos': 0.148, 'co...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>If we gotta watch more warriors basketball at ...</td>\n",
              "      <td>{'neg': 0.0, 'neu': 1.0, 'pos': 0.0, 'compound...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>JIMMY BUTLER’S DOING IT AGAIN</td>\n",
              "      <td>{'neg': 0.0, 'neu': 1.0, 'pos': 0.0, 'compound...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e7cd3ac8-b5f5-4611-b219-5aa0f9b03352')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-e7cd3ac8-b5f5-4611-b219-5aa0f9b03352 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-e7cd3ac8-b5f5-4611-b219-5aa0f9b03352');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 50
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df['compound']  = df['scores'].apply(lambda score_dict: score_dict['compound'])\n",
        "\n",
        "df.head()\n",
        "avg_sentiment = df[\"compound\"].mean()\n",
        "print(avg_sentiment)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fDGu-iH9DFB5",
        "outputId": "6956b665-4157-445a-8a64-afd56c245902"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.05573274634694642\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v7_JnqqYIfEh"
      },
      "source": [
        "Neural Network using full season averages"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "k3sb2fj4FJSj"
      },
      "outputs": [],
      "source": [
        "## This function will compile and fit a very simple NN using the training data\n",
        "## pulled from the classification data function above. \n",
        "\n",
        "#Coded by Davin Stephens\n",
        "def fit_nn():\n",
        "  # Grab train and test data\n",
        "  x_train, y_train, x_test, y_test = classification_train_test_data()\n",
        "\n",
        "  # #Reshape the data to be used in the model\n",
        "  # x_train = x_train.values.reshape(x_train.shape[0], 1, 28)\n",
        "\n",
        "  model = tf.keras.models.Sequential()\n",
        "  # model.add(LSTM(50, activation=tf.nn.relu6, input_shape=(1, 28)))\n",
        "  # model.add(Dense(1, activation=tf.nn.sigmoid))\n",
        "  # model.summary()\n",
        "\n",
        "  model.add(Dense(32, activation=tf.nn.relu6, input_dim=len(x_train.columns)))\n",
        "  model.add(Dense(32, activation=tf.nn.relu6))\n",
        "  model.add(Dense(1))\n",
        "\n",
        "  #Compile the model using adam as the optimizer and MSE as the loss function\n",
        "  model.compile(optimizer='adam', loss='binary_crossentropy', metrics='accuracy')\n",
        "  model.fit(x_train, y_train, epochs=100, batch_size=32)\n",
        "  return model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wCtKu0sSGNY6",
        "outputId": "81c3ce99-86a4-4cf2-945d-533679b61fe2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "160/160 [==============================] - 1s 2ms/step - loss: 7.8136 - accuracy: 0.4935\n",
            "Epoch 2/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 3/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 4/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 5/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 6/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 7/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 8/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 9/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 10/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 11/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 12/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 13/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 14/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 15/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 16/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 17/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 18/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 19/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 20/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 21/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 22/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 23/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 24/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 25/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 26/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 27/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 28/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 29/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 30/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 31/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 32/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 33/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 34/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 35/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 36/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 37/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 38/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 39/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 40/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 41/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 42/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 43/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 44/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 45/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 46/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 47/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 48/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 49/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 50/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 51/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 52/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 53/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 54/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 55/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 56/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 57/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 58/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 59/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 60/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 61/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 62/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 63/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 64/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 65/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 66/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 67/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 68/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 69/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 70/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 71/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 72/100\n",
            "160/160 [==============================] - 0s 3ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 73/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 74/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 75/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 76/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 77/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 78/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 79/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 80/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 81/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 82/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 83/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 84/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 85/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 86/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 87/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 88/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 89/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 90/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 91/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 92/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 93/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 94/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 95/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 96/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 97/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 98/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 99/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "Epoch 100/100\n",
            "160/160 [==============================] - 0s 2ms/step - loss: 7.8135 - accuracy: 0.4935\n",
            "40/40 [==============================] - 0s 1ms/step\n",
            "NN accuracy score of: 0.5283320031923384\n"
          ]
        }
      ],
      "source": [
        "## This function will get the accuracy of the NN using our test data (average data over last season)\n",
        "\n",
        "# Coded by Davin Stephens\n",
        "def nn_accuracy():\n",
        "  x_train, y_train, x_test, y_test = classification_train_test_data()\n",
        "  model = fit_nn()\n",
        "  x_test = np.asarray(x_test).astype('float32')\n",
        "  # x_test = x_test.reshape(x_test.shape[0], 1, 28)\n",
        "  predictions = model.predict(x_test)\n",
        "  predictions = [0 if val < 0.5 else 1 for val in predictions]\n",
        "\n",
        "  print(\"NN accuracy score of: \" + str(accuracy_score(y_test, predictions)))\n",
        "\n",
        "nn_accuracy()"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Our results using the full season averages were dissapointing, so we moved to using five day averages"
      ],
      "metadata": {
        "id": "KVoQN9mnRG4r"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**5 Day Average Prediction (NEW FOR MILESTONE 2 CODE DEMO)**"
      ],
      "metadata": {
        "id": "gKUgGDtPXjQB"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N3QFkOpoAtz4"
      },
      "source": [
        "# 5 Day average (Wins/Losses)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ruLCZGH6BKOy"
      },
      "source": [
        "Importing box score data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "RiOVvUi3A_N6"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "\n",
        "## Coded by Davin Stephens\n",
        "def five_day_average_data():\n",
        "  # access file from github containing nba game box scores over the last 10 seasons\n",
        "  nba_file = 'https://raw.githubusercontent.com/dcsteph7/Data-Capstone/main/nba_games1.csv'\n",
        "  df = pd.read_csv(nba_file, index_col = 0)\n",
        "  # sort values chronologically\n",
        "  df = df.sort_values(\"date\")\n",
        "  df = df.reset_index(drop=True)\n",
        "\n",
        "  # remove duplicated stats and null values\n",
        "  del df[\"mp.1\"]\n",
        "  del df[\"mp_opp.1\"]\n",
        "  del df[\"index_opp\"]\n",
        "  df[\"won\"].value_counts()\n",
        "  nulls = pd.isnull(df).sum()\n",
        "  nulls = nulls[nulls > 0]\n",
        "  valid_columns = df.columns[~df.columns.isin(nulls.index)]\n",
        "  df = df[valid_columns].copy()\n",
        "  # Convert won column to 1 for a win, 0 for a loss\n",
        "  df['won'] = df['won'].apply(lambda x: 1 if x else 0)\n",
        "  # compute the rolling averages of the last 5 games played for each game\n",
        "  team_stats = df.groupby('team')[['fg', 'fga', '3p', '3pa', '3p_max', 'ft', 'fta', 'orb', 'drb', 'ast', 'stl', 'blk',\n",
        "                                  'tov', 'fg_opp', 'fga_opp', '3p_opp', '3pa_opp', '3p_max_opp', 'ft_opp', 'fta_opp', \n",
        "                                  'orb_opp', 'drb_opp', 'ast_opp', 'stl_opp', 'blk_opp', \n",
        "                                  'tov_opp', 'won']].rolling(5, min_periods=1).mean().reset_index()\n",
        "\n",
        "  # merge rolling values with normal using the original index\n",
        "  merged_df = df.merge(team_stats.reset_index()[['level_1', 'fg', 'fga', '3p', '3pa', '3p_max', 'ft', 'fta', 'orb', 'drb',\n",
        "                                                'ast', 'stl', 'blk', 'tov', 'fg_opp', 'fga_opp', '3p_opp', '3pa_opp',\n",
        "                                                '3p_max_opp', 'ft_opp', 'fta_opp', 'orb_opp', 'drb_opp', 'ast_opp',\n",
        "                                                'stl_opp', 'blk_opp', 'tov_opp', 'won']], left_index=True, right_on='level_1', suffixes=('', '_r'))\n",
        "  df = merged_df\n",
        "\n",
        "  # calculate team ids if they are necessary (currently not in use)\n",
        "  teams = list(set(df['team'].values))\n",
        "  teamToId = {t: i for i, t in enumerate(teams)}\n",
        "  homeId = [teamToId[id] for id in list(df['team'].values)]\n",
        "  df['home_id'] = homeId \n",
        "  awayId = [teamToId[id] for id in list(df['team_opp'].values)]\n",
        "  df['away_id'] = awayId\n",
        "\n",
        "  # set target and testing data\n",
        "  # Initially \"won\" will be the label we are testing for\n",
        "  target = df['won']\n",
        "  df = df[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r', 'won_r']]\n",
        "  return df, target\n",
        "\n",
        "  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UcoOCwngBsr2"
      },
      "source": [
        "Getting data ready for machine learning"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tj5FP-3DBYIn"
      },
      "outputs": [],
      "source": [
        "#grab datasets that will be used for x and y data\n",
        "df, target = five_day_average_data()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "imNkMMrsDBQG",
        "outputId": "47e8ab49-c76e-4cdb-b7ca-0a53c9bacd3e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Logistic Regression Train Accuracy: 0.7196726103478515\n",
            "Logistic Regression Test Accuracy: 0.7199922057677319\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.7196726103478515, 0.7199922057677319)"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ],
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "#Normalize the data to be used in training the model\n",
        "xx_train = tf.keras.utils.normalize(df, axis=1)\n",
        "#Split the data into test and training sets and an 80/20 split\n",
        "X_train, X_test, y_train, y_test = train_test_split(xx_train, target, test_size=.2)\n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "# Will run logistic regression on the 5 day average data set\n",
        "def logistic_regression_5_day():\n",
        "\n",
        "  #Initialize and fit simple logistic regression\n",
        "  #Parameters taken from GridSearchCV\n",
        "  logis = LogisticRegression(C=10.0, penalty='l1', solver='liblinear', max_iter=2000)\n",
        "  logis.fit(X_train, y_train.values.ravel())\n",
        "\n",
        "  # Predict on training data\n",
        "  train_pred = logis.predict(X_train)\n",
        "\n",
        "  # Predict on testing data \n",
        "  y_pred = logis.predict(X_test)\n",
        "\n",
        "  # Print accuracy score\n",
        "  print(\"Logistic Regression Train Accuracy: \" + str(accuracy_score(y_train, train_pred)))\n",
        "  print(\"Logistic Regression Test Accuracy: \" + str(accuracy_score(y_test, y_pred)))\n",
        "  return accuracy_score(y_train, train_pred), accuracy_score(y_test, y_pred)\n",
        "\n",
        "logistic_regression_5_day()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_zcqnjmJDoPm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "de0b35be-8d01-40c9-e726-a9f95cb46667"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Logistic Regression Train Accuracy: 0.7196726103478515\n",
            "Logistic Regression Test Accuracy: 0.7199922057677319\n"
          ]
        }
      ],
      "source": [
        "#Hold the results from logistic regression to be used in a plot later \n",
        "train_lR, test_lR = logistic_regression_5_day()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tLyl95VxE9J_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a49f774b-d935-4f66-8f73-8f7d527409c2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.7031569716457177\n",
            "0.7155105222135619\n"
          ]
        }
      ],
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "#Random forest classifier to predict wins/losses on 5 day averages\n",
        "#Coded by Tyler Wulff\n",
        "def Random_Forest_5_Day():\n",
        "\n",
        "  forest = RandomForestClassifier(oob_score=True)\n",
        "  #Fit random forest on training data\n",
        "  forest.fit(X_train, y_train)\n",
        "  forest_train = forest.oob_score_\n",
        "\n",
        "  #Predict on testing set\n",
        "  y_pred = forest.predict(X_test)\n",
        "  forest_test = accuracy_score(y_test, y_pred)\n",
        "  return forest_train, forest_test\n",
        "\n",
        "#Hold values for plot later to compare models\n",
        "forest_train, forest_test = Random_Forest_5_Day()\n",
        "print(forest_train)\n",
        "print(forest_test)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Hyperparamater tuning on our logistic regression model increased accuracy around 3 percent"
      ],
      "metadata": {
        "id": "-wWIxch9RcjV"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "iIRoVZ_xdKJo"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import GridSearchCV\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "#Coded by Tyler Wulff\n",
        "#Use GridSearchCV to optimize hyperparamaters for logistic regression\n",
        "# param_grid = {\n",
        "#     'penalty' : ['l1','l2', 'elasticnet'], \n",
        "#     'C'       : [100, 10, 1.0],\n",
        "#     'solver'  : ['newton-cg', 'lbfgs', 'liblinear']\n",
        "# }\n",
        "# log = LogisticRegression(max_iter=2000)\n",
        "# grid_search = GridSearchCV(log, param_grid, cv=5, scoring='accuracy')\n",
        "\n",
        "# grid_search.fit(X_train, y_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "29cTQTqwhhl-"
      },
      "outputs": [],
      "source": [
        "#Paramaters to be used are shown below ('C': 1.0, 'penalty': 'l1', 'solver': 'liblinear')\n",
        "# grid_search.best_params_"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.to_csv('check.csv')"
      ],
      "metadata": {
        "id": "CBR0uZmdT8-Q"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s4W32kiJB9N-"
      },
      "source": [
        "Splitting data and training model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VCLDxcdyB8L1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "79b75c02-8d75-40fb-8085-c534a2ad7476"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "642/642 [==============================] - 4s 3ms/step - loss: 0.6811 - accuracy: 0.5620\n",
            "Epoch 2/50\n",
            "642/642 [==============================] - 5s 8ms/step - loss: 0.6130 - accuracy: 0.6657\n",
            "Epoch 3/50\n",
            "642/642 [==============================] - 4s 6ms/step - loss: 0.5854 - accuracy: 0.6903\n",
            "Epoch 4/50\n",
            "642/642 [==============================] - 2s 4ms/step - loss: 0.5772 - accuracy: 0.6942\n",
            "Epoch 5/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5752 - accuracy: 0.6937\n",
            "Epoch 6/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5720 - accuracy: 0.6961\n",
            "Epoch 7/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5705 - accuracy: 0.6991\n",
            "Epoch 8/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5690 - accuracy: 0.7007\n",
            "Epoch 9/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5671 - accuracy: 0.7010\n",
            "Epoch 10/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5656 - accuracy: 0.7033\n",
            "Epoch 11/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5627 - accuracy: 0.7052\n",
            "Epoch 12/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5627 - accuracy: 0.7065\n",
            "Epoch 13/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5606 - accuracy: 0.7043\n",
            "Epoch 14/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5601 - accuracy: 0.7039\n",
            "Epoch 15/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5603 - accuracy: 0.7065\n",
            "Epoch 16/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5596 - accuracy: 0.7086\n",
            "Epoch 17/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5563 - accuracy: 0.7106\n",
            "Epoch 18/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5564 - accuracy: 0.7105\n",
            "Epoch 19/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5547 - accuracy: 0.7119\n",
            "Epoch 20/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5537 - accuracy: 0.7122\n",
            "Epoch 21/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5578 - accuracy: 0.7071\n",
            "Epoch 22/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5527 - accuracy: 0.7138\n",
            "Epoch 23/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5557 - accuracy: 0.7091\n",
            "Epoch 24/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5537 - accuracy: 0.7104\n",
            "Epoch 25/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5539 - accuracy: 0.7123\n",
            "Epoch 26/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5511 - accuracy: 0.7135\n",
            "Epoch 27/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5507 - accuracy: 0.7134\n",
            "Epoch 28/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5496 - accuracy: 0.7126\n",
            "Epoch 29/50\n",
            "642/642 [==============================] - 2s 4ms/step - loss: 0.5500 - accuracy: 0.7126\n",
            "Epoch 30/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5486 - accuracy: 0.7158\n",
            "Epoch 31/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5509 - accuracy: 0.7103\n",
            "Epoch 32/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5493 - accuracy: 0.7143\n",
            "Epoch 33/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5513 - accuracy: 0.7098\n",
            "Epoch 34/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5465 - accuracy: 0.7128\n",
            "Epoch 35/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5492 - accuracy: 0.7131\n",
            "Epoch 36/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5468 - accuracy: 0.7151\n",
            "Epoch 37/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5495 - accuracy: 0.7124\n",
            "Epoch 38/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5454 - accuracy: 0.7139\n",
            "Epoch 39/50\n",
            "642/642 [==============================] - 2s 3ms/step - loss: 0.5464 - accuracy: 0.7145\n",
            "Epoch 40/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5465 - accuracy: 0.7148\n",
            "Epoch 41/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5476 - accuracy: 0.7110\n",
            "Epoch 42/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5453 - accuracy: 0.7170\n",
            "Epoch 43/50\n",
            "642/642 [==============================] - 1s 2ms/step - loss: 0.5443 - accuracy: 0.7146\n",
            "Epoch 44/50\n",
            "642/642 [==============================] - 2s 2ms/step - loss: 0.5455 - accuracy: 0.7160\n",
            "Epoch 45/50\n",
            "642/642 [==============================] - 3s 5ms/step - loss: 0.5458 - accuracy: 0.7145\n",
            "Epoch 46/50\n",
            "642/642 [==============================] - 6s 9ms/step - loss: 0.5472 - accuracy: 0.7148\n",
            "Epoch 47/50\n",
            "642/642 [==============================] - 3s 5ms/step - loss: 0.5453 - accuracy: 0.7127\n",
            "Epoch 48/50\n",
            "642/642 [==============================] - 4s 6ms/step - loss: 0.5472 - accuracy: 0.7146\n",
            "Epoch 49/50\n",
            "642/642 [==============================] - 5s 7ms/step - loss: 0.5428 - accuracy: 0.7179\n",
            "Epoch 50/50\n",
            "642/642 [==============================] - 4s 6ms/step - loss: 0.5446 - accuracy: 0.7136\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f8471a548e0>"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "#Coded by Davin Stephens\n",
        "# normalize and split data into training and testing\n",
        "xx_train = tf.keras.utils.normalize(df, axis=1)\n",
        "X_train, X_test, y_train, y_test = train_test_split(xx_train, target, test_size=.2)\n",
        "\n",
        "# neural network model, not fully optimized yet \n",
        "model = tf.keras.models.Sequential()\n",
        "model.add(tf.keras.layers.Dense(64, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(32, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(1, activation=tf.nn.sigmoid))\n",
        "model.compile(optimizer='adam', loss='binary_crossentropy', metrics='accuracy')\n",
        "model.fit(X_train, y_train, epochs=50, batch_size=32)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hIDBCqbcCLvR"
      },
      "source": [
        "Evaluation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8K_tqjHbEYA6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3009e3fc-4d71-4d3c-c476-a6df36f017ea"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.716457188129425"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "#Grab accuracies on training data from neural network\n",
        "_, train_nn = model.evaluate(X_train, y_train, verbose = 0)\n",
        "train_nn"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "##Summary of neural network model\n",
        "print(model.summary())"
      ],
      "metadata": {
        "id": "5piNqRlhggxK",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "95cf3659-8e7e-46cf-abe8-d41cc79f2dab"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense (Dense)               (None, 64)                1856      \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 32)                2080      \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 1)                 33        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 3,969\n",
            "Trainable params: 3,969\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YUtqhXz8CMYJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 507
        },
        "outputId": "927c361e-320d-49c9-e314-936a892710af"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "161/161 [==============================] - 0s 1ms/step\n",
            "NN Test Accuracy: 0.7213561964146532\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "#Coded by Davin Stephens\n",
        "# show accuracy on test set for neural network\n",
        "y_pred = model.predict(X_test)\n",
        "y_pred = [0 if val < 0.5 else 1 for val in y_pred]\n",
        "nn_test_accuracy = accuracy_score(y_test, y_pred)\n",
        "print(\"NN Test Accuracy: \" + str(nn_test_accuracy))\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# calculate the confusion matrix\n",
        "cm = confusion_matrix(y_test, y_pred)\n",
        "\n",
        "# plot the confusion matrix\n",
        "labels = ['Win', 'Loss']\n",
        "sns.heatmap(cm, annot=True, cmap='Blues', fmt='d', xticklabels=labels, yticklabels=labels)\n",
        "\n",
        "# add title and axis labels\n",
        "plt.title('Confusion Matrix')\n",
        "plt.xlabel('Predicted')\n",
        "plt.ylabel('Actual')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WKBvdDDlE6o_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 508
        },
        "outputId": "f6f2587f-df8f-4de2-8820-a876225487a3"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "#Coded by Tyler Wulff\n",
        "# Plot model accuracies showing train and test accuracies for each model\n",
        "def plot_model_accuracies():\n",
        "\n",
        "  # Models that will be plotted\n",
        "  models = (\"Neural Network\", \"Random Forest\", \"Logistic Regression\")\n",
        "  # Each model will plot the train and test accuracies that were calculated above\n",
        "  accuracies = {\n",
        "      'Train': (train_nn, forest_train, train_lR),\n",
        "      'Test': (nn_test_accuracy, forest_test, test_lR)\n",
        "  }\n",
        "\n",
        "  x = np.arange(len(models))\n",
        "  width = .5\n",
        "  multiplier = 0\n",
        "  fig, ax = plt.subplots(layout='constrained')\n",
        "\n",
        "  # loop through the accuracies and add plot\n",
        "  for attribute, measurement in accuracies.items():\n",
        "      offset = width * multiplier\n",
        "      rects = ax.bar(x + offset, measurement, width, label=attribute)\n",
        "      ax.bar_label(rects)\n",
        "      multiplier += 1\n",
        "\n",
        "  # Set title and styles\n",
        "  ax.set_title('Model Accuracies for W/L')\n",
        "  ax.set_xticks([.25, 1.25, 2.25], models)\n",
        "  ax.set_ylim(0, 1)\n",
        "  plt.show()\n",
        "\n",
        "#Blue bars represent training data, orange represnts testing accuracies\n",
        "plot_model_accuracies()"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Spread Prediction"
      ],
      "metadata": {
        "id": "pHBBG0ULpCdk"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Data pre-processing for spread prediction**"
      ],
      "metadata": {
        "id": "cHl9rScioIEB"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ka-9GgHyHB05",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 840
        },
        "outputId": "7650e8de-8088-42b7-c1ca-169b3cf4817c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-2-cfe6acb049fc>:51: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df2['PointDifference'] = df['point_difference']\n",
            "<ipython-input-2-cfe6acb049fc>:54: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df2['PointDifference'][i] = df['point_difference'][count]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       home  fg_r  fga_r  3p_r  3pa_r  3p_max_r  ft_r  fta_r  orb_r  drb_r  \\\n",
              "24810     1  32.0   90.0   8.0   32.0       2.0  12.0   20.0   18.0   21.0   \n",
              "4262      0  36.0   79.0   7.0   20.0       3.0  15.0   22.0   18.0   36.0   \n",
              "5138      1  40.0   85.0   5.0   15.0       2.0  14.0   18.0    9.0   31.0   \n",
              "11223     0  38.0   77.0   3.0   13.0       1.0  12.0   31.0   15.0   31.0   \n",
              "865       1  39.0   75.0   6.0   13.0       3.0  23.0   28.0    7.0   34.0   \n",
              "...     ...   ...    ...   ...    ...       ...   ...    ...    ...    ...   \n",
              "8583      1  39.8   86.2  15.6   40.2       6.6  13.2   16.2   10.8   32.8   \n",
              "1770      0  37.0   84.8  15.0   36.6       4.8  15.2   20.0    9.8   34.6   \n",
              "1771      1  36.2   82.8  15.0   36.6       5.0  15.6   21.4    9.4   34.2   \n",
              "8584      0  39.0   86.2  14.6   41.0       6.0  12.6   16.0    8.8   32.4   \n",
              "8585      1  38.8   87.0  14.6   41.2       5.8  12.0   14.6    9.4   32.8   \n",
              "\n",
              "       ...  orb_opp_r  drb_opp_r  ast_opp_r  stl_opp_r  blk_opp_r  tov_opp_r  \\\n",
              "24810  ...       18.0       36.0       22.0        7.0        5.0       20.0   \n",
              "4262   ...       18.0       21.0       26.0       11.0       10.0       12.0   \n",
              "5138   ...       15.0       31.0       24.0        6.0        5.0       14.0   \n",
              "11223  ...        9.0       31.0       22.0        9.0        5.0       11.0   \n",
              "865    ...        5.0       31.0       25.0        8.0        5.0        8.0   \n",
              "...    ...        ...        ...        ...        ...        ...        ...   \n",
              "8583   ...        9.0       32.0       25.8        6.6        6.0       12.8   \n",
              "1770   ...        9.4       32.8       21.4        8.8        4.6       13.2   \n",
              "1771   ...        8.8       32.4       22.8        9.8        4.0       12.8   \n",
              "8584   ...        9.4       34.2       25.0        5.8        5.6       15.0   \n",
              "8585   ...       10.2       33.8       23.8        6.0        6.0       17.0   \n",
              "\n",
              "       team  team_opp        date  PointDifference  \n",
              "24810   CLE       WAS  2012-10-30               10  \n",
              "4262    DAL       LAL  2012-10-30              -10  \n",
              "5138    LAL       DAL  2012-10-30               -8  \n",
              "11223   BOS       MIA  2012-10-30                8  \n",
              "865     MIA       BOS  2012-10-30               13  \n",
              "...     ...       ...         ...              ...  \n",
              "8583    BOS       GSW  2022-06-10              -10  \n",
              "1770    BOS       GSW  2022-06-13               10  \n",
              "1771    GSW       BOS  2022-06-13               10  \n",
              "8584    GSW       BOS  2022-06-16              -10  \n",
              "8585    BOS       GSW  2022-06-16              -13  \n",
              "\n",
              "[25657 rows x 31 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-091270ce-713f-4140-a2ce-cc7ea565f8b1\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>home</th>\n",
              "      <th>fg_r</th>\n",
              "      <th>fga_r</th>\n",
              "      <th>3p_r</th>\n",
              "      <th>3pa_r</th>\n",
              "      <th>3p_max_r</th>\n",
              "      <th>ft_r</th>\n",
              "      <th>fta_r</th>\n",
              "      <th>orb_r</th>\n",
              "      <th>drb_r</th>\n",
              "      <th>...</th>\n",
              "      <th>orb_opp_r</th>\n",
              "      <th>drb_opp_r</th>\n",
              "      <th>ast_opp_r</th>\n",
              "      <th>stl_opp_r</th>\n",
              "      <th>blk_opp_r</th>\n",
              "      <th>tov_opp_r</th>\n",
              "      <th>team</th>\n",
              "      <th>team_opp</th>\n",
              "      <th>date</th>\n",
              "      <th>PointDifference</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>24810</th>\n",
              "      <td>1</td>\n",
              "      <td>32.0</td>\n",
              "      <td>90.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>32.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>...</td>\n",
              "      <td>18.0</td>\n",
              "      <td>36.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>CLE</td>\n",
              "      <td>WAS</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4262</th>\n",
              "      <td>0</td>\n",
              "      <td>36.0</td>\n",
              "      <td>79.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>36.0</td>\n",
              "      <td>...</td>\n",
              "      <td>18.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>DAL</td>\n",
              "      <td>LAL</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>-10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5138</th>\n",
              "      <td>1</td>\n",
              "      <td>40.0</td>\n",
              "      <td>85.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>...</td>\n",
              "      <td>15.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>24.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>LAL</td>\n",
              "      <td>DAL</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>-8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11223</th>\n",
              "      <td>0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>77.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>...</td>\n",
              "      <td>9.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>BOS</td>\n",
              "      <td>MIA</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>865</th>\n",
              "      <td>1</td>\n",
              "      <td>39.0</td>\n",
              "      <td>75.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>28.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>...</td>\n",
              "      <td>5.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>25.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MIA</td>\n",
              "      <td>BOS</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>13</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8583</th>\n",
              "      <td>1</td>\n",
              "      <td>39.8</td>\n",
              "      <td>86.2</td>\n",
              "      <td>15.6</td>\n",
              "      <td>40.2</td>\n",
              "      <td>6.6</td>\n",
              "      <td>13.2</td>\n",
              "      <td>16.2</td>\n",
              "      <td>10.8</td>\n",
              "      <td>32.8</td>\n",
              "      <td>...</td>\n",
              "      <td>9.0</td>\n",
              "      <td>32.0</td>\n",
              "      <td>25.8</td>\n",
              "      <td>6.6</td>\n",
              "      <td>6.0</td>\n",
              "      <td>12.8</td>\n",
              "      <td>BOS</td>\n",
              "      <td>GSW</td>\n",
              "      <td>2022-06-10</td>\n",
              "      <td>-10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1770</th>\n",
              "      <td>0</td>\n",
              "      <td>37.0</td>\n",
              "      <td>84.8</td>\n",
              "      <td>15.0</td>\n",
              "      <td>36.6</td>\n",
              "      <td>4.8</td>\n",
              "      <td>15.2</td>\n",
              "      <td>20.0</td>\n",
              "      <td>9.8</td>\n",
              "      <td>34.6</td>\n",
              "      <td>...</td>\n",
              "      <td>9.4</td>\n",
              "      <td>32.8</td>\n",
              "      <td>21.4</td>\n",
              "      <td>8.8</td>\n",
              "      <td>4.6</td>\n",
              "      <td>13.2</td>\n",
              "      <td>BOS</td>\n",
              "      <td>GSW</td>\n",
              "      <td>2022-06-13</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1771</th>\n",
              "      <td>1</td>\n",
              "      <td>36.2</td>\n",
              "      <td>82.8</td>\n",
              "      <td>15.0</td>\n",
              "      <td>36.6</td>\n",
              "      <td>5.0</td>\n",
              "      <td>15.6</td>\n",
              "      <td>21.4</td>\n",
              "      <td>9.4</td>\n",
              "      <td>34.2</td>\n",
              "      <td>...</td>\n",
              "      <td>8.8</td>\n",
              "      <td>32.4</td>\n",
              "      <td>22.8</td>\n",
              "      <td>9.8</td>\n",
              "      <td>4.0</td>\n",
              "      <td>12.8</td>\n",
              "      <td>GSW</td>\n",
              "      <td>BOS</td>\n",
              "      <td>2022-06-13</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8584</th>\n",
              "      <td>0</td>\n",
              "      <td>39.0</td>\n",
              "      <td>86.2</td>\n",
              "      <td>14.6</td>\n",
              "      <td>41.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>12.6</td>\n",
              "      <td>16.0</td>\n",
              "      <td>8.8</td>\n",
              "      <td>32.4</td>\n",
              "      <td>...</td>\n",
              "      <td>9.4</td>\n",
              "      <td>34.2</td>\n",
              "      <td>25.0</td>\n",
              "      <td>5.8</td>\n",
              "      <td>5.6</td>\n",
              "      <td>15.0</td>\n",
              "      <td>GSW</td>\n",
              "      <td>BOS</td>\n",
              "      <td>2022-06-16</td>\n",
              "      <td>-10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8585</th>\n",
              "      <td>1</td>\n",
              "      <td>38.8</td>\n",
              "      <td>87.0</td>\n",
              "      <td>14.6</td>\n",
              "      <td>41.2</td>\n",
              "      <td>5.8</td>\n",
              "      <td>12.0</td>\n",
              "      <td>14.6</td>\n",
              "      <td>9.4</td>\n",
              "      <td>32.8</td>\n",
              "      <td>...</td>\n",
              "      <td>10.2</td>\n",
              "      <td>33.8</td>\n",
              "      <td>23.8</td>\n",
              "      <td>6.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>17.0</td>\n",
              "      <td>BOS</td>\n",
              "      <td>GSW</td>\n",
              "      <td>2022-06-16</td>\n",
              "      <td>-13</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>25657 rows × 31 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-091270ce-713f-4140-a2ce-cc7ea565f8b1')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-091270ce-713f-4140-a2ce-cc7ea565f8b1 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-091270ce-713f-4140-a2ce-cc7ea565f8b1');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ],
      "source": [
        "import pandas as pd\n",
        "\n",
        "#Coded by Davin Stephens\n",
        "def spread_data():\n",
        "  # access file from github\n",
        "  nba_file = 'https://raw.githubusercontent.com/dcsteph7/Data-Capstone/main/nba_games1.csv'\n",
        "  df = pd.read_csv(nba_file, index_col = 0)\n",
        "  df = df.sort_values(\"date\")\n",
        "  df = df.reset_index(drop=True)\n",
        "\n",
        "  #Calculate point difference\n",
        "  df[\"point_difference\"] = df[\"total_opp\"] - df[\"total\"]\n",
        "\n",
        "  # remove duplicated stats and null values\n",
        "  del df[\"mp.1\"]\n",
        "  del df[\"mp_opp.1\"]\n",
        "  del df[\"index_opp\"]\n",
        "  df[\"won\"].value_counts()\n",
        "  nulls = pd.isnull(df).sum()\n",
        "  nulls = nulls[nulls > 0]\n",
        "  valid_columns = df.columns[~df.columns.isin(nulls.index)]\n",
        "  df = df[valid_columns].copy()\n",
        "\n",
        "  # compute the rolling averages of the last 5 games played for each game\n",
        "  team_stats = df.shift(periods = 1).groupby('team')[['fg', 'fga', '3p', '3pa', '3p_max', 'ft', 'fta', 'orb', 'drb', 'ast', 'stl', 'blk',\n",
        "                                  'tov', 'fg_opp', 'fga_opp', '3p_opp', '3pa_opp', '3p_max_opp', 'ft_opp', 'fta_opp', \n",
        "                                  'orb_opp', 'drb_opp', 'ast_opp', 'stl_opp', 'blk_opp', \n",
        "                                  'tov_opp']].rolling(5, min_periods=1).mean().reset_index()\n",
        "                                \n",
        "  # merge rolling values with normal using the original index\n",
        "  merged_df = df.merge(team_stats.reset_index()[['level_1', 'fg', 'fga', '3p', '3pa', '3p_max', 'ft', 'fta', 'orb', 'drb',\n",
        "                                                'ast', 'stl', 'blk', 'tov', 'fg_opp', 'fga_opp', '3p_opp', '3pa_opp',\n",
        "                                                '3p_max_opp', 'ft_opp', 'fta_opp', 'orb_opp', 'drb_opp', 'ast_opp',\n",
        "                                                'stl_opp', 'blk_opp', 'tov_opp', 'team']], left_index=True, right_on='level_1', suffixes=('', '_r'))\n",
        "  df2 = merged_df\n",
        "\n",
        "  # calculate team ids if they are necessary (currently not in use)\n",
        "  teams = list(set(df2['team'].values))\n",
        "  teamToId = {t: i for i, t in enumerate(teams)}\n",
        "  homeId = [teamToId[id] for id in list(df2['team'].values)]\n",
        "  df2['home_id'] = homeId \n",
        "  awayId = [teamToId[id] for id in list(df2['team_opp'].values)]\n",
        "  df2['away_id'] = awayId\n",
        "  df2['won'] = df2['won'].apply(lambda x: 1 if x else 0)\n",
        "\n",
        "  # set target and testing data\n",
        "  target = df2['won']\n",
        "  df2 = df2[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r', 'team', 'team_opp', 'date']]\n",
        "\n",
        "  #Calculate the point difference to compare with the spread from sportsbooks\n",
        "  df2['PointDifference'] = df['point_difference']\n",
        "  count = 0\n",
        "  for i in (df2.index):\n",
        "    df2['PointDifference'][i] = df['point_difference'][count]\n",
        "    count = count + 1\n",
        "  return df2\n",
        "\n",
        "df2 = spread_data()\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Process spread data from Kaggle dataset**\n",
        "\n",
        "In order to get odds from previous years, we had to use a kaggle dataset that had odds from the early 2000s to 2018. A lot of cleaning and data processing had to be completed, which is what the next few blocks of code consist of. (https://www.kaggle.com/datasets/ehallmar/nba-historical-stats-and-betting-data)"
      ],
      "metadata": {
        "id": "P128-kfGoWHN"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "E3HZh-8fA6Fc"
      },
      "outputs": [],
      "source": [
        "#We had to pull multiple datasets from kaggle to find spreads over the last 10 years\n",
        "#These data sets consisted of different game ids that had to be mapped to merge with other dataset\n",
        "#Coded by Tyler Wulff\n",
        "\n",
        "## this function maps the team ids with the team name\n",
        "def team_ids():\n",
        "    teams = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_teams_all.csv')\n",
        "   \n",
        "    team_dict = {}\n",
        "    ##Loop through the dataframe and convert ids to names\n",
        "    for i in range(len(teams.index)):\n",
        "        if isinstance(teams['abbreviation'][i], str):\n",
        "            team_dict[teams['team_id'][i]] = teams['abbreviation'][i]\n",
        "\n",
        "    return team_dict\n",
        "\n",
        "## This function maps the game ids to a date\n",
        "def game_ids():\n",
        "    teams = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_games_all.csv')\n",
        "    \n",
        "    team_dict = {}\n",
        "    #Loop through dataframe and convert ids to dates\n",
        "    for i in range(len(teams.index)):\n",
        "        if isinstance(teams['game_date'][i], str):\n",
        "            team_dict[teams['game_id'][i]] = teams['game_date'][i]\n",
        "\n",
        "    return team_dict\n",
        "\n",
        "#Add spreads into dataframe using previous two functions to get spread using team name and date\n",
        "def add_teams():\n",
        "    moneylines = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_money_line.csv')\n",
        "    over_under = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_totals.csv')\n",
        "    spreads = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_spread.csv')\n",
        "    games = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_games_all.csv')\n",
        "    \n",
        "    df = pd.DataFrame()\n",
        "    temp_dict = team_ids()\n",
        "    game_dict = game_ids()\n",
        "    #Loop through spreads and add spread along with date, book name, and teams\n",
        "    for i in range(len(spreads.index)):\n",
        "        if (spreads['book_name'][i] == 'Pinnacle Sports'):\n",
        "            temp = pd.DataFrame({\"date\":[game_dict[spreads['game_id'][i]]],\n",
        "                            \"book_name\":[spreads['book_name'][i]],\n",
        "                            \"team\":[temp_dict[spreads['team_id'][i]]],\n",
        "                            \"away_team\": [temp_dict[spreads['a_team_id'][i]]],\n",
        "                            \"spread1\": [spreads['spread1'][i]],\n",
        "                            \"spread2\": [spreads['spread2'][i]]})\n",
        "            #df = df.append(temp, ignore_index = True)\n",
        "            df = pd.concat([df, temp])\n",
        "    return df\n",
        "\n",
        "#Add spreads using bovada sportsbook same as previous function\n",
        "def add_teams_2():\n",
        "    moneylines = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_money_line.csv')\n",
        "    over_under = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_totals.csv')\n",
        "    spreads = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_spread.csv')\n",
        "    games = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_games_all.csv')\n",
        "    \n",
        "    df = pd.DataFrame()\n",
        "    temp_dict = team_ids()\n",
        "    game_dict = game_ids()\n",
        "    #Loop through spreads and add spread along with date, book name, and teams\n",
        "    for i in range(len(spreads.index)):\n",
        "        if (spreads['book_name'][i] == 'Bovada'):\n",
        "            temp = pd.DataFrame({\"date\":[game_dict[spreads['game_id'][i]]],\n",
        "                            \"book_name\":[spreads['book_name'][i]],\n",
        "                            \"team\":[temp_dict[spreads['team_id'][i]]],\n",
        "                            \"away_team\": [temp_dict[spreads['a_team_id'][i]]],\n",
        "                            \"spread1\": [spreads['spread1'][i]],\n",
        "                            \"spread2\": [spreads['spread2'][i]]})\n",
        "            #df = df.append(temp, ignore_index = True)\n",
        "            df = pd.concat([df, temp])\n",
        "    return df\n",
        "        \n",
        "prev_lines = add_teams()\n",
        "prev_lines = prev_lines.reset_index(drop=True)\n",
        "bovad_lines = add_teams_2()\n",
        "bovad_lines = bovad_lines.reset_index(drop=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "en1XuuZoYzIi"
      },
      "outputs": [],
      "source": [
        "import math\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "\n",
        "#Read in previous data that holds spread information calculated above\n",
        "yup = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/AllData.csv', index_col = 0)\n",
        "\n",
        "#Will add bovada lines into dataframe that will be used with existing statistical information\n",
        "## Coded by Austin Fett / Tyler Wulff\n",
        "def get_box_score_average_data_3():\n",
        "\n",
        "  # For future use, we will be pulling in all teams data\n",
        "  team_list = [\"Atl\"]\n",
        "\n",
        "  ind = 0\n",
        "  Df = []\n",
        "  Teams = []\n",
        "  for team in team_list:\n",
        "\n",
        "    # Initializing list of dicts\n",
        "    rows = [{\"Date\": ''}, {\"Spread_Result\": ''}, {\"O/U_Result\": ''}, {\"Team\": ''}]\n",
        "    # Initialize dataframes that will be used to hold features (statistical information)\n",
        "    for i in df.index:\n",
        "        d1 = {}\n",
        "        rows.append(d1)\n",
        "    \n",
        "    result_found = []\n",
        "\n",
        "    for i in yup.index:\n",
        "      result_found.append(False)\n",
        "\n",
        "    index = 0\n",
        "    count = 0\n",
        "    var = 0\n",
        "\n",
        "    # Loop through dataframe\n",
        "    for i in range(len(yup.index)):\n",
        "          #Check if a line has not been added and if the date is in the kaggle dataset\n",
        "          if (math.isnan(yup[\"Line\"][i]) and yup[\"Date\"][i] < '2018-11-06'):\n",
        "                #Loop through the bovada lines and add line into dataframe\n",
        "                # if it is the proper game\n",
        "                for j in range(len(bovad_lines.index)):\n",
        "                    if bovad_lines[\"date\"][j] == yup[\"Date\"][i]:\n",
        "                          if bovad_lines[\"team\"][j] == yup[\"Home_Team\"][i] or bovad_lines[\"team\"][j] == yup[\"Away_Team\"][i]:\n",
        "                              yup[\"Line\"][i] = float(bovad_lines[\"spread1\"][j])\n",
        "                              Teams.append(bovad_lines[\"team\"][j])\n",
        "                              var = var + 1\n",
        "                              break       \n",
        "          count = count + 1    \n",
        "    ind = ind + 1\n",
        "    \n",
        "    return yup, Teams\n",
        "\n",
        "newLines = get_box_score_average_data_3()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "kbujO4A5EmcV"
      },
      "outputs": [],
      "source": [
        "#Add dataframe into csv file with current bovada lines\n",
        "one, two = newLines\n",
        "one.to_csv('Spreads.csv')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-Ur-4zL-zTm5"
      },
      "outputs": [],
      "source": [
        "#Display game line information for presentation\n",
        "game_lines_2023_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2021_2023_Season_Odds.csv'\n",
        "game_lines_2023_data = pd.read_csv(game_lines_2023_file)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "QzvmA3uaznzu"
      },
      "outputs": [],
      "source": [
        "#Coded by Austin Fett/Tyler Wulff\n",
        "#Add lines into previous dataframe using ESPN scraped line information \n",
        "game_lines_2023_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2021_2023_Season_Odds.csv'\n",
        "game_lines_2023_data = pd.read_csv(game_lines_2023_file)\n",
        "\n",
        "#Subset dataframe to only include ESPN scraped games\n",
        "yup2 = yup.loc[yup['Date'] > '2018-11-06']\n",
        "#loop through dataframe\n",
        "for i in range(len(yup.index)):\n",
        "  #Check date to ensure it is in ESPN game range\n",
        "  if yup['Date'][i] > '2018-11-06':\n",
        "    #Loop through ESPN games\n",
        "    for j in range(len(game_lines_2023_data)):\n",
        "      #Find game based on date and teams and add line\n",
        "      if game_lines_2023_data[\"Date\"][j] == yup[\"Date\"][i]:\n",
        "        if game_lines_2023_data[\"Home\"][j] == yup[\"Home_Team\"][i]:\n",
        "          yup['Line'][i] = game_lines_2023_data['Line'][j]\n",
        "        if game_lines_2023_data[\"Away\"][j] == yup[\"Away_Team\"][i]:\n",
        "          yup['Line'][i] = game_lines_2023_data['Line'][j] * -1"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JWs5hcNGNJ9R"
      },
      "outputs": [],
      "source": [
        "#Coded by Tyler Wulff\n",
        "#Read in previous csv file that contains several spreads\n",
        "spreads = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/Finally5.csv')\n",
        "game_lines_2023_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2021_2023_Season_Odds.csv'\n",
        "game_lines_2023_data = pd.read_csv(game_lines_2023_file)\n",
        "spreads['Date'] = pd.to_datetime(spreads['Date']).dt.date\n",
        "\n",
        "#Loop through spreads dataframe\n",
        "for i in range(len(spreads.index)):\n",
        "  #Check dates to make sure we aren't looping unecessarily\n",
        "  if str(spreads['Date'][i]) > '2018-10-01':\n",
        "    #Loop through game lines data\n",
        "    for j in range(len(game_lines_2023_data.index)):\n",
        "      #Check dates and teams to ensure that lines are being added to the correct games\n",
        "      if str(spreads['Date'][i]) == str(game_lines_2023_data['Date'][j]):\n",
        "        if spreads['Team'][i] == game_lines_2023_data['Home'][j] or spreads['Team'][i] == game_lines_2023_data['Away'][j]:\n",
        "          if game_lines_2023_data['Favorite'][j] == spreads['Team'][i]:\n",
        "            spreads['Line'][i] = game_lines_2023_data['Line'][j]\n",
        "            break\n",
        "          else:\n",
        "            spreads['Line'][i] = float(game_lines_2023_data['Line'][j]) * -1\n",
        "            break"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "spreads = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/Finally5.csv')"
      ],
      "metadata": {
        "id": "QfU-kLksyCAw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2l-wbWmbTIAl"
      },
      "outputs": [],
      "source": [
        "#Coded by Tyler Wulff\n",
        "#combine spreads with box score statistical data and remove extra columns\n",
        "spreads = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/Finally5.csv')\n",
        "spreads = spreads.drop(columns = ['Unnamed: 0.5', 'Unnamed: 0.4', 'Unnamed: 0.3', 'Unnamed: 0.2', 'Unnamed: 0.1', 'Unnamed: 0'])\n",
        "spreads\n",
        "df2 = df2.sort_index()\n",
        "frames = [df2, spreads]\n",
        "df3 = pd.concat(frames, axis=1)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Coded by Tyler Wulff\n",
        "#Remove games that did not have a spread line available\n",
        "df4 = df3[df3['Line'].notnull()]\n",
        "df4 = df4.drop(columns = ['Date', 'Home_Team', 'Away_Team', 'Team'])\n",
        "\n",
        "#Create a column to indicate whether a team covered or not\n",
        "df4['Covered'] = df4['PointDifference']\n",
        "df4 = df4.reset_index()\n",
        "#Loop through games and give team a 1 if they covered the spread and a 0 if they did not\n",
        "for i in range(len(df4.index)):\n",
        "  if (df4['Line'][i] > 0):\n",
        "    if (df4['PointDifference'][i] < df4['Line'][i]):\n",
        "      df4['Covered'][i] = 1\n",
        "    else:\n",
        "      df4['Covered'][i] = 0\n",
        "  if (df4['Line'][i] < 0):\n",
        "    if (df4['PointDifference'][i] < df4['Line'][i]):\n",
        "      df4['Covered'][i] = 1\n",
        "    else:\n",
        "      df4['Covered'][i] = 0\n",
        "  if df4['Line'][i] == 0:\n",
        "    if (df4['PointDifference'][i] > 0):\n",
        "      df4['Covered'][i] = 0\n",
        "    else:\n",
        "      df4['Covered'][i] = 1\n"
      ],
      "metadata": {
        "id": "O-BvkNk6poZO",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0f54f246-d65a-45d3-fa9d-0a99a0c8a94c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-24-d6188e1dcba2>:15: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df4['Covered'][i] = 0\n",
            "<ipython-input-24-d6188e1dcba2>:13: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df4['Covered'][i] = 1\n",
            "<ipython-input-24-d6188e1dcba2>:20: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df4['Covered'][i] = 0\n",
            "<ipython-input-24-d6188e1dcba2>:18: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df4['Covered'][i] = 1\n",
            "<ipython-input-24-d6188e1dcba2>:23: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df4['Covered'][i] = 0\n",
            "<ipython-input-24-d6188e1dcba2>:25: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df4['Covered'][i] = 1\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Display for presentation\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "\n",
        "temp = df4[['PointDifference','home', 'team', 'team_opp', 'date']]\n",
        "df4 = df4.drop(columns=['PointDifference','home', 'team', 'team_opp', 'date'])\n",
        "df4 = df4.shift(periods=1)\n",
        "df4['PointDifference'] = temp['PointDifference']\n",
        "df4['home'] = temp['home']\n",
        "df4['team'] = temp['team']\n",
        "df4['team_opp'] = temp['team_opp']\n",
        "df4['date'] = temp['date']\n",
        "df4 = df4.drop([0])"
      ],
      "metadata": {
        "id": "-pNic9O83MGo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Logistic Regression for spreads (last 5 game averages)"
      ],
      "metadata": {
        "id": "HJAcQCOBctWa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "#Coded by Tyler Wulff\n",
        "# set target, testing and training data\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "target = df4[['date', 'Covered']]\n",
        "df4 = df4[['date', 'home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r', 'Line']]\n",
        "# normalize and split data into training and testing\n",
        "# Testing set will be the 2021-2022 season\n",
        "X_train = df4[df4[\"date\"] < '2021-10-10']\n",
        "X_test = df4[df4[\"date\"] > '2021-10-09']\n",
        "y_train = target[target[\"date\"] < '2021-10-10']\n",
        "y_test = target[target[\"date\"] > '2021-10-09']\n",
        "\n",
        "#Drop date column from test and train data\n",
        "y_test = y_test.drop(columns=['date'])\n",
        "X_train = X_train.drop(columns=['date'])\n",
        "X_test = X_test.drop(columns=['date'])\n",
        "y_train = y_train.drop(columns=['date'])\n",
        "xx_train = tf.keras.utils.normalize(X_train, axis=1)\n",
        "\n",
        "#Initialize and fit logistic regression\n",
        "logis = LogisticRegression(C=10.0, penalty='l1', solver='liblinear', max_iter=2000)\n",
        "logis.fit(xx_train, y_train.values.ravel())\n",
        "\n",
        "# Predict on training data\n",
        "train_pred = logis.predict(X_train)\n",
        "\n",
        "# Predict on testing data \n",
        "y_pred = logis.predict(X_test)\n",
        "\n",
        "# Print accuracy score\n",
        "print(\"Logistic Regression Train Accuracy: \" + str(accuracy_score(y_train, train_pred)))\n",
        "print(\"Logistic Regression Test Accuracy: \" + str(accuracy_score(y_test, y_pred)))\n"
      ],
      "metadata": {
        "id": "zZvfCjoB4cIv",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "71b679bb-07e9-4f38-cf1f-481432c0506c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Logistic Regression Train Accuracy: 0.695466313444518\n",
            "Logistic Regression Test Accuracy: 0.6163757273482959\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n"
      ],
      "metadata": {
        "id": "ddy0nieVwxRo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Neural Network (Spreads)"
      ],
      "metadata": {
        "id": "EUVjXYpmpcgY"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Neural Network for predicting spreads**"
      ],
      "metadata": {
        "id": "F-9BjflPn60o"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "#Coded by Davin Stephens\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "#Covered is the target label to predict\n",
        "target = df4['Covered']\n",
        "#Features that will be predicted\n",
        "df4 = df4[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "sentiment = 0\n",
        "\n",
        "# normalize and split data into training and testing\n",
        "xx_train = tf.keras.utils.normalize(df4, axis=1)\n",
        "X_train, X_test, y_train, y_test = train_test_split(xx_train, target, test_size=.2)\n",
        "\n",
        "# neural network model, not fully optimized yet \n",
        "model = tf.keras.models.Sequential()\n",
        "model.add(tf.keras.layers.Dense(64, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(32, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(1, activation=tf.nn.sigmoid))\n",
        "model.compile(optimizer='adam', loss='binary_crossentropy', metrics='accuracy')\n",
        "model.fit(X_train, y_train, epochs=50, batch_size=32)"
      ],
      "metadata": {
        "id": "emAVHKhF82FB",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5a5ec017-132c-4e5e-a163-8ab63f1b7ff5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "476/476 [==============================] - 2s 2ms/step - loss: 0.6894 - accuracy: 0.5369\n",
            "Epoch 2/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6843 - accuracy: 0.5498\n",
            "Epoch 3/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6680 - accuracy: 0.5933\n",
            "Epoch 4/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6477 - accuracy: 0.6175\n",
            "Epoch 5/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6372 - accuracy: 0.6327\n",
            "Epoch 6/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6344 - accuracy: 0.6325\n",
            "Epoch 7/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6320 - accuracy: 0.6414\n",
            "Epoch 8/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6326 - accuracy: 0.6357\n",
            "Epoch 9/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6311 - accuracy: 0.6377\n",
            "Epoch 10/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6303 - accuracy: 0.6390\n",
            "Epoch 11/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6280 - accuracy: 0.6425\n",
            "Epoch 12/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6282 - accuracy: 0.6436\n",
            "Epoch 13/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6289 - accuracy: 0.6415\n",
            "Epoch 14/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6289 - accuracy: 0.6396\n",
            "Epoch 15/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6284 - accuracy: 0.6434\n",
            "Epoch 16/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6271 - accuracy: 0.6419\n",
            "Epoch 17/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6288 - accuracy: 0.6398\n",
            "Epoch 18/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6271 - accuracy: 0.6412\n",
            "Epoch 19/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6269 - accuracy: 0.6392\n",
            "Epoch 20/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6260 - accuracy: 0.6444\n",
            "Epoch 21/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6262 - accuracy: 0.6444\n",
            "Epoch 22/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6283 - accuracy: 0.6402\n",
            "Epoch 23/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6256 - accuracy: 0.6427\n",
            "Epoch 24/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6257 - accuracy: 0.6427\n",
            "Epoch 25/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6271 - accuracy: 0.6401\n",
            "Epoch 26/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6250 - accuracy: 0.6451\n",
            "Epoch 27/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6259 - accuracy: 0.6408\n",
            "Epoch 28/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6245 - accuracy: 0.6474\n",
            "Epoch 29/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6262 - accuracy: 0.6419\n",
            "Epoch 30/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6258 - accuracy: 0.6430\n",
            "Epoch 31/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6250 - accuracy: 0.6411\n",
            "Epoch 32/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6231 - accuracy: 0.6471\n",
            "Epoch 33/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6241 - accuracy: 0.6440\n",
            "Epoch 34/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6255 - accuracy: 0.6453\n",
            "Epoch 35/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6234 - accuracy: 0.6421\n",
            "Epoch 36/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6238 - accuracy: 0.6430\n",
            "Epoch 37/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6219 - accuracy: 0.6488\n",
            "Epoch 38/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6237 - accuracy: 0.6436\n",
            "Epoch 39/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6220 - accuracy: 0.6462\n",
            "Epoch 40/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6209 - accuracy: 0.6473\n",
            "Epoch 41/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6216 - accuracy: 0.6423\n",
            "Epoch 42/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6217 - accuracy: 0.6413\n",
            "Epoch 43/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6209 - accuracy: 0.6451\n",
            "Epoch 44/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6210 - accuracy: 0.6465\n",
            "Epoch 45/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6218 - accuracy: 0.6455\n",
            "Epoch 46/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6210 - accuracy: 0.6474\n",
            "Epoch 47/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6201 - accuracy: 0.6448\n",
            "Epoch 48/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6204 - accuracy: 0.6456\n",
            "Epoch 49/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6200 - accuracy: 0.6471\n",
            "Epoch 50/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6189 - accuracy: 0.6467\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f8446d85750>"
            ]
          },
          "metadata": {},
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Prediction on the test set from the neural network on the spreads\n",
        "y_pred = model.predict(X_test)\n",
        "y_pred = [0 if val < 0.5 else 1 for val in y_pred]\n",
        "nn_test_accuracy = accuracy_score(y_test, y_pred)\n",
        "print(nn_test_accuracy)"
      ],
      "metadata": {
        "id": "2skqVwbv9XF2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1905487b-c55c-4757-da87-8bf9d7524856"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "119/119 [==============================] - 0s 1ms/step\n",
            "0.6352879305811202\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')"
      ],
      "metadata": {
        "id": "6plpo9jc0kfJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Linear Regression on predicting spreads"
      ],
      "metadata": {
        "id": "KX6Udh2ISDT4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.metrics import mean_squared_error\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "df5 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2023Season.csv')\n",
        "# Coded by Tyler Wulff\n",
        "def run_linear_regression():\n",
        "  # Initialize train and test sets\n",
        "  # x_train, y_train, x_test, y_test = regression_train_test_data()\n",
        "\n",
        "  y_train = df4[['Line']]\n",
        "  \n",
        "  x_train = df4[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "  # normalize and split data into training and testing\n",
        "  # Testing set will be the 2021-2022 season\n",
        "  y_test = df5[['Line']]\n",
        "  x_test = df5[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "\n",
        "  #initialize and fit linear regression on training data\n",
        "  lin_reg = LinearRegression()\n",
        "  lin_reg.fit(x_train, y_train)\n",
        "\n",
        "  #Predict with training set\n",
        "  trainPrediction = lin_reg.predict(x_train)\n",
        "  train_lin_mse = mean_squared_error(y_train, trainPrediction)\n",
        "  train_lin_rmse = np.sqrt(train_lin_mse)\n",
        "  print(\"Linear Regression Train Data RMSE: \" + str(train_lin_rmse))\n",
        "\n",
        "  #Predict with test data\n",
        "  prediction = lin_reg.predict(x_test)\n",
        "  # Find Mean squared error for prediction\n",
        "  lin_mse = mean_squared_error(y_test, prediction)\n",
        "  # Find RMSE\n",
        "  lin_rmse = np.sqrt(lin_mse)\n",
        "  print(\"Linear Regression Test Data RMSE: \" + str(lin_rmse))\n",
        "  return prediction, y_test, x_test\n",
        "\n",
        "val, test, xTest = run_linear_regression()\n"
      ],
      "metadata": {
        "id": "4FFc38dddlbt",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0f2d7ed8-0ebd-4980-f0d5-5fee1896c688"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Linear Regression Train Data RMSE: 6.669061946456832\n",
            "Linear Regression Test Data RMSE: 6.217489537424567\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Random Forest to predict spreads"
      ],
      "metadata": {
        "id": "pAGYXBMRSJ2z"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import RandomForestRegressor\n",
        "\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "df5 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2023Season.csv')\n",
        "def run_forest_regressor():\n",
        "  #Initialize train and test data\n",
        "  y_train = df4[['PointDifference']]\n",
        "  x_train = df4[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r', 'Line']]\n",
        "  # normalize and split data into training and testing\n",
        "  # Testing set will be the 2021-2022 season\n",
        "  y_test = df5[['PointDifference']]\n",
        "  x_test = df5[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r', 'Line']]\n",
        "\n",
        "  #Initialize random forest regressor and fit on training data\n",
        "  forestReg = RandomForestRegressor()\n",
        "  forestReg.fit(x_train, y_train.values.ravel())\n",
        "\n",
        "  #Predict using training data\n",
        "  predict = forestReg.predict(x_train)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  tree_mse = mean_squared_error(y_train, predict)\n",
        "  tree_rmse = np.sqrt(tree_mse)\n",
        "\n",
        "  #Predict using testing data\n",
        "  test_predict = forestReg.predict(x_test)\n",
        "\n",
        "  #Find MSE and RMSE\n",
        "  test_tree_mse = mean_squared_error(y_test, test_predict)\n",
        "  test_tree_rmse = np.sqrt(test_tree_mse)\n",
        "  print(\"Random Forest Train Data RMSE: \" + str(tree_rmse))\n",
        "  print(\"Random Forest Test Data RMSE: \" + str(test_tree_rmse))\n",
        "  return test_predict, y_test\n",
        "\n",
        "val, test = run_forest_regressor()\n",
        "win = 0\n",
        "loss = 0\n",
        "total = 0\n",
        "for i in range(len(val)):\n",
        "  if (abs(val[i] - test['PointDifference'][i]) > 3):\n",
        "    if (val[i] < test['PointDifference'][i]):\n",
        "      win = win + 1\n",
        "    else:\n",
        "      loss = loss + 1\n",
        "    total = total + 1\n",
        "print(win)\n",
        "print(loss)\n",
        "print(total)\n",
        "print(win/total)"
      ],
      "metadata": {
        "id": "mLi9jVt5ivx0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6d55eef7-c236-4de3-c273-ef5477e5e8b5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Random Forest Train Data RMSE: 4.463324551033339\n",
            "Random Forest Test Data RMSE: 11.86160597977968\n",
            "955\n",
            "963\n",
            "1918\n",
            "0.49791449426485923\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Neural Network to predict spreads (regression)"
      ],
      "metadata": {
        "id": "rIrhT-IeSN_a"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "#Coded by Davin Stephens\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "#Covered is the target label to predict\n",
        "target = df4['PointDifference']\n",
        "#Features that will be predicted\n",
        "df4 = df4[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "\n",
        "# normalize and split data into training and testing\n",
        "xx_train = tf.keras.utils.normalize(df4, axis=1)\n",
        "X_train, X_test, y_train, y_test = train_test_split(xx_train, target, test_size=.2)\n",
        "\n",
        "# neural network model, not fully optimized yet \n",
        "model = tf.keras.models.Sequential()\n",
        "model.add(tf.keras.layers.Dense(64, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(32, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(1, activation=tf.nn.sigmoid))\n",
        "model.compile(optimizer='adam', loss='mae', metrics='mae')\n",
        "model.fit(X_train, y_train, epochs=50, batch_size=32)"
      ],
      "metadata": {
        "id": "_VQ_sm80hTmc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "df255b1e-68d6-4036-f67a-ce04ef4cd449"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "476/476 [==============================] - 4s 4ms/step - loss: 11.3753 - mae: 11.3753\n",
            "Epoch 2/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.3486 - mae: 11.3486\n",
            "Epoch 3/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2769 - mae: 11.2769\n",
            "Epoch 4/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2358 - mae: 11.2358\n",
            "Epoch 5/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2206 - mae: 11.2206\n",
            "Epoch 6/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2124 - mae: 11.2124\n",
            "Epoch 7/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2093 - mae: 11.2093\n",
            "Epoch 8/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2048 - mae: 11.2048\n",
            "Epoch 9/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2019 - mae: 11.2019\n",
            "Epoch 10/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.2050 - mae: 11.2050\n",
            "Epoch 11/50\n",
            "476/476 [==============================] - 2s 5ms/step - loss: 11.1982 - mae: 11.1982\n",
            "Epoch 12/50\n",
            "476/476 [==============================] - 2s 3ms/step - loss: 11.1988 - mae: 11.1988\n",
            "Epoch 13/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.2027 - mae: 11.2027\n",
            "Epoch 14/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.1983 - mae: 11.1983\n",
            "Epoch 15/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.1988 - mae: 11.1988\n",
            "Epoch 16/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1987 - mae: 11.1987\n",
            "Epoch 17/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1984 - mae: 11.1984\n",
            "Epoch 18/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1970 - mae: 11.1970\n",
            "Epoch 19/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1955 - mae: 11.1955\n",
            "Epoch 20/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1963 - mae: 11.1963\n",
            "Epoch 21/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1958 - mae: 11.1958\n",
            "Epoch 22/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1968 - mae: 11.1968\n",
            "Epoch 23/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.1939 - mae: 11.1939\n",
            "Epoch 24/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.1920 - mae: 11.1920\n",
            "Epoch 25/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.1957 - mae: 11.1957\n",
            "Epoch 26/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1902 - mae: 11.1902\n",
            "Epoch 27/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1928 - mae: 11.1928\n",
            "Epoch 28/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1942 - mae: 11.1942\n",
            "Epoch 29/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1958 - mae: 11.1958\n",
            "Epoch 30/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1934 - mae: 11.1934\n",
            "Epoch 31/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1924 - mae: 11.1924\n",
            "Epoch 32/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1927 - mae: 11.1927\n",
            "Epoch 33/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1916 - mae: 11.1916\n",
            "Epoch 34/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1980 - mae: 11.1980\n",
            "Epoch 35/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1920 - mae: 11.1920\n",
            "Epoch 36/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1930 - mae: 11.1930\n",
            "Epoch 37/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 11.1952 - mae: 11.1952\n",
            "Epoch 38/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1875 - mae: 11.1875\n",
            "Epoch 39/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1928 - mae: 11.1928\n",
            "Epoch 40/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1948 - mae: 11.1948\n",
            "Epoch 41/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1957 - mae: 11.1957\n",
            "Epoch 42/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1944 - mae: 11.1944\n",
            "Epoch 43/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1983 - mae: 11.1983\n",
            "Epoch 44/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1924 - mae: 11.1924\n",
            "Epoch 45/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1939 - mae: 11.1939\n",
            "Epoch 46/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1901 - mae: 11.1901\n",
            "Epoch 47/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1946 - mae: 11.1946\n",
            "Epoch 48/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1911 - mae: 11.1911\n",
            "Epoch 49/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1900 - mae: 11.1900\n",
            "Epoch 50/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 11.1967 - mae: 11.1967\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f845f0bf460>"
            ]
          },
          "metadata": {},
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "2023 Season Results\n",
        "\n",
        "Below are the model results from the 2023 season, training on all of the previous data that has been cleaned and processed. \n",
        "\n",
        "Neural Network for 2023 season results"
      ],
      "metadata": {
        "id": "xfkJuhxhSS4y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "#Coded by Davin Stephens/Tyler Wulff\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "df5 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2023Season.csv')\n",
        "\n",
        "y_train = df4[['Covered']]\n",
        "X_train = df4[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "y_test = df5[['Covered']]\n",
        "X_test = df5[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "\n",
        "\n",
        "# neural network model, not fully optimized yet \n",
        "model = tf.keras.models.Sequential()\n",
        "model.add(tf.keras.layers.Dense(64, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(32, activation=tf.nn.relu6))\n",
        "model.add(tf.keras.layers.Dense(1, activation=tf.nn.sigmoid))\n",
        "model.compile(optimizer='adam', loss='binary_crossentropy', metrics='accuracy')\n",
        "model.fit(X_train, y_train, epochs=50, batch_size=32)\n",
        "\n",
        "#Prediction on the test set from the neural network on the spreads\n",
        "y_pred = model.predict(X_test)\n",
        "y_pred = [0 if val < 0.5 else 1 for val in y_pred]\n",
        "nn_test_accuracy = accuracy_score(y_test, y_pred)\n",
        "print(nn_test_accuracy)"
      ],
      "metadata": {
        "id": "0VYdlKECh67K",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4340a9cd-567f-4cce-a4b4-dd5b92fdbbae"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "595/595 [==============================] - 7s 7ms/step - loss: 0.6698 - accuracy: 0.5872\n",
            "Epoch 2/50\n",
            "595/595 [==============================] - 4s 6ms/step - loss: 0.6542 - accuracy: 0.6155\n",
            "Epoch 3/50\n",
            "595/595 [==============================] - 2s 4ms/step - loss: 0.6474 - accuracy: 0.6202\n",
            "Epoch 4/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6443 - accuracy: 0.6261\n",
            "Epoch 5/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6396 - accuracy: 0.6319\n",
            "Epoch 6/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6378 - accuracy: 0.6335\n",
            "Epoch 7/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6372 - accuracy: 0.6317\n",
            "Epoch 8/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6360 - accuracy: 0.6331\n",
            "Epoch 9/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6372 - accuracy: 0.6322\n",
            "Epoch 10/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6344 - accuracy: 0.6367\n",
            "Epoch 11/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6326 - accuracy: 0.6376\n",
            "Epoch 12/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6313 - accuracy: 0.6400\n",
            "Epoch 13/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6348 - accuracy: 0.6344\n",
            "Epoch 14/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6300 - accuracy: 0.6385\n",
            "Epoch 15/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6310 - accuracy: 0.6407\n",
            "Epoch 16/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6302 - accuracy: 0.6395\n",
            "Epoch 17/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6311 - accuracy: 0.6372\n",
            "Epoch 18/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6304 - accuracy: 0.6409\n",
            "Epoch 19/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6303 - accuracy: 0.6384\n",
            "Epoch 20/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6284 - accuracy: 0.6436\n",
            "Epoch 21/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6298 - accuracy: 0.6414\n",
            "Epoch 22/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6306 - accuracy: 0.6364\n",
            "Epoch 23/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6295 - accuracy: 0.6393\n",
            "Epoch 24/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6284 - accuracy: 0.6429\n",
            "Epoch 25/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6297 - accuracy: 0.6381\n",
            "Epoch 26/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6262 - accuracy: 0.6419\n",
            "Epoch 27/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6266 - accuracy: 0.6444\n",
            "Epoch 28/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6277 - accuracy: 0.6413\n",
            "Epoch 29/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6284 - accuracy: 0.6388\n",
            "Epoch 30/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6262 - accuracy: 0.6427\n",
            "Epoch 31/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6264 - accuracy: 0.6381\n",
            "Epoch 32/50\n",
            "595/595 [==============================] - 3s 5ms/step - loss: 0.6240 - accuracy: 0.6447\n",
            "Epoch 33/50\n",
            "595/595 [==============================] - 3s 5ms/step - loss: 0.6263 - accuracy: 0.6393\n",
            "Epoch 34/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6228 - accuracy: 0.6450\n",
            "Epoch 35/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6256 - accuracy: 0.6388\n",
            "Epoch 36/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6238 - accuracy: 0.6416\n",
            "Epoch 37/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6222 - accuracy: 0.6407\n",
            "Epoch 38/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6248 - accuracy: 0.6380\n",
            "Epoch 39/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6238 - accuracy: 0.6391\n",
            "Epoch 40/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6239 - accuracy: 0.6425\n",
            "Epoch 41/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6223 - accuracy: 0.6401\n",
            "Epoch 42/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6228 - accuracy: 0.6415\n",
            "Epoch 43/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6218 - accuracy: 0.6414\n",
            "Epoch 44/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6225 - accuracy: 0.6414\n",
            "Epoch 45/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6213 - accuracy: 0.6451\n",
            "Epoch 46/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6218 - accuracy: 0.6427\n",
            "Epoch 47/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6262 - accuracy: 0.6384\n",
            "Epoch 48/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6238 - accuracy: 0.6394\n",
            "Epoch 49/50\n",
            "595/595 [==============================] - 1s 2ms/step - loss: 0.6210 - accuracy: 0.6434\n",
            "Epoch 50/50\n",
            "595/595 [==============================] - 2s 3ms/step - loss: 0.6213 - accuracy: 0.6438\n",
            "75/75 [==============================] - 0s 2ms/step\n",
            "0.6128896377422073\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Coded by Tyler Wulff\n",
        "\n",
        "#This loops through our neural network model predictions \n",
        "#and outputs what the profit for the 2023 season would have been if \n",
        "#a bettor placed a 10 dollar wager on every game, assuming all odds were -110\n",
        "count = 0\n",
        "win = 0\n",
        "profit = 0\n",
        "profitArr = []\n",
        "\n",
        "for i in y_test.index:\n",
        "  if y_test['Covered'][i] == y_pred[count]:\n",
        "    win = win + 1\n",
        "    profit = profit + 9.09\n",
        "    \n",
        "  else:\n",
        "    profit = profit - 10\n",
        "\n",
        "  profitArr.append(profit)\n",
        "  count = count + 1\n",
        "\n",
        "print(win)\n",
        "print(count)\n",
        "\n",
        "print(win / count)\n",
        "print(profit)"
      ],
      "metadata": {
        "id": "h_Q6-OvC4J-c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "40d14f3c-3219-41c4-b87a-e2a5dc7b7a28"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1362\n",
            "2374\n",
            "0.573715248525695\n",
            "2260.579999999978\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.dates as mdates\n",
        "\n",
        "#Coded by Tyler Wulff\n",
        "#This is a plot of our profit over the 2023 season\n",
        "# https://www.python-graph-gallery.com/basic-time-series-with-matplotlib\n",
        "fig, ax = plt.subplots(figsize=(8, 6))\n",
        "half_year_locator = mdates.MonthLocator(interval=1)\n",
        "ax.xaxis.set_major_locator(half_year_locator)\n",
        "\n",
        "fig.suptitle('2023 Season Profit From Neural Network (Unit = $10)', fontsize=16)\n",
        "plt.xlabel('Date During 2023 Season', fontsize=14)\n",
        "plt.ylabel('Profit in Dollars', fontsize=14)\n",
        "ax.plot(df5['date'], profitArr);\n",
        "fig.savefig('NNProfit.jpg')\n"
      ],
      "metadata": {
        "id": "1nI2QolS5t20",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 607
        },
        "outputId": "30b79e20-0d04-4d68-c681-2260c910e33b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsUAAAJOCAYAAABfvHoDAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACufUlEQVR4nOzdd1hT1/8H8HcgQNhDGYIguPdeqDhwr1qlw1Hrav3Zqh22zjrbftVaa1tbbWuH1qqtrVWr4kYR9957gKAIosje5Pz+wFyIBAghECDv1/PwPHece+8nJIRPTs79HJkQQoCIiIiIyIiZGDoAIiIiIiJDY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1KsB5mZmQgKCsLUqVPRpk0bODg4wMzMDG5ubnjppZcQGBhY5Dn279+Pfv36oWrVqrC0tET9+vXxySefICkpSWP7mzdv4ptvvkG/fv3g4eEBc3Nz2NnZoU2bNli0aFGBxz148ADTpk1Dz5494e3tDVtbW1hYWMDLywtDhw7FkSNHdP49HDx4EEOHDkWNGjWgUChga2sLHx8fdOvWDZ988glOnDih87mNWXBwMGQyWb4fW1tbNGvWDDNmzMDjx4/LNKbt27fDz88PdnZ2UjzBwcEAIK3rm7e3t8bfQ96fb775Ru/XNbTRo0dLj2/SpEkFtvv8888hk8kwevTosguujM2fPx8ymQzz58/X+piwsDDp92dpaYkHDx4U2FYul0MmkyEsLKzkwVYSqvefrl276v3c3377LWQyGf7991+17V27dtXqeVa9HvQVm+q14u3trZfzlQfavp7PnTuHpUuXYtiwYahbty5MTEwgk8mwbt06ra7zzz//oGvXrnB0dIS1tTWaNWuGJUuWIDMzU2P7Hj16wM7ODlFRUcV5OKVObugAKoNDhw6hZ8+eAAA3Nzd06tQJ1tbWuHbtGrZv347t27dj/Pjx+PHHHzUmC19//TWmTJkCmUwGPz8/uLq64vDhw1i4cCH+/fdfHDlyBFWrVlU7pnv37nj48CEUCgVat26Nzp07Izo6GsePH8eZM2fw66+/4sCBA/Dy8lI77saNG/jyyy/h6OiIhg0bok2bNsjKysL169exceNGbNy4EV988QWmTZtWrN/BtGnT8OWXXwIAatasiZ49e8LW1haPHj3CuXPnEBwcjJs3b2LTpk3FOi+pGzVqFABACIH79+/jxIkTuHTpEtasWYPg4GDUr1+/1GO4cOECAgICoFQq4e/vj2rVqkEmk8HNza3Q47p27YpDhw7h4MGDJfon1rFjR9SuXVvjvoYNG+p83opg1apV+PDDD1GrVi1Dh1IhpaWlYe7cufjtt9/K5Hqq93shRJlcryKJiYnB/Pnz0aZNGwQEBBg6nCJ5e3vj/v37CA0NrVRJs8qnn36K//77T6djP/jgA3z77beQy+Xw9/eHjY0NDhw4gOnTp2P79u3Yu3cvLC0t1Y5ZvHgx2rRpg5kzZ2L16tX6eAj6IajEgoKCREBAgAgJCcm376+//hKmpqYCgPj999/z7T937pyQyWTC1NRU7Ny5U9qenJwsunfvLgCIgICAfMf5+/uLX3/9VSQmJqptDw0NFY0aNRIARLdu3fIdFxUVJc6fPy+ys7Pz7duwYYMwNTUVJiYm4tq1a1o9diGE2LFjhwAg5HK5+PPPP/Ptz8jIEIGBgeL777/X+pyU6+DBgwKA0PTnevPmTVG9enUBQPj5+ZVJPPPmzRMAxKxZszTuv379urh+/Xq+7V26dBEAxMGDB3W6bo0aNQQAsXr1ap2Or6hGjRolAAgrKysBQLz++usa23322WcCgBg1alTZBliGVK+9efPmaX1MaGioACBkMplQKBTC1NRUXLlyRWNb1Xt1aGioXuIt6O+2IlG9/3Tp0kWv5504caIAIAIDA/PtU71XFPU8q14P+ootIyNDXL9+Xdy5cyffPtX7j75eG6UlLS1NfPbZZ6JRo0ZCoVAIAEKhUAgvLy/h7+8v5syZI5RKZb7jFi1aJGbNmiU2bdok7t69Kz0Hf/zxR6HX27JliwAgbGxsxNmzZ6XtMTExokmTJgKA+OijjzQeO2DAACGTycSFCxdK9qD1qGL/tVYQ48aNEwBE9+7d8+179dVXBQDx1ltv5dsXFhYmTExMBACNSUZBDh8+LL0ZR0REFCtWf39/AUAsX75c62PeeOMNAUAMGzasWNci7RSWFAshxOrVq6X9kZGRpR7PmDFjdEpOmRTrRpUUT548WdjY2AiZTKb2z0eFSbFmqqTY1NRUTJ06VQAQL730ksa2TIrzK42k+NmzZ8La2lp4eHho7KAxVFJcmIqQFCuVSqkzTaFQiH79+gkAYuDAgaJz587C0tJSABCZmZlFnkvbpLhNmzYCgPj888/z7VPlIhYWFiIuLi7fflVCPWbMGO0fZCnjmOIy0KJFCwBARESE2vaMjAxpvPHw4cPzHVejRg107NgRALBly5ZiX0/TNYsil+eMqLGwsND6mOjoaACAi4tLsa6l8uzZM8ybNw/NmzeHra0trKys0KRJE3z++edISUnJ1z4xMRE///wzhgwZgjp16sDa2hrW1tZo0qQJPvnkE8TFxWm8zqNHj/D++++jbt26UCgUsLKygqenJ7p3746lS5dqPObUqVN47bXX4O7uDnNzc7i4uGDgwIHYt2+fxvaq8Z9r1qxBaGgoRo4cCTc3N1hYWKBWrVqYPXs20tPTdfo9FaRVq1bS8v379wEAa9askcaXxsbG4oMPPkCtWrVgYWGhNnQhKysLP/74Izp06AB7e3soFArUqVMH7733Hh4+fKh2HdX4PdVXXWPGjJHGauY954tjilVjEg8dOgQA6Natm9o44DVr1uj19wHkjj8OCwvDf//9B39/fzg5OamNfQZyhhONGTMGNWrUgIWFBZycnNC9e3f8/fffGs+bd0xrZGQk3nrrLbi7u8PS0hKNGzfGr7/+qnbu4cOHw83NDQqFAs2aNcPGjRt1fkwuLi746KOPIITA9OnTi318ZGQkpkyZggYNGsDKygq2trZo06YNvv/+e2RlZeVrrxrXmff3lVdB43vzbg8PD8e4cePg6ekJMzMztfHOmzdvxltvvYXGjRvD0dERCoUCPj4+GDt2LG7evFnsx6eNmTNnwtHREdu2bcPRo0eLffymTZvQp08fODs7w9zcHB4eHnjjjTdw7do1tXaq34HKi2Pfw8LCsG3bNshkMrz00kv5rvPuu+9CJpPBzMwMCQkJavtCQkIgk8nQuXPnfMeV5PVc2HNVkJiYGHTo0AEymQyvv/661u9tq1evRnJyMkaOHAkTE/2nIXnf/5KTkzFz5kzUrl0bFhYWcHNzw6hRo/K9vwGaxxSrzqV6b/Xx8VF7Lgv6+zCEPXv2ICgoCG5ubrh27ZqUXyxfvhyHDh1CVFQUfvnlF739zh8+fIjTp08D0JzDdOrUCZ6enkhPT8fOnTvz7e/fvz+qVq2KP//8E7GxsXqJqaQ4prgM3L59GwBQrVo1te23bt2Skr7WrVtrPLZ169Y4fPgwzp8/X+zrabpmYQIDA3Hw4EEoFAr06tVL6+NU45Y3bdqEqVOnwsPDQ+tjr127hj59+iAiIgLVqlVDp06dYGZmhlOnTmHOnDn4999/ERwcDHt7e+mYixcvYvz48XB2dka9evXQqlUrPHv2DGfPnsXChQvx999/48SJE6hSpYp0TFRUFFq3bo3IyEh4eXmhT58+UCgUiIyMxIULF3D27Fl8/PHHarH9/PPPmDBhApRKJVq0aIGuXbvi/v372LFjB3bs2IH58+dj3rx5Gh/XhQsX8P7778PR0RFdunRBbGwsjh49iv/973+4evVqsT7kFCXvP80XP8w8efIErVu3RlxcHPz8/NCqVSuYm5sDANLT0zFgwADs378fCoUC3bp1g52dHY4dO4bvvvsOf/75J/bs2YOWLVsCAJo3b45Ro0bhyJEjuHv3rtrY3sLGMqv+Ce3evRvR0dHo3bu32vjjgsYH68NXX32F77//Hq1bt0afPn0QGRkJU1NTADmv91deeQVpaWmoV68ehgwZgsePH+PQoUM4cOAA9uzZo5bk5hUeHi79Lv38/BATE4OQkBC89dZbiIuLQ8eOHdGrVy+4u7ujW7duuH//Po4fP46hQ4cCAF5//XWdHs/HH3+MH374Afv378e+ffukexmKEhISgpdffhnPnj2Dt7c3evbsifT0dJw6dQqTJ0/G9u3bsWPHDpiZmekUlya3b99GixYtYG5ujo4dO0IIoXZvxGuvvQYLCws0bNgQ/v7+yMrKwpUrV7B69Wr8/fff2Lt3Lzp06KC3eADA0dERM2bMwPTp0zF9+nStbyzOysrCiBEj8Pfff8PCwgKtWrWCh4cHbt26hfXr12Pz5s3YvHkz+vTpAyD3b+X3338HkHsvgIqNjQ26du0KuVyOQ4cOISsrS+qQAHJuvFZdNzg4WC1xVu3r0aOH2jlL8nou6rnS5NatW+jXrx/u3r2LadOmYfHixVrfYLt161aNj0Hf4uPj0aFDB4SHh8PPzw+NGzfG8ePHsXbtWhw6dAgXL15U+9+iSe3atTFq1Chs2rQJycnJCAgIgI2NjbS/qHspytKFCxcAAEOHDoWPj0++/XZ2dhg3bpzerqfKS5ycnDReD8jJYSIiInD+/HkMGzZMbZ+ZmRm6du2KTZs2Yc+ePfn2G4SBe6orvUePHgl7e3uNQxK2bdsmAAgHB4cCj1+2bJkAIFq3bq31NV9//XUBQLRs2bLQdu+8844YNWqUCAgIkMb+2Nraik2bNml9LSGEOHXqlJDL5QKAsLS0FK+88or45ptvREhIiEhOTi7wuJSUFFGrVi0BQMyePVukp6dL+5KTk8WwYcM0frUSEREh9u/fn+9rt+TkZPHmm28KAOLdd99V27dgwQIBQIwfPz7feKqMjAyxf/9+tW2XLl0ScrlcyGQysXbtWrV9O3fuFObm5gKA2Lt3r9o+1VfdAMQnn3wisrKypH2XL18W1tbWAoA4duxYgb+XFxU1fOLjjz+Wvi5LSUkRQqgPqejevbuIj4/Pd9z06dMFAFGrVi21rwQzMjKkIT8+Pj5qz0vex1jQMIaCYi3L4ROqtqampuK///7Ltz8qKkr6u/z888/VXhOnT58Wjo6OAoBYtWqV2nGqr2sBiAkTJqh9Dan6e7a1tRU1atTId95vvvlGABC1a9cu1uNW/b4/++wzIYQQy5cvl/6+856/oOETjx49ElWqVBEymUysXLlS7e/myZMn0pCpBQsWqB1X1PNV0FCGvL+jN954Q6SlpWk8/q+//hJJSUlq25RKpVixYoUAIBo1apTvb7WkwyeEyHnfUY3Df/G1UdDwiVmzZgkAol27duLevXtq+/755x9hamoqHB0dxbNnz9T2FfZ3K4QQvr6+AoA4evSotO3+/fsCgGjatKk0bKaoY/Txei7oudI0fCIkJEQ4OTkJU1NT8eOPPxb4+DRJSUkR5ubmwsTERCQkJGhsU9LhE3nf/3r37q32/hcbGyuaN28uAIiFCxeqHad6rdSoUSPftUo6fEL1mIr7U5zrfffddwKAGDx4sLRN15i1GT6hei9q3rx5gW3ee+89AUC88sorGvercpxx48YVO8bSwKS4FGVmZkrje5o0aZIvuVi/fr0AIDw8PAo8x6pVqwQAUbduXa2uqXozMDU1FYcPHy60repNVPXj7Ows/v33X62u86Lt27dL/2jy/piZmYmePXvmSx6FEOKHH34QAMSAAQM0njMxMVG4uLgIuVwuYmNjtYojOTlZyOVy4ezsrLb93XffFQDE5s2btTqPKikcMmSIxv2TJk0SAETPnj3VtqsSmFatWmm8mWHChAkCgPj000+1ikMIzUmxUqkU9+/fF59//rn0geS9996T9qteB2ZmZuLu3bv5zpmamipsbGwEALFt27Z8+5OTk4Wrq6sAINavX6/xMRoqKS7oJ+8/RlXbsWPHajyXKoFs1aqVxv1Lly4VAESdOnXUtqv+CXt5eYnU1NR8x6kSmbZt2+Z7/jMzM4WTk5MAIO7fv6/1434xKc7IyBA1a9YUANRubC0oKVZ9+Jk0aZLG8z948ECYmZkJZ2dntZhLmhQ7OTlpHEeoDVXid/XqVa2uWZgXk2IhhPjll1+kxDvvB1dNSfHTp0+FpaWlUCgU4sGDBxqvoXp/+e6779S2F5UUz5kzRwAQ8+fPl7b9+uuvAoD47bffhIuLi6hfv760Lz4+XsjlcmFnZ6f2gaykr+fCnqsXk+INGzYICwsLYWNjo3ZzuLZOnz4t/Q0VRF9JsbW1tcb7LP766y8BQPj7+6ttL82keNGiRWLUqFHF/omJidH6GqGhoVKHzYQJE8Tx48dLNSn+3//+JwCIjh07FthG9YGyV69eGvfv2bNHABAtWrQodoylgcMnStGECRMQFBSEKlWqYNOmTdLX1qUlKCgI//d//wcAWLJkCTp16lRoe9XY29jYWFy5cgULFy5EQEAAhg4dinXr1klfM2tjwIAB6N27N/bs2YP9+/fj9OnTuHDhAlJSUrBv3z7s27cPc+fOxYIFC6RjVOOdCvoq2cbGBq1bt8bOnTtx+vTpfEM6jh07hsOHDyM8PBwpKSlS2SNzc3PExMTg2bNncHR0BAC0bdsWK1euxIwZMyCEQK9evdS+AnuRapxYQePqxo0bh++//x6HDx9GdnZ2vt/VgAEDNH6V2KBBAwDQOJ5NGwV9PTl8+HAsWbIk3/YWLVqgZs2a+bafOXMGSUlJcHJywsCBA/Ptt7KywtChQ/Htt9/i4MGDGseLGUpBJdk0DeF45ZVXNJ5D9fy++LW2yrhx4/Dxxx/j9u3biIyMhLu7u9r+bt26QaFQ5DuuTp06uHTpEvr27ZvvuZLL5fD29kZsbKw0jEcXZmZm+PzzzzF8+HDMnj0bAQEBhQ57KOrvzMPDA3Xq1MG1a9dw+/Zt1K1bV6e4XtSjR48iv5q+c+cOdu/ejTt37iAxMRHZ2dkAcu9TuHnzZqmU2Rs9ejSWLVuGq1ev4vfff8fYsWMLbHvw4EGkpqaie/fuBQ4N69q1K1auXIljx44VWkv6RT169MBnn32G/fv3S0OxVMMjevXqhX379uHPP//Ew4cP4eHhgeDgYGRlZaFLly5qwy1K+nrW5rkCgIULF2L27NmoVq0aAgMD0bx5c60fq4rquc07vK20tG7dWuMQwpK+D+tixowZpX4Nb29v/P333xg7dix+/PFH/PjjjwBy/h917NgRr7zyitZDrsqK6nWgel0YGpPiUvL+++/j119/haOjI/bt26fxH42trS0AIDk5ucDzqCbhsLOzK/R6R44cwaBBg5CRkYF58+ZhypQpWsfq5OSEzp07w8/PDwMHDsRff/2Fjh07FuvNHcj5Zz1gwAAMGDAAQM6Y1eDgYMyePRtnzpzBp59+iv79+6Nt27YAgHv37gEARo4ciZEjRxZ67piYGGn58ePHCAgIKHI8YEJCgpQUjxw5Evv27cP69esREBAAU1NTNGzYEJ06dcIrr7wCf39/tWNVb5YFjZNS1YlNS0vD06dP891kWFDCo3oe09LSCo29IKp/ejKZDFZWVvDx8UGfPn3QuHFjje0LqqdZ1OMDch9jWf7j0MZbb72l9QQVuj5+BwcHODk5ITY2Fg8ePMiXRBT0/Ko+aBW0X/U3r+vzrzJ06FB8+eWXOH/+PH766adC/1ZVf2d+fn5FnjcmJkZvSXFhtVyzs7MxadIk/PTTT4XW8H3xJjN9MTU1xcKFC/Hyyy9j3rx5GD58uMYPOUDu7y8oKKjIMbN536e04evrC2tra5w8eRJJSUmwtrbGgQMH0KBBA3h4eKBHjx74888/sX//fowaNarA8cQlfT1rU3f36NGjOHToEBQKBUJCQnSulR0fHw+g8P9p2tZ3Vu0v6Hkprffh8mzQoEG4f/8+tmzZgoMHD2L16tW4du0arl69ilWrVsHf3x+bN2/W6kNQUfSRw6i2P3v2rMTx6AOT4lLw0UcfYfny5XBwcMDevXvVqkHkpXojiouLQ2JiovQCy0tVPaKwN61jx46hX79+SE5OxieffFKs2Z7yUt2tGxgYiC1bthQ7KX6RhYUFevfujY4dO6J+/fp4+PAh/vvvPykpViqVAIA+ffrA1dW10HPVqFFDWn7rrbdw5MgR+Pr6YsGCBWjWrBkcHR2l3jJ3d3c8evRI7Q3VxMQE69atw6xZsxAYGIijR4/i6NGj+OGHH/DDDz9g4MCB2LJlS7F6xwtTGndUAyh2pYYXC6Ybm9J6/EU9v6X1/KvIZDIsXrwYvXv3xmeffVbohwTV39krr7wCa2vrQs9bnN471XkLUtjv/ttvv8WPP/4INzc3LFu2DB06dICrq6uUmA4fPhx//vlnqU56MWjQIHTo0EG6sXTq1Kka26keZ+3ataVqQAUp7uQ5ZmZm6Ny5M3bt2oXg4GB4eXkhOjpa6tVXJb/79u0rNCkuKW3+Tho1agQzMzOcOXMGkydPxr///qvT35eDgwOAwj/wqF6nhSVbQG7CVdC3fqX9d1gcixcvxo0bN4p93NKlS4u86fFFNjY2UmfT6tWrcenSJVy/fh3Tp0/HgQMH8OGHH+plAhtVXlJYlauichjVhyRVB5ahMSnWs2nTpmHZsmWwt7fH3r17C6wqAQD16tWDlZUVUlJScObMGXTr1i1fmzNnzgCAVAHgRSdOnECfPn2QmJiIWbNm4fPPPy9R/Ko3I31OG2xjYwNfX19s2rQJT548kbZ7enrixo0bGDduXIFfc78oOTkZO3fuhImJCXbu3Cm9webdX9i0kQ0bNkTDhg0xdepUCCFw4MABDB8+HNu3b8fatWsxZswYADlfKd+9exf37t3T2Aur6j1SKBRwcnLSKvbyRPU1cGhoaIFtVI+xONVEKgoPDw/cuHFDeowvio+Pl0oEldfH36tXL3Tv3h1BQUH46quvCvxA5+npidu3b2P69OmFvh+9SDXcKzExUeN+VYkqXahKhP30008aS5LlraBTmr744gv4+flh0aJFePvttzW28fT0BJDzfl0a5QN79OiBXbt2Yf/+/VLPpirp9fLyQp06dRAUFISHDx/i+vXrcHd3zzekpCxezw4ODti2bRsGDBiAXbt2oW/fvtixY0ehw9A0UX2r9vTp0wLbqH4Pd+7cKfRcqteJrkORytLu3bulspTFMX/+/GInxS+ysbHBq6++ilq1aqFVq1bYunWrXpJiVYff06dPERoaqvGbiqJyGNXroKiOsbJSfj5GVQIzZszAl19+CXt7e+zbtw9t2rQptL25uTn69+8PANiwYUO+/ffv38exY8cAAIMHD863/9SpU+jdu7eUEP/vf/8r8WMICgoCgGJ9hapNb054eDgAoHr16tK2vn37AkCBNTQ1iY+PR3Z2Nuzs7PIlxACwbt06rXuXZDIZunfvLo2XVZWzASDV3S3on6DqDcXPz09tbF9F0bp1a9jY2CA2Nhbbtm3Ltz81NRV//fUXAGj8sKYLVZKlqSZuWVM9v6qSWS9SPb916tQpt0kxkJPUyWQyfPXVVwV+da/L3xmQmzxdv349376UlBQcPHiwmNHmUiVoeb8BUrl69ara32Jp6tSpEwYOHIhnz55h0aJFGtt0794d5ubmCA4OLnZngerbq8Je83l7g/fv3w+5XK5W97tHjx6IiorCN998I8XzorJ6PdvZ2WH37t3o1asXDh06hB49ehT7a+9GjRrB3NwcDx48KPADl2o4W1BQkNST+KJnz57hwIEDau1LU0nfv4KDgyFyihsU60efU0qrzqWvYSPVq1eX8hxNOcyRI0cQEREBCwsL9OvXT+M5rly5AkC93r4hMSnWk9mzZ+OLL76Ag4ODVgmxyowZM6QJEXbv3i1tT0lJwbhx45CdnY2AgIB8X8udOXMGvXr1QkJCQrES4lWrVmksjJ+ZmYlVq1Zh+fLlAIDx48drdT4g5yaO2bNna/xUn5qaivnz5+PUqVOQy+VqPcLjx49HjRo18M8//2D69Oka3yCjoqLw888/S+uurq5wdHREXFwc/vjjD7W2J06cwMyZMzXGuHbtWpw9ezbf9sTEROkmlbz/oN9//33I5XJs3boV69atUztm7969+OmnnwAgX23jikKhUGDixIkAcob75O31y8zMxPvvv4+oqCj4+Pho3YtfFNUHoqtXr+rlfCXx9ttvw87ODufOncPChQvVPkidP39e+saloK/Uy4tWrVrh1VdfRWJiIn755ReNbaZOnQoHBwcsW7YMX331FTIyMvK1CQ0Nzfc6VyVrK1asUBtXnpycjPHjxxd7YqC8VDc6rVixQm0YxqNHj/Dmm2+W6QenhQsXwsTEBN99953GISGurq6YPHkykpOTMXDgQFy+fDlfm/T0dGzbti3f1+PavOabNGkCFxcXXLt2DQcPHkT79u3VhtKpnofvv/9ebT2vsnw9W1lZYfv27RgyZAhOnjyJrl27FusmKUtLS7Rv3x5KpRInT57U2Obll19GvXr1kJSUhJEjR+ZLjOPi4vDGG28gOTkZ9evXx6BBg0r0mLRRnt6/CrJhwwZs27ZN4+tYqVRKrwNfX1+9XXPWrFkAcoaHnDt3Ttr+9OlTvPvuuwCASZMmFTiGWdXxVxYfbLRSZnUuKrH//vtPKr3TunXrAkurFDT/t6pOn0wmE127dhWvvfaaqFatmgAg6tWrp7Eki6rupIODQ6HlXF6cHlpVZqVWrVripZdeEsOHDxfdu3cXbm5uAoAwMTERixYtKtbjHzRokPT4a9asKQYOHCiGDx8uevToIcVpamoqVq5cme/YK1euCG9vb+mxdO7cWQwfPly8/PLLomHDhkImkwlXV1e1Y77++mvpeu3atRPDhg0THTt2FDKZTIwcOVJj6RxVjO7u7qJfv35ixIgRol+/flJZusaNG+ermfnTTz9J02y3bNlSDB8+XLoOXiijpFJUuTJVqaDiTMVbVJ1iXa+TlpYmlQy0tLQU/fr1E6+//rrw8vISAESVKlXEmTNn8h2na0m2HTt2CADC3NxcDBgwQIwdO1aMGzdOrd5qYXSpU1xYKaLt27cLhUIhAIj69euLYcOGie7du0sl7jRNPVpUSbCifje6lKV7sSTbi27fvi3MzMyk37um5/zQoUOiatWqAoBwcXER/v7+YsSIEWLAgAFSrfB27dqpHZORkSFat24tAAh7e3vRv39/0bdvX+Hs7Cw8PDzE2LFjCy3JVlg5rRMnTkilo2rXri1ee+010adPH2FpaSkaNWokBg8erPH3qK+SbC8aPXq0Wnm/F183mZmZYvjw4dJ7ZIsWLURAQIB4/fXXRceOHaX647t27VI7TlVDvGrVquK1114T48aNE+PGjRNPnjxRa6eqyQ7krxf97Nkz6X0IgHj48KHGx1Aar2chCp7mOSsrS4wcOVIAOSVDw8PDCzzHi1T/86ZNm1Zgm2vXrknvRba2tqJ3795ixIgRonfv3sLW1lYq63bt2rV8xxb1/ldQ6bXCSrJ9//33AoCwsbERQ4YMkZ7LGzduaP24S5vq+axevbp48803xdy5cwUAMXz4cFGnTh2pTJ2m9/UdO3aIdu3aST+q33GtWrXUtmuiqkVsZmYm+vTpIwICAoSDg4NUrk1VP/9FGRkZomrVqkKhUIinT5/q9XehKybFepC3UHhhP5r+0FT27dsn+vTpI5ycnISFhYWoU6eOmDlzZoHFzbW5nqZ/voGBgWL8+PGiWbNmwtnZWcjlcmFraysaNWok3nnnHXHx4sViP/4HDx6I1atXizfeeEM0a9ZMqi1sa2srmjZtKiZNmiSuXLlS4PEJCQliyZIlwtfXVzg4OAgzMzNRrVo10aZNGzF16lSNE11s3bpVdOjQQTg4OAgbGxvRunVrsXLlSqFUKjUmRCEhIeKDDz4Qbdu2FW5ubsLc3Fy4ubkJX19f8d133+WbREDlxIkT4pVXXhFubm5CLpeLKlWqiP79+2usuyxExUqKhcj5Z79y5UrRvn17YWtrK8zNzUWtWrXE5MmTC6zJqmtSLIQQP//8s2jZsqWwsrKS2mmT5Aqh/6RYiJx/vKNGjRLVq1cXZmZmwsHBQXTr1k389ddfGtuXx6RYiNw6uYU959HR0WLOnDmiZcuW0nNdvXp10aFDBzFv3jxx6dKlfMc8e/ZMTJo0Sfr9eHh4iPHjx4vo6Ogi6xQXlbheunRJvPTSS6JatWpCoVCIOnXqiGnTpomEhIQCf4+llRSHh4dLCWVhr5udO3eKIUOGCA8PD+n10qBBAzF06FCxYcOGfJMVpaamimnTponatWtLHwI0nV9VmxiAxg+Jbdq0EQBEgwYNCn2s+n49C1FwUixETr30d955R/r/dvv27ULjU3n27JmwtrYW7u7uanWiNbVbuHChaN++vXBwcJAmSfH19RULFy7MN1mKSmkkxdnZ2WLRokWiUaNGaq8VXeuul4anT5+KlStXipdeeknUq1dPqkVvZmYmvLy8xKhRo8TNmzc1HqttHlOQjRs3is6dOws7OzthaWkpGjduLBYvXpxvfoa8Nm/eXOAHNkORCVGKt/cSERERvWDSpElYsWIFtm3bprFWOumHTCZDaGioXscm68vAgQMRGBiIc+fO6VTzujQwKSYiIqIypaqJXbt2bZw+fdrQ4VRa5TUpPn36NNq2bYvRo0dj9erVhg5HwhvtiIiIqEw5Oztj/vz5OHPmDDZt2mTocKiMzZw5E7a2tgVWfTEU9hQTERERkdFjTzERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERERk9JgUExEREZHRkxs6gIpKqVQiMjIStra2kMlkhg6HiIiIiF4ghEBiYiLc3d1hYlJ4XzCTYh1FRkbC09PT0GEQERERUREiIiJQvXr1QtswKdaRra0tgJxfsp2dnYGjISIiIqIXJSQkwNPTU8rbCsOkWEeqIRN2dnZMiomIiIjKMW2GuvJGOyIiIiIyekyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIyekyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIyekyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIyekyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIqExGxKdh09oGhw9BIbugAiIiIiMg4+C05CACwVcjRu5GbgaNRx55iIiIiIipTFyPiDB1CPkyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIdBZ46RG8ZwTinzMRhg6lRJgUExEREVGR4lMzNW6fuOEcAGDqpktan8vaovzVemBSTERERESFar8wCM0W7MWKg3dKdJ76brYAgOaeDnqISr+YFBMRERFRoaIS0gCg3NYY1gcmxURERERUKtIys/HxPxdxOzrR0KEUiUkxEREREWlNqRTwnhEI7xmByMhSFtp2ztYr2HT2AXp+HVJG0emOSTERERERae3W49xe387PZ6gryN2YJGk5K1uJG1E5x2YpRekEVwJMiomIiIhIayJPPqsaawwAb7T3QlpmNv4+E4G/T+cvz+b/1SFpuTzOaFf+6mEQERERUZlLycjCjahEtPRyRFxKBpp/ug92Cjkuze+t1s7K3FRa/iKgCW5HJ+GXI6FYdyIcu69E40lSOgCga31n3H+aIrUNj81dTsvMLuVHU3zsKSYiIiIiNJy7B0NWHsPuK1FSlYmEtCy1NqFPkjHguyMAcpLj19t44dLDeGm/KiEGgGfJmXianCGtzxvYsDTDLzEmxUREREQkOXInBso8YyRe7NVNTMtCbRcbfDu0BQDgVGistO/LV5pKy72/Ub+5bkxHn9IIV2+YFBMRERFRgTp9cUBt/aeRrbD3g87o2dA1X9tXW3tqdc4G1ez0Eps+MSkmIiIiogI9ScodAlGzqjV6N3KDiYmsyOM+7lU337ad7/lhau96GNjMXa8x6gNvtCMiIiIiyfnwOKw7ES6tf/16M5y4G4uNZyIwV4txwefn9ER8aia8q1pj15UoXI1MkPY1dLdDQ/fy10sMMCkmIiIiojzyJrEAMLhFdQxuUR1f5BkvXBhHa3M4WpsDABRmpkW0Lj84fIKIiIiISsXHveoBAPo0cjNwJEVjUkxEREREEidrc9z6vC8A7ZJZhVnB6aRvrSoIW9wfP45spbf4SguHTxARERGRxNLMFOZyE4Qt7q9V+/8mdsKQlUfx13jfUo6sdDEpJiIiIjJyIu/czcVUz80WVz/to8doDINJMREREZGRUioF9l+Pxsrgu9I2W4VxpofG+aiJiIiIjNzjhDS8+dsp3IhKVNu+dmxbA0VkWEyKiYiIiIzQsbtPcSMqEVbmphjdwRtjOvrA2dbC0GEZDJNiIiIiIiMkkDOOuFUNR0zrU9/A0RgeS7IRERERkdFjUkxERERUSR288RjeMwKx/uR9te2xyRlYtu8WAODw7SeGCK3cYVJMREREVEm9s/4sAOCTLVcAANEJaZi/7So6LA5CRGyqIUMrdzimmIiIiKiSyszOrT8shEC7hUEGjKZ8Y08xERERkRF4eeUxafmTfg3gV6eqAaMpf5gUExERERmBixFx0rJXFSu87VcTAFDT2dpAEZUvTIqJiIiIKrCsbCVCbsVo3FfF2lxafrdrLbV9nes6I2xxfxz4qGtphldhMCkmIiIiqsDqzt6FN387hQ83Xsi3b2hbLwDAS83cWYu4CEyKiYiIiCow5fN76TSVVjOVyQAAtgrWVigKk2IiIiIiMnpMiomIiIgqqIdxutUarsWb6/JhXzoRERFRBTP3vys4fPsJImJTpG01qlgVedw/E3xx93ESarvYlmZ4FRKTYiIiIqIK5Oz9Z1h7/H6+7b0buQIAvGcEAgDCFvdHRna2Wps23k5o4+1U+kFWQBw+QURERFSh5M5S99/EjhjS0kNaj8wznGL06lNYGXwXAKAwMy278CooJsVEREREFYiFPCfBrWavQDNPB7V92crchDn4ZgyEAHo2dMW4Tj5lGWKFxOETREREROVY2JNk/Hz4HsKeJmP50BaFtrUyz+0RfqmZO97rXpvjh7XEpJiIiIionEnLzEarz/bBxESG5PQsqRbxqdBYXI1MAAA8ik/Ld5zcNHcQwNJXm8FczkEB2mJSTERERFTOjFl9GskZ2fm2v7P+XIHHnLwXC//6rqUZVqXGjw9ERERE5czxe0+l5cD3OhXatrG7PQAg6MZj9Pv2cKnGVZkxKSYiIiIqxxo9T3oBwNnWAvs+7Ky2f0xHb6x/qx2aezogI1tZ1uFVGhw+QURERFRBdKxVBXVcbXH90z5QmOX0bcpkMnSsXRUdalXBgRuPsTL4LtzsFBxPXExMiomIiIgqGEvz/HWHZTIZujdwRfcGHFesi3L3EWLRokVo06YNbG1t4eLigpdffhk3b95Ua5OWloaJEyeiSpUqsLGxQUBAAKKjo9XahIeHo3///rCysoKLiwumTp2KrKwstTbBwcFo2bIlLCwsULt2baxZs6a0Hx4RERGRztr4cDa60lLukuJDhw5h4sSJOHHiBPbt24fMzEz06tULycnJUpsPP/wQ27dvxz///INDhw4hMjISQ4YMkfZnZ2ejf//+yMjIwLFjx/D7779jzZo1mDt3rtQmNDQU/fv3R7du3XDhwgV88MEHeOutt7Bnz54yfbxERERERbk4txd+fKMVRrSrYehQKi2ZEEIU3cxwYmJi4OLigkOHDqFz586Ij4+Hs7MzNmzYgFdeeQUAcOPGDTRo0ADHjx9H+/btsWvXLgwYMACRkZFwdc35CuHHH3/E9OnTERMTA3Nzc0yfPh2BgYG4cuWKdK2hQ4ciLi4Ou3fvLjKuhIQE2NvbIz4+HnZ2dqXz4ImIiMgoec8IlJbDFvc3YCQVW3HytXLXU/yi+Ph4AICTU87XBWfPnkVmZiZ69Oghtalfvz68vLxw/PhxAMDx48fRpEkTKSEGgN69eyMhIQFXr16V2uQ9h6qN6hwvSk9PR0JCgtoPERERkT6dD3+G8WvPGDoMo1Sub7RTKpX44IMP0LFjRzRu3BgAEBUVBXNzczg4OKi1dXV1RVRUlNQmb0Ks2q/aV1ibhIQEpKamwtLSUm3fokWLsGDBAr09NiIiIqK85my9gj9O3FfbNrqDt2GCMULlOimeOHEirly5giNHjhg6FMycORNTpkyR1hMSEuDp6WnAiIiIiKgy2XM1p+NuQNNq+KBHHdR2sTVwRMal3A6fmDRpEnbs2IGDBw+ievXq0nY3NzdkZGQgLi5OrX10dDTc3NykNi9Wo1CtF9XGzs4uXy8xAFhYWMDOzk7th4iIiEhXEbEpmLj+HNafvI+MrNxJN97tWpsJsQGUu6RYCIFJkyZhy5YtOHDgAHx8fNT2t2rVCmZmZggKCpK23bx5E+Hh4fD19QUA+Pr64vLly3j8+LHUZt++fbCzs0PDhg2lNnnPoWqjOgcRERFRadp15RECLz/CJ1uuoPuyYDxOTDd0SEat3A2fmDhxIjZs2ID//vsPtra20hhge3t7WFpawt7eHuPGjcOUKVPg5OQEOzs7TJ48Gb6+vmjfvj0AoFevXmjYsCFGjhyJJUuWICoqCrNnz8bEiRNhYWEBAJgwYQK+//57TJs2DWPHjsWBAwfw999/IzAwsMDYiIiIiPQl74zMEbGp0nJqZrYBoqFy11P8ww8/ID4+Hl27dkW1atWkn40bN0ptvv76awwYMAABAQHo3Lkz3NzcsHnzZmm/qakpduzYAVNTU/j6+uKNN97Am2++iU8//VRq4+Pjg8DAQOzbtw/NmjXDV199hV9++QW9e/cu08dLRERExm1A02qY0be+tJ6exaTYEMp9neLyinWKiYiISFfZSoHBK4/i0oN41Hezxe4POmPt8TCE3IrBL6PaGDq8SqM4+Vq5Gz5BREREVFltPf8Qm84+QGRcKu49yZmt90ZUIgDgTV9vvOnrbcDojBuTYiIiIqIykJGlxAcbLxg6DCpAuRtTTERERFQZKfOMWP2gRx3ITWQGjIZexKSYiIiIqIy95VcT1hb8wr48YVJMREREREaPSTERERERGT0mxURERERk9JgUExERERmAjPfZlStMiomIiIgMYE7/hgCAke1rGDgSAlinmIiIiKjUZWQpseX8Q7VtAa2qI6BVdQNFRC9iUkxERERUStKzsvHH8fv45XAoohLSAADeVaxgZWZq4MjoRUyKiYiIiPQkK1uJ2p/sgpeTFUKmdcP3B+7guwN3AACudhZ4q1NNDGvnBRNO3FHuMCkmIiIi0pOgG48BAOGxKQCAJ0npAIAhLT2waEgTWMjZQ1xe8UY7IiIiIj3JzFZKyyfuPZWWfapYMyEu59hTTERERKQnrnYKaXnoqhMGjISKiz3FRERERHpib2kmLcvzjBvefz3aEOFQMTApJiIiItIzJ2tzHPy4q7R+PSrRcMGQVpgUExEREZUCTycrQ4dAxcCkmIiIiIiMHpNiIiIiolImZ13ico9JMREREZGeZCsFACA2OQMAMLV3PQDAV682M1hMpB2WZCMiIiIqofSsbGw59xAzNl9W2z6xW21M7FbbQFFRcTApJiIiIiqBkb+exOHbTwwdBpUQh08QERERlUDehLhDrSoGjIRKgkkxERERUQk0crcDACx9tRmGt/MycDSkKybFRERERCXgYJUzi52ZqQw1q9oYOBrSFccUExEREelJQ3c7zB3QEA2q2Rk6FComJsVEREREejS2k4+hQyAdcPgEERERERk9JsVERERExSCEwLoT9/HgWYqhQyE94vAJIiIiqrSEEHjwLBWeTlZ6O+dfpyMwe+sVAMCGt9shNCZZb+cmw2FSTERERJWWz8ydAID5AxtidEf9jPW99CBOWh7+80kAgLmpCeq42Orl/GQYHD5BRERElUZaZja8ZwRi2b5batt3Xo7S2zVMZDJp2UJugjd9ayDooy5o6M6KExUZe4qJiIio0vg55B4AYHnQbZy9HyttN5frrx+wtbcj1p8MBwAcme4PZ1sLvZ2bDIc9xURERFRpPIxLlZaP3nkqLXdv4KK3a9ha5EzW0dzTgQlxJcKkmIiIiCqNiw/ipeVRvjVQxyVnhjlri/xfjp+89xTeMwJx5WHOMZnZSqRlZpdNoFTuMCkmIiKiSkMIIS0vGNQY1R0t1fZ5zwiE94xApGVm4/VVJwAAA747gmV7b6L15/vRY9khZGUrNZ43MS2z9B8AGQzHFBMREZFRuPwwtxe565fBavuWH7gDAIhPzURCWhacrM3V9quqWBz8uGupxkiGw55iIiIiqlCUSoErD+ORkZW/R7cw9pZm0nJUQpravsYe2lWOCL75uFjXpIqDSTERERGVO9/uvw3vGYFYHnRb2paRpcTfZyLQ4+tDGPDdESzceV2nc5ubmmD+wIZq27ZN7KT18aLoJlQBcfgEERERlTtf78+pM7xs3y28170Otp5/iC9238Cj+Nwe3qjny0IIpGUqYWluWug5s5Q56WxGthKjO/pg/vZr0r48pYcL9ceJ+4hPyRlbbGaq5UFUIbCnmIiIiMq9T7ZcxqP4NLjYWqBj7Spq+3xm7kSDubtxOzoRKRmaq0dsOBmO7l8d0vp6FyPi4D0jEKuPhqptvxeTjKfJGajuaImJ3WoX/4FQucWeYiIiIir3Mp5XhPj3nQ44dCtGrQaxSs+vQ/JtC7n9BABwISJO62uF3IrBdwdyhm0s2H4NY16YHvqb15tjQNNqkJuyb7Ey4bNJREREFYa8mEMWspW5I4B7N3JV26e6ua5nQ1fkqeSGDzZewN2YZGk9JjFdmhFv87sd8HILDybElRB7iomIiKhCysxW4p8zEWrb2no74VRYbL62bb2d8PXrzdFw7h5p247JfrjzOAm1XWygVBZ8+1znJQelZY4irrz4MYeIiIjKtSO3nyAzOydpVQpA+bxbN+jGY0zddElq91HPuvjs5cZqx9aoYgUAGNvJG1bmcoQu6oewxf2l/bWfz3iX1/q32qmtp2ZmF7v8G1U87CkmIiKiciFbKWAiA2QvlIJ449eT0vLjhDQEXnokrTvbWiAmMV1ar+dmi9BF/aRzHJraDUIIaf3Fc6vk3dzY3R4eDpZ4GJcKAPjlzdb4JugWniRmoGbV/Ek0VQ5MiomIiMjg0jKzUX/ObgDAmdk91PYpzEyQlpnTUysAJKRlSfsOT+smHTewmTuA/IlvQYnwi21+HdUaSelZsLcyQ/cGLlh7/D4AoEdDV/Ro6KqWXFPlw6SYiIiIDO7aowRpuePiA2r7jk73R6vP9wMAar3QU6swM0XY4v5QKgVMTEqWsHZvkHsj3qeDGmNIy+po6mEvbWNCXLkxKSYiIiKDS8tTXzg9Swlrc1MkZ2Rj/5TOqGJjgXNzeiIpLacXV5OSJsSaNPd00Ps5qfxiUkxEREQGF5eaKS1veLsdfGtWUeuZdbI2h5O1uSFCIyPBpJiIiIjKlQ61qha6X2HG4lmkf3xVERERUYWyeEhTAMDMvvUNHAlVJuwpJiIiogqlnputWq1hIn1gTzERERERGT0mxURERERk9JgUExEREZHRY1JMREREREaPSTERERERGT0mxURERGQQgZcewXtGIB7Fpxo6FCImxURERGQYEzecAwBM2XjRwJEQMSkmIiIiA7vyMB5CGDoKMnZMiomIiMigEtOzpF5jIkNhUkxEREQG4VenqqFDIJIwKSYiIqIykZCWiZdXHMU7687iRlQCrM3lAIApPeuinY+TgaMjYyc3dABERERkHK5FJuBCRBwuRAC7r0ZJ44idrM3x1/j28Jm507ABklFjUkxERERlytREhmxl7p1158Kf4Y32NRC2uL8BoyJjx+ETREREVKa8q1gh8L1O0vr1R4kGjIYoB5NiIiIiKnON3O2lZVNmI1QO8GVIREREBjG0jScA4JN+DQ0cCRHHFBMREZGBLA5oisUBTQ0dBhEA9hQTERERETEpJiIiIiLi8AkiIiIqNX+cuI/AS5EY5esNBytzQ4dDVCAmxURERFRq/jwZjmuPEnDiXqy0TRTSnshQOHyCiIiISuSt30+jyfw9AID0rGxsOBmOyw/iAeQmwDJZbvt7McllHCFR0dhTTERERCWy//pjADkz00XFp2HWlssAgN6NXHErOmdijm9eb44fD93D9UcJBouTqDBMiomIiEgvktKykJSeJa3vuRotLVuZy7HrfT98vuMaejR0NUR4RIXSa1IcFhaGffv2QaFQYPDgwbCxsdHn6YmIiKiCaORuhxpVrLDzchSAnF7kng1dMXsAJ+qg8kmnMcULFy6Ej48Pnj17Jm0LDg5G48aNMWHCBIwePRotW7ZEbGxsIWchIiKiyuZZcgYAIFspsHJEK2l7VrbSUCERaUWnpHjr1q3w9vaGo6OjtG369OlQKpVYsGAB3nnnHdy5cwfffPONvuIkIiKicu6rvTexaNcNAMCNqEQDR0NUPDolxWFhYWjYMPfrj8jISJw+fRoTJ07E7Nmz8f3336N79+7YsmVLsc8dEhKCgQMHwt3dHTKZDFu3blXbP3r0aMhkMrWfPn36qLWJjY3FiBEjYGdnBwcHB4wbNw5JSUlqbS5dugQ/Pz8oFAp4enpiyZIlxY6ViIiIcl18XnEiL1uLnJGaI9t7l3E0RMWjU1KckJAABwcHaT0kJAQymQwDBw6UtrVs2RLh4eHFPndycjKaNWuGFStWFNimT58+ePTokfTz559/qu0fMWIErl69in379mHHjh0ICQnB+PHj1eLv1asXatSogbNnz+LLL7/E/PnzsWrVqmLHS0RERDk8HCxhaiJT23Z5QW+ELe4PrypWBoqKSDs63Wjn6uqK+/fvS+v79u2DhYUF2rVrJ21LS0uDTCbTdHih+vbti759+xbaxsLCAm5ubhr3Xb9+Hbt378bp06fRunVrAMB3332Hfv36YenSpXB3d8f69euRkZGB3377Debm5mjUqBEuXLiAZcuWqSXPREREpL0Nb7fD1H8u4VQY7ymiikennuI2bdrgv//+w44dO7B//35s3LgR3bp1g4WFhdQmNDQU7u7uegs0r+DgYLi4uKBevXp455138PTpU2nf8ePH4eDgICXEANCjRw+YmJjg5MmTUpvOnTvD3Dx3usnevXvj5s2bajcP5pWeno6EhAS1HyIiosrk2J0neJyQVuzjbCxY4ZUqPp2S4lmzZiErKwuDBg1C7969kZaWhlmzZkn709PTERISotZzrC99+vTB2rVrERQUhC+++AKHDh1C3759kZ2dDQCIioqCi4uL2jFyuRxOTk6IioqS2ri6qtdIVK2r2rxo0aJFsLe3l348PT31/dCIiIgM5kJEHIb/chJtFwYZOhQig9Dpo13Lli1x4sQJ/PHHHwCA1157DW3btpX2nz9/Ht26dcPw4cP1E2UeQ4cOlZabNGmCpk2bolatWggODkb37t31fj2VmTNnYsqUKdJ6QkICE2MiIqo0zodr/qZUEyEEbj9Ogk9Va5iZqvevNa1uz+ETVCHplBSHh4fD1dUVS5cu1bi/ffv2OlWe0EXNmjVRtWpV3LlzB927d4ebmxseP36s1iYrKwuxsbHSOGQ3NzdER0ertVGtFzRW2cLCQm14CBERUWUWEZuC7w/cwfyXGsHS3FRt38bTEZix+TK8nKzwQY86UAoh7ZvVrwG8qlihf5NqZR0yUYnoNHzCx8cHn3zyib5j0cmDBw/w9OlTVKuW88fn6+uLuLg4nD17Vmpz4MABKJVKaTiHr68vQkJCkJmZKbXZt28f6tWrp1Z7mYiIyFj5LTmIjWciMHHDOQDAW7+fwYx/LwEAHsalAgDCY1Mw5e+LSMnIlo4zMZHhTV9vVLFhRxJVLDolxY6OjnByctJ3LACApKQkXLhwARcuXACQc8PehQsXEB4ejqSkJEydOhUnTpxAWFgYgoKCMGjQINSuXRu9e/cGADRo0AB9+vTB22+/jVOnTuHo0aOYNGkShg4dKt34N3z4cJibm2PcuHG4evUqNm7ciG+//VZteAQREVFFEfYkGRGxKdJ6tlIU0rp4bj9ORFR8GvZfj8ZfpyOgzHPu2i42sLc0k9blpjqlFUTlgk7DJ/z8/KRKDvp25swZdOvWTVpXJaqjRo3CDz/8gEuXLuH3339HXFwc3N3d0atXL3z22WdqQxvWr1+PSZMmoXv37jAxMUFAQACWL18u7be3t8fevXsxceJEtGrVClWrVsXcuXNZjo2IiCqc9KxsdF0aDAD49x1f/HToHoJuPMbiIU3wamtPCCGQpRT5xv5qKz4lEwt3XpfW86bbnWpXxYc96+KP42HIzBZwt1eU4JEQGZZOSfGiRYvQvn17fPrpp5g1axbkcv2VYunatSuEKPgT7p49e4o8h5OTEzZs2FBom6ZNm+Lw4cPFjo+IiKg8iUvJHQoY8MNxaflceBxebe0Jn5k7AQDn5/SEo7U5spUi3wQbhUlIy8K2i5Hq21JzrpmWmQ17SzNM8q9TkodAVC7olM0uWbIETZo0wYIFC/DTTz+hWbNmcHV1zTdZh0wmw6+//qqXQImIiCi/jCyl2rqJDFAKYOv5h/j85cbS9uP3nsLc1ARvrT0DGws5rizojc92XMOvR0JxcW4vra+3eNd1/H48ZwKvv05HYHFAU/08ECID0ykpXrNmjbSsmmpZEybFREREpSsjOzcp3japI4auOoGUjGykZmajzzcham03nAoHACSlZwEAfj0SCgD4Nug2PJ0spXaztlyWlrvVc8b4zrUw7OcTAICfD4eWzgMhMjCdkuLQUP5BEBERlTdu9gq18mi3HydJy2fvP1Mbnrj7Su5kVamZWbgWmTtT64aTOclzG29HLBrSFI8Tc2e5a+HlgPPhcaURPpFB6ZQU16hRQ99xEBERkZaW7L6BlcF38eMbLeFd1Vpt36IhTfDhxosY3cEbdgo5lh+4AyC3V1hlwrrc0qV/n3mgVrGiWz1nvNO1Ntr65FSaSkzLHbf874QOCHuajGE/n8Bvo9vo/bERGQonKyciIjKQSw/icPlhPEa0K7yzKSNLiXfWnYW9pRkm+tfGyuC7AIAJ686ptTOVyTC4RXUMblFd2qZKil9kq5AjMS1nGMWLJdxWj2mr6RBJTWcbnJzVo9A2RBVNiZLitLQ0nD59GpGRkUhPT9fY5s033yzJJYiIiCqtl74/CgBoUM0OLb0KnjzqzuMkBN3Ima1164WHBbYrrE7w4iFNMGNz7ljhYzP88c3+2/jjxH280qo6OtaqKk3UQWSMdE6KV6xYgTlz5iA+Pl7jfiEEZDIZk2IiIqLnzt6Pxbjfz6C9TxVM6VVX2n4rKrHQpFjkqQ784rwc0/rUw5LdNws8tq23E06FxeL1Np7Yey0aB54n17YKM8zu3wCz+jWQSrR5OnVEDSfrAs9FVJnpVMl78+bNmDx5Mjw9PbF06VIIITBo0CAsXLgQffr0gRACAQEB+O233/QdLxERUYV1JuwZ4lIysftqFHq/UBlCGy62Ftjybgdp3VYhx7hOPtK6tblpvmP+nuCLsMX985VNBXKqROWtWdy0ugPsrczytavmYJnnmGKHTVQh6NRT/M0338DFxQXHjx+HlZUVPvroIzRv3hzTp0/H9OnTsWHDBowaNQoTJ07Ud7xEREQVnq2FHInPy6IVRakUCLqe07v7ODEdLfL0KJvIZLCQmyLooy4wlclKbZplGws5dn/gB0szU43JNVFloNNfz6VLl/DSSy/ByspK2padnS0tDx8+HP7+/vj0009LHiEREVEl07ORK3ZM7iSt13G1ydcmK1uJLecfoO+3h7Fs3y21fare3Q61qgAAajnb5KtCoUn7mk46x1zfzQ41qnBoBVVeOvUUZ2ZmwtnZWVq3tLREXFycWptmzZph1apVJQqOiIiosmrsYY9q9go8ik+DhTz/sIe5265K9YJfdOOzPgi+GYMeDVyKdc3xnWuhtosN2vlU0SlmospMp55id3d3tVnsatSogfPnz6u1uX//PuRyVnwjIiLSxZ3nE28Mb+elNl0zAJiZmqBnQ1edhjL413eFtQX/PxO9SKekuE2bNjh3LrdsS58+fXD06FEsWrQIV69exU8//YTNmzejTRsW9SYiIiqJTrWrYmgbTwBA0+r2Bo6GqPLSKSl+9dVXkZ6ejrCwMADAzJkzUb16dcyePRtNmzbFO++8AxsbGyxZskSfsRIREVVaiWmZ8J4RiLXHw/Ltk5uaIGxxf2yb1Cn/gUSkFzp9fzJ48GAMHjxYWnd2dsaFCxfwyy+/4N69e6hRowZGjhwJDw8PvQVKRERUWaVkZOOrvTk308397yre9PU2bEBERkhvg4ocHR0xdepUfZ2OiIio0rN8Xlf4zd9OIi1TaeBoiIxb6RQ0JCIiIgDA7ehEXH+UoHHfsteao7mnAxNionJAq57itWvX6nwBTvNMRESVXWpGNizkJjAxUa8GkZWtRM+vc2auu/FZn3zHNfd0wJZ3O+DAjccY9/sZtX1JadpN7kFE+qFVUjx69Ohil30RQkAmkzEpJiKiSi02OQMtP9sHAAhb3F9tX3pWbg/w0+QMjcfLZDJ0b+AKhZmJ1GP8xe4buPa8d/lCRBz6NalWGqETUR5aJcWrV68u7TiIiIgqpNNhsVq1W3fiPhQaJunQ5Ifgu9Kys42FTnERUfFolRSPGjWqtOMgIiKq8J4kpaNqniRWbpr7LesPwXdR2Jeu1uZypGXm9CY393RAQEsPZCsFRnXwLq1wiSgPTmlDRESkg4wsJczlJrBTmEnbOi85iDEdvTHerxbsrcwgQ24WbKeQI+H5OOG821Xa16qCwEs5s8VuebeDTrPVEZHuWH2CiIiomBrM2Y26s3dh1+VHcLbN7RlOycjGioN30WnJAVyIiFM75sDHXfHpoEZoVcMRL7dwz3fO3o3cpGUmxERlT6ue4po1a+p0cplMhrt37xbdkIiIqAJJzcwGAPx5OgJzBzQEANhbmuHLV5ris8BriIhNxa7Lj3D0zhPpGDMTE7zp613gxBwvNXNHw2q28K5iXerxE1F+WiXFSqVSp0+tQohiH0NERFQeXXkYD6UQ+PVIqLRNnqcEm0wG9Grkht1XohAR+xA/hdxTOz4pIwv2VmYoTG0XW/0GTURa0yopDgsLK+UwiIiIyq+9V6Mw/o+z+bZ3rlM137bN5x9Ky9XsFXgUnwYAMOGICKJyjWOKiYiIinD5Yby03L9pbs3g7w7cwdOk9AKPOzLdX1quZm9ZOsERkV7opfpEcnIyEhISYGdnB2trjoUiIqLKpbmnAwDAwcoMK4a3ROClQAA5E3KM+OWkxmNMTWQwNZHlm9CDiMonnXuKMzIy8L///Q916tSBnZ0dqlevDjs7O9SpUwcLFy5ERobmmXuIiIgqGtVtNTWcrPLty1Lm3D/D0RFEFZtOPcWpqano3r07Tp48CVNTU9SpUwfVqlVDVFQU7t69izlz5mDHjh0ICgqCpSW/LiIiosrJr05VvNG+Bn45fA8tvBwNHQ4RlYBOPcVffPEFTpw4gddeew13797FjRs3cPDgQVy/fh337t3D66+/jhMnTmDJkiX6jpeIiAhAwRWOhBAIfZJcJjHUdbVF70Zu+GdCB8zq10Btn0LO23aIKhKd/mI3btyIli1b4s8//4Snp6favurVq2PDhg1o1aoV/vrrL70ESURElJf3jED4zNyJuzFJ+O/CQwz/+QT+OHEfAOAzcye6LQ3Gzy+URNOnvR92xhvtvTCjb/18+756tRkAYNWbrUvt+kSkfzolxWFhYejVq1ehbXr06MFSbkREVKq6f3UI7/91AcfuPsX650mxyq4rOVMm770ahUsP4vR63bqutvj85SYwM83/bzSgVXWELe6PjrXzl2sjovJLpzHFVlZWiImJKbRNTEwMrKzy35BARESkT2amMmRmax5KERmXKtUXntanHkJjkvHP2Qd4z782pvSqV5ZhElE5p1NPcfv27fHXX3/h6tWrGvdfu3YNGzduhK+vb4mCIyIiKkzvRq74/OXGAIAbUYk4e/+Z2v6059MxA8CS3Tfxz9kHAIDlB+6UXZBEVCHo1FM8a9Ys7N27F23atMG4cePQpUsXuLq6Ijo6GsHBwVi9ejUyMzMxc+ZMfcdLRESEOi42uP04CaM6eCPk1hNpe8APx9Ta5R3e4OlkiYjY1DKLkYgqFp2S4o4dO2LDhg14++23sWLFCqxcuVLaJ4SAvb09fv/9d3Ts2FFvgRIREanEFDCLnKmJDNnK/EMprMxNceCjrmg8bw/Ss5To0cC1tEMkogpG5xntXn31VfTp0wf//fcfzp8/L81o16JFCwwaNAi2trb6jJOIiAj3YpKwKuQe4lIyAQB3Y9RLr+2f0gXdlgYDANp4O6ntMzM1wcBm7th09gEcrMzKJF4iqjhKNM2zra0t3njjDbzxxhv6ioeIiCifhLRMzNp8GYGXHyFveeKYhDS1dj5VreFXpyoO334C76rW+c6z6fmY4k1nH6CKtTkmdKkFR2vzUo2diCqGElUWVyqVePLkCZ48eQKlUqmvmIiIiNTsuRKFHZdyEuLu9V3U9tlbqvf6Hr6dM8Z45ubLOHjzcYHn/CnkHvyWHMSqkLsAcob/7b4ShfSs7AKPIaLKq9hJcWpqKpYuXYo2bdrAwsICrq6ucHV1hUKhQNu2bfHNN98gLS2t6BMREREVwntGILxnBCIrWymVXOvRwBW/jm4jtXmtjSfGdfIBALzfvU6+c8z9T71KUlUbC2m5YTU7JKVnYeHOG4hPycSsLZcxYd1Z1Ju9uzQeDhGVc8UaPhEREYE+ffrgxo0bEEJALpfDxSXnE/vTp09x5swZnD17Fr/99ht27doFDw+PUgmaiIiMx8KdN1DTOWcohIksZ1vY4v5qbV5cV1GYmSAtUylVoTgyvRv+PhOBoW28IDeRoeasnQCAjGwlTtyLLaVHQEQVgdZJsRACw4YNw/Xr1+Hv749PPvkEHTt2hLl5zlisjIwMHDlyBP/73/9w8OBBDB8+HIcOHSq1wImIyDj8djRUWtY8RYe6ze92wJe7b+KXUa2RnJGFP47fl5JqhZkp3vT11jkWba5PRBWT1knx3r17cezYMQwfPhx//PEHZDKZ2n5zc3P4+/ujW7duGDFiBDZu3Ij9+/ejR48eeg+aiIiM075r0UW2aenliD/HtwcAWFvI8VEJZ64Le5KMWZuvAAAuPYgv0bmIqPzSekzxP//8A4VCgeXLl+dLiPOSyWRYvnw5LCwssHHjRr0ESURExsdCnvMvasfkTga5/qjfTsF7RiC6Lg1GVALvlSGq7LROis+ePQs/Pz84OTkV2bZq1arw8/PD2bNnSxQcERGRo7U5mnjYl/l1D92KKfNrEpHhaJ0UR0REoEGDBlqfuGHDhggPD9cpKCIiMj4HbzyG94xA7L0aZehQMHr1KWl5XCcfvNHey4DREFFZ0DopTkhIgIODg9YntrOzQ2Jioi4xERGRERqz5jQAYPwf+b9ltDI3LdNYgm/m9hJXs1fAyrxEc10RUQWgdVKclZUFExPtyxqbmJggKytLp6CIiMi4pWdlq81c90VAUwDAp4Malcr1ktKzEBGbIq0Pa6veMzzJvzYASDWRiajyKdZH3+TkZDx+XPDsQHklJSXpFBAREVGXJcHIyM6ZKdVCbgIPB8sCaxHrQ7elwdJyQMvqWDSkCf48lTMEsHcjN9gpzEr1+kRkeMVKipcuXYqlS5eWVixEREQAgKiENLjaWeC97nXUZqErTTWdrTGhSy0MaZEz8VTY4v7IVgqYmhRccYmIKg+tk+LOnTsXWoqNiIioOIQQ8JmZM6Pc3YX91PYtHtIEg1t6wEJeNmOJZ/Wrj7c61YTJCwkwE2Ii46F1UhwcHFyKYRARkbG5GpkgLfdcljsD6gc96mBo27Kt9tC/qXu+hJiIjAtvpyUiIoNQ5rmT7t6TZNgp5BjVwRv/17lWmcUwrK0XouJT4eFgWWbXJKLyiUkxEREZhJO1ubT8Sb8GGNbOCzYWZftvadGQJmV6PSIqv5gUExGRQZiZ5pT5NJEBb3euaeBoiMjYaV94mIiIqBTwJm4iKg+YFBMRERGR0WNSTERERERGj0kxERHp1b2YJGw+9wApGVmGDoWISGtMiomISK9mbbmMKX9fROclwVhzNBTpWdmGDomIqEglqj5x6tQpnD59GnFxccjOzv+mJ5PJMGfOnJJcgoiIKpik9Jwe4idJ6Zi//RpWhdzD54Mbw7++q4EjIyIqmE5JcWxsLF5++WUcPXoUIk/x9RcxKSYiqrguRsRBYWaKem62Oh0f0LI6jt55gsj4NHyx6yaTYiIq13RKiqdMmYIjR46ga9euGDVqFKpXrw65nCWPiYgqi+T0LAxacRQAcG9hP52mQB7YrBpeb+OJ1346jsxsJQDg+wO3sXTvLeyY3AnOthZ6jZmIqCR0ymR37NiBtm3bIigoiPUliYgqoZjEdGk5SylgrkNSDAAv/otYuvcWAGDm5sv4ZVRrneMjItI3nW60S01NRefOnZkQExGRTnjzHRGVNzolxc2bN0dYWJieQyEiosrsbkyStHzncVIhLYmIyp5OSfG8efOwbds2nDhxQt/xEBFRJXTvSTJ6LDskrSsFkJCaCQDIVhZ8wzYRUVnRaUxxVFQU+vfvjy5dumDEiBFo2bIl7OzsNLZ98803SxQgERFVXPefpkjLLxYr6vl1SBlHQ0RUMJ2S4tGjR0Mmk0EIgTVr1mDNmjX5xhcLISCTyZgUExFVQGkvjPlNSMvE0j03Mdm/DpxtLTB45VGcD49D6KJ+iIhNxfZLkejV0BV1XNXLtz1OTJOW937YGb2YCBNROaVTUrx69Wp9x0FERGUsK1uJo3efomZVa3g6WQHISWL/OH4f3x24I7WTyYCAlcdw+3ES1h6/j++GtcD58DgAQPtFQYhJTIdSAJcfxOPHka0KvF7dFxLmBS81wrxtV/X/wIiIdKBTUjxq1Ch9x0FERGUkLTMb/557gJ8O3UN4bAoaudsh8D0/eM8I1NjeRCZDdUdL3H5+c9zkP89L+6ITcku3pWZqX1GiuacDRnXwxrnwZ6juaKnjIyEi0h+dbrQjIqKKae/VKPgtOYhPtlxBeGzOeN+rkQl45Ydjau0WD2kiLV+IiEOT6g7Suq1Fbn+Ks60FxneuqfX1PxvUCACw9NWmAIBvh7bA1N71i/04iIj0jUkxEZER+fnwPcQkpqOavQJDWnpI28/cfyYtu9pZYEjL6nC0MgMABPxwDMuDbgMA3vStgcPTu0ltlwQ0zTcsojAjfb0Rtrg/arvoNnU0EVFp0SopNjExgVwux61bt6R1U1PTIn849TMRUc5YW/Fi6QUDUZU/W/BSI7TwcpS25+3tlUEGc7kJtk/uhNdaV0feyey2XYyEg5V5mcVLRFRWtMpaVbPXWVlZqa0TEVHhZm+9jHUnwmEhN8HNz/saOhx1eRL1Wf0aYFXIPbXd1R2tsOSVZpjQpRb8v8qpMRyXklmmIRIRlRWtkuLg4OBC14mISLP91x4DANKzlACAKw/jcfFBHEa0qwEAeJyQBidrc8hNy+9otprONgXuszArv3ETERUHxzcQEZWhAd8dAZDT43oyNBYht2Lwaqvq+PLVZgaOrHCfvdwYc7ZewWT/2gCAptXtcelBPHxrVsG/5x4WeFxbb6eyCpGIqESYFBMRGcCXe25Ky/eeJAMAfgi+iy9238DZ2T1QxcbCUKGhtkv+nuGR7WtgZPsa0vq2SZ20Oldrbyf8+44vajvzxjoiKt/4vRcRkYG08HIAAJwLz6n88MXuGwCAhTtvlNH1HdXWD0/rhtdaV8dPhUzAoYtWNZxg/7ySBRFRecWeYiIiAxjYzB1OVmY4Hx4HIYDRq09J+5LSy+ZmtsYe9tg0wRc+Va0BAJ5OOTfWEREZo3LXUxwSEoKBAwfC3d0dMpkMW7duVdsvhMDcuXNRrVo1WFpaokePHrh9+7Zam9jYWIwYMQJ2dnZwcHDAuHHjkJSUpNbm0qVL8PPzg0KhgKenJ5YsWVLaD42IjFDecmYrDuZOnWxlZoqk9NwZ4IJvxkjLyenazwxXUq29nQw6VIOIqLwod0lxcnIymjVrhhUrVmjcv2TJEixfvhw//vgjTp48CWtra/Tu3RtpaWlSmxEjRuDq1avYt28fduzYgZCQEIwfP17an5CQgF69eqFGjRo4e/YsvvzyS8yfPx+rVq0q9cdHRMalZ0NXaTnvOOIXDW6RO5HGkTtPSi2e7PJRLpmIqNwpd8Mn+vbti759NdfyFELgm2++wezZszFo0CAAwNq1a+Hq6oqtW7di6NChuH79Onbv3o3Tp0+jdevWAIDvvvsO/fr1w9KlS+Hu7o7169cjIyMDv/32G8zNzdGoUSNcuHABy5YtU0ueiYhKytPJSlquUcUK95/mTK1sYWaCrDw9wl+/3hxbzhdcxeFFK4PvQAYZ3ulaS6v2aZnZ+OfsA1yMiAMA3IlJQi+tr1Y4VW/42fvPsOLgHaRlKvV0ZiKisqNTT/HYsWOxbdu2Qtvs2LEDY8eO1SmogoSGhiIqKgo9evSQttnb26Ndu3Y4fvw4AOD48eNwcHCQEmIA6NGjB0xMTHDy5EmpTefOnWFunjsrU+/evXHz5k08e5Y71SkRkb4MbuGB/VO6SOt1XG0xrpMPAMC7ipXGYw7dikH48yQ6r5SMLCzZfRNf7L6B1IzCh1rEp2TCe0Yg6s/ZjTlbr0jbQ2OSdXkYGnWsXRX1XG2RlJ6FL/fcxJ3HSUUfRERUzuiUFK9ZswYXLlwotM3Fixfx+++/63L6AkVFRQEAXF1d1ba7urpK+6KiouDi4qK2Xy6Xw8nJSa2NpnPkvcaL0tPTkZCQoPZDRKRJWmY2UjKy8m03yzNBhwxAQ3c7hC3uj+Cp3fK1vR2diFG/nULnLw/m25eSJxFOzSw8KV66N3fIhoeDpTbhF5urnQI73/fDN683R408Cb65vNyN0CMiKlCpvWOlpaVBLi93ozN0tmjRItjb20s/np6ehg6JiMohIQTqz9mNhnP3ICOr4GEEfnWqFnqeE6GxeomniYe9tBw8tatezqmJqYkMLz/vDV8S0BSTutVGqxqORR9IRFRO6Jy1ymQyjduFEIiIiMCuXbvg7u6uc2CauLm5AQCio6NRrVo1aXt0dDSaN28utXn8+LHacVlZWYiNjZWOd3NzQ3R0tFob1bqqzYtmzpyJKVOmSOsJCQlMjIkon/Q8iXBkXGq+/fcW9kNCWiYcrMzz7csr71CHkrA0NwUA+NasotZTXVrMTE3wWhu+NxJRxaP1O6SJiQlMTU1haprzBjt//nxpPe+PXC6Hj48Pzp07h6FDh+o1WB8fH7i5uSEoKEjalpCQgJMnT8LX1xcA4Ovri7i4OJw9e1Zqc+DAASiVSrRr105qExISgszM3Fqg+/btQ7169eDoqLlnw8LCAnZ2dmo/RESFOXP/GQIvP1LbZmIiKzIhLkpyeu7QDE3DNLRR3VHzOGYiImOldU9x586dpd7hkJAQeHl5wdvbO187U1NTODk5wd/fH2+//XaxA0pKSsKdO7m1PENDQ3HhwgU4OTnBy8sLH3zwAT7//HPUqVMHPj4+mDNnDtzd3fHyyy8DABo0aIA+ffrg7bffxo8//ojMzExMmjQJQ4cOlXquhw8fjgULFmDcuHGYPn06rly5gm+//RZff/11seMlIirIx/9cBADIZDk3oxXH/wY3xidbrqDq8xrCWdlKbD73EPeeJOPHQ3eldt2/OoSZfetjWDsvWMhNizzv72Pb4pfD9zCxm3ZVK4iIjIXWSXFwcLC0bGJigjFjxmDu3Ll6D+jMmTPo1i33phPVkIVRo0ZhzZo1mDZtGpKTkzF+/HjExcWhU6dO2L17NxQKhXTM+vXrMWnSJHTv3h0mJiYICAjA8uXLpf329vbYu3cvJk6ciFatWqFq1aqYO3cuy7ERUYm9OLJsWFtPjO9cS5o1Thv9m1TLNx639ie7NLZNz1Ji/vZrWBVyD9+PaImWXoWP4+1S1xld6jprHQsRkbHQaUyxUll6NSi7du0KIQquLi+TyfDpp5/i008/LbCNk5MTNmzYUOh1mjZtisOHD+scJxGRJqZ5suI9H3RGPTdbrY/t36QaAi8/wpRedZGZnfs+++J74uIhTTBj82UAwNTe9bD2eBgi49Pwz5kHRSbFRESkGevlEBHp4KdDd3HgRnShbVztijd98ooRLRG2uD9qOdtI254kpSPgh2Nq7XrkmSVvWFsvvOnrDQBQKnOS5++Cbuc7hoiICqdVT/HYsWMhk8mwcOFCuLq6aj0ph0wmw6+//lqiAImIypvb0YlYtOsGACBscf9SuUZ6nlnhzoXHSctfvdqsyGO/2ncLAPDZjms4HZZT2q2AgkFERPScVknxmjVrIJPJMH36dLi6umLNmjVanZxJMRFVRuGx+WeZ07ewp7kzzk3oUgtjO3nDxTbn3om81ScUZgV/4ffrkVAAgLmpCQY112+JTCKiykarpDg0NOeN1cPDQ22diMgYmZrkdru+9fsZNKxmi+UH7uCrV5uVSvI5o299tXVrCzlGtq8BmQywMi/8bfz/utTEuI4+cLFTFNqOiMjYaZUUX7x4EfXr15dmqKtRo0apBkVEVJ7lrfG7/3o09l/PGVv80T8XcfTOE71co6AJklQ+e7lxvm0XH8ThRlTuFPTjOvlgZt8GeomHiKiy0+pGu8GDB+Ovv/6S1mvWrKlW4oyIyJjk6SjGgKbV1PZtPv9QL9fo9fxmOjPTogcDN/Gwh4kMuBGViL7f5lbVMZfzXmoiIm1p9Y5pZmamNvtbWFgY4uLiSismIqIKwU4hx/fDW6pt86uTM0mHhdwECrOiJ9MoiMLMFGGL++P2//oV2bZzXWfs+aAz+jephrzV23hvHRGR9rQaPuHl5YUjR44gOztbmua5qK/2iIiM0R/j2uHSgziYyGQlSoqLq46rLVaMaIl3HsZjwHdHAKBYE4YQERk7rZLi4cOH49NPP4WTkxOqVKkCAPj666+xevXqQo+TyWS4e/duoW2IiCqCzGwl4lMzpWmXC9O0ukPpB1SAxh720jKHTxARaU+rpHj27NlQKBQIDAxEZGQkZDIZhBCFzjwH5J+FiYiooknJyMLG0xH4OeQeohLS8M8EX5x/Xjc4IS1LrW0tZ/bMEhFVVFolxXK5HDNmzMCMGTMAACYmJvjwww8xd+7cUg2OiMhQspUCP4Xcxc8h9/AsJfeeitG/nUZiunoyfOqT7li29xZm9mOlByKiikqrpPhF8+bNQ9euXfUcChFR+XE6LBZLdt8EAHg5WSFbKfAwLjVfQgwALrYKLA5oWtYhFsm3ZhVDh0BEVGHonBQTEVVmqZnZAIA6LjbY9b4f/j33AKtC7mFwCw+4O1hiyt8XDRxhwe4t7Ie0rOwiJ/YgIqJcJXrHPHr0KNasWYMLFy4gISEBdnZ2aNGiBd5880106tRJXzESERmMhZkJ5KYmeL2NF15v4wUACHo+WUd5ZWIiY0JMRFRMOr9rfvjhh1i+fLl0M53q5ruzZ8/i119/xfvvv49ly5bpLVAiovKipZejoUMgIiI906lez++//45vv/0WderUwfr16xEZGYmsrCw8evQIGzZsQN26dfHtt99i7dq1+o6XiKjU7L0ahY+0GBbhaG2Oqwt64+7CoifWICKiikEmdKib1r59e0RGRuLy5cuwt7fPtz8+Ph5NmjSBu7s7Tpw4oZdAy5uEhATY29sjPj4ednZ2hg6HiPTAe0YgAGDliJawNDfFmNWn0djDDjsm+xk4MiIi0kVx8jWdeoqvXr2KgIAAjQkxANjb2yMgIABXr17V5fRERAYV9jTZ0CEQEVEZK7XpjjgNNBERERFVFDolxY0aNcK///6LpKQkjfsTExPx77//olGjRiUKjojIEB48SzV0CEREVMZ0Sor/7//+Dw8ePICvry/+/fdfPHnyBADw5MkTbNq0CR06dMCDBw/wzjvv6DVYIqKysOFkOD7fcc3QYRARURnSqSTbmDFjcP78eXz//fd47bXXAORM/axUKgEAQghMnjwZo0aN0l+kRESlzMrcFCkZOZN23I3huGIiImOic53i5cuX49VXX9U4eceoUaPg58e7tYnI8IQQuPM4CTWdbWBqUvi9Dq52CoQ+ScaSV5ri6J0n2HYxEnVcbMsoUiIiMiSdkuKQkBDY2dnBz8+PyS8RlTveMwJR28UG+6d0wZpjYViw/RpqOVtjSs96WLz7OiJiUxG2uD/SMrPxeeA1/F/nWvB0skJiWiYAwKeqNV5r7Yn5AxvBVsGZ4YiIjIFOY4q7deuGVatW6TsWIqISuxgRBwC48zjnRuCHz2+auxuTjIkbziEiNvX5/kRM+fsC1p0Ih9+Sg/h8xzU8ScoAADx4lgIgZ5IOuWmpFekhIqJyRKd3excXFygUCn3HQkRUYmmZ2dLyhecJMgDUdbWBjUVur29MYgbOh+fu/+VIqLT8KD6tVGMkIqLyR6ekuGfPnggODoYOk+EREZUqR2tzafnlFUelZNe/visOT+sm7UvNzFJLftv6OEnL1uYcMkFEZGx0SooXL16Mp0+fYvz48YiNjdV3TEREOrMyN5WW895Xt+dqlFrCnJmt/qH+7//zxSutqgMAhrfzKt0giYio3NGpO+SNN96Ag4MDfvvtN6xbtw4+Pj5wdXXNN4udTCZDUFCQXgIlIioOC7kJdr7vh+5fHQIAhD5RL7H2f3+czXfM0lebYemrzcokPiIiKl90SoqDg4Ol5fT0dNy4cQM3btzI145TPRORIdVytjF0CEREVEHoNHxCqVRq9ZOdnV30yYiIDODV50MliIiIAB2TYiKi8qa4N/6O7ugNDwfLUoqGiIgqmmIlxcePH4e/vz9sbW1hZ2eHnj174uTJk6UVGxGRVvp9exg+M3dix6VIhMfm1BhOz1IWekzDanZwtDYri/CIiKgC0HpM8eXLl9G9e3ekpeWWMAoKCsKxY8dw6tQpNGrUqFQCJCIqyrVHCQCASRvOF9jGzDTnHoewxf2lbV8ENEX/5UfwSb8GpRsgERGVe1r3FC9evBhpaWn45JNPEBUVhaioKMyZMwepqan44osvSjNGIiKd7ZjcCZZmptg2qVO+fY3c7RG2uD/e7lzTAJEREVF5IhNaDsTz8vKCt7c3QkJC1LZ36dIFYWFhuH//fqkEWF4lJCTA3t4e8fHxsLOzM3Q4RGUqNSMbb/52ElN61oNvrSqGDgfeMwKl5WWvNcOUvy8CUO8VJiIi41OcfE3rnuLo6Gi0b98+3/Z27dohOjq6+FESUYWSnpWNJ0npAIBPd1zD6bBnGPbzCQNHlaNnQ1cAwKIhTTCkZXW08HLgBBxERFQsWo8pzszMhI1N/pqf1tbWyMzM1GtQRFT+1Ju9GwBwbIY/9l8v3x+Et7zb0dAhEBFRBaPT5B1EZLw6LD5g6BCIiIj0rlhJ8bp163DihPrXpXfu3AEA9OvXL197mUyGwMDAfNuJqHJoXcPRoNd/nJiG346E4eidJwAAE06iSUREOipWUnznzh0pCX7R7t27823jNM9Elc/Yjj6oYmOOL/fcxJn7z5CQlgk7RdnW+83MVuJ/gdex4VQ4Mp7XI67vZgu/Os5lGgcREVUeWifFoaGhpRkHEZVzDlZmiEvJxPB2Xvg88Jq0vfOSg/i/zrUwqkMNWJmXzYisixFxWHMsDADQ0ssB73atDf/6LjBhVzEREelI6/9gNWrUKM04iKgCUeYp5BiXkokvdt/A78fCsG1SR7jYKUr9+hnZOb3DtZyt8e87HfitFBERlVixpnkmInrRsteawU4hR1RCGq5GJpTptU1NZEyIiYhIL5gUE1GJDGlZHQlpWQCAE6FPDRwNERGRbpgUE5He/HToHkb+ehLnw58ZOhQiIqJiYVJMREVKy8xGXEruJD19G7sV2Pbw7ScYvPIYFu+6URahERER6QWTYiIqUGJaJn48dBd+Sw5K2x7GpWJYWy8cntYNoYty6pObPq/68PObrdG/aTUAwK4rj0otrqzsnDv9HsWlldo1iIjIuHBGOyLS6Hz4M4z67ZQ0XlglLTMbAODpZCVtu7uwH4QQkMlkcLI2Q+Cl0kmIk9Kz8M+ZCCzYnlMSLjE9q4gjiIiItMOkmIg0OnQrBglpWfBwsMSHPevi438uFtpe31Ug4lIy0H/5ESwc0gRd6jrDewZnxyQiotLD4RNEVKiu9ZzxSqvqZX7daZsu4WFcKkb9dirfvgbV7Mo8HiIiqtyYFBNRsTStbl8m13mcmK627miVM5X072PbYsu7HQAAw9p6lkksRERU+XH4BBFp5fL8XohNzkA1e0uDXN/VToFnKZmQm8igMDNF2OL+BomDiIgqJybFRKQVW4UZbBVmpXqNxLRMRCekIej6Y1yIiCvVaxEREeXFpJiIyo0m8/dq3P79gdu4/zSljKMhIiJjwjHFRFQutarhKC0v3XsLqZnZqOlsjYa8yY6IiEoBe4qJqNwJmdoN58Kf4ez9nOmi67vZ4p2utdC/STXITflZnoiI9I9JMRGVG5ZmpkjNzIZMBvRs6IqPetZFIw87dKvnovc6yERERHmxy4WISlVCWib+OHEf2UpRrOOsLeSY3L0O/Ou7MiEmIqJSx55iIipVTZ/fPHf+/jMse725NDMdS6oREVF5wp5iIpJ4zwjU23TKyelZiE5Ik9ZP34/FnceJejk3ERGRvjEpJiIAwKP4VGn5vwsPIYo32kFSxdoCJjLgSVIGOi85KG2PiE3FtguRam2XB92G94xALNp5XbeLERER6QmHTxARAEAhN5WW3//rgrRc3NzYu6o1/hrviy9235CqR6gsP3BHbX3ZvlsAgJ9C7qGFlyMyspXFvBoREZF+sKeYiAAAJia5N7OZ5yl7tuFkeLHP1dbHCZsm+GL1mDZaHzNh3VlkKwWcrM1R1cai2NckIiIqCSbFRJRPyLRuJT6HTCZDt3ou2PW+H9p6O+Hy/F7Y92FnaX9aZrZaezuFHJO61ca+DzvD0tz0xdMRERGVKg6fIKJ8qtiYQyaDzuOK82pQzQ5/T/AFAKRl5g6P6PTFQbV2R2f4w1ZhVvILEhER6YA9xURUZrKUuUnxk6R0tX1MiImIyJCYFBMZmS9238DBm4+LbGdrof8vkuJSMqXlL19pqvfzExER6YpJMZEROXnvKX4Ivosxq0+rbc9WChy4Ea22bd1b7QAAa4pxs1xxvNras1TOS0REpAuOKSbSg6xsJbKFgIXc8DeIJadnwW/JQcwZ0ACDW1RX2xf6JFltPTNbiU1nH+CnQ3cR9jQFAODpZAm5iQxNqztw1jkiIjIa7CkmKiEhBGp/sgv1Zu9GRpbh6+x+sfsGYpMz8OHGiwCAa5EJWLD9KpRKAdM8ZdfGrz2Dd9efw8zNlxH2NAUOVmZ4v3sdbJvYCTKZrKDT69Vf49sDANaObVsm1yMiIioIe4qJSkiZp0JDeGwKarvYGC4YAE+TMtTW+y0/DACoamOBjrWrStv3XssdLjG4hQc+f7kxrEthHHFelmbqPenta1ZhbzQREZUL7Ckm0lG2UmDn5UcYtOKItE1uUjY9rNpKSMu9se12dKJafP2bVJOWW9VwLPWEGMiZ7c5OIUcTD/tSvxYREVFxsKeYSAcT/jiL3Vej8m23syxfZcU6LjogLTtYmUvL1ewVWDGiJd6NjMfJe7F4uYVHmcV0aX7vMrsWERGRtthTTKSDvAnxZP/a0vK0TRex71o0vGcEYv3J+4YITU1iepa0rGlYRyN3e4zt5AObMuglJiIiKs+YFBOV0Ic96krL+68/xttrzwAAPtlyxSDxNHS3k5Z/frO1tDx76xV8ueemIUIiIiIq95gUU4V3KzoRc7ZeweqjoaV6nftPkxGfZ/IJlbyFGmo6W5dqDNrwrpITQzsfJ/Rs6Kq279CtGADAo/i0Mo+LiIioPON3plRhnQt/hh+C72Lf8yoKlmamGNPRp1Su9TghDV2+DAaAfNUSZDIZ3urkg9iUDHz1ajOM+/0MDtx4jJpVDZ8g59WnkZvGcdBERETEpJgqqJ9D7uF/O6+rbcvOWxtNz649Sih0/+wBDaXljrWr4sCNx2hSvWwrLKRmZOPvMxFYFXIPAKSaxDc+64NrjxLQ0ssRf50Kx4zNl8s0LiIiooqASTFVSGfuxwIAutR1xtt+NfHGrydL9Xp5J72YuP5cqV5LF/+efYCFO6/jaXJOjeKqNuYY3cEbAKAwM0VLL0cAQE1nw9ZQJiIiKq+YFFOFsfpoKNrXrIIG1XJvJOvZ0BW1XEp/mIKHg6W0HHj5kc7nEUJg7fH78K/vAk8nK32EBgBYtu8WniZnwNPJEuM718KrrapDYZZ/yum2Pk54v3sdtPF20tu1iYiIKgMmxVQh7L8WjQXbrwEA5gxoiPQymE65+1fBuBuTjIWDm6BDrSrS9p4NXaVxzIVRTe7xU8g9xKdkYNvkTth+MRLztl3FvG1X9TqTW2Z2zu/jxzdaoZF74cM2PuxZt9D9RERExqjCVZ+YP38+ZDKZ2k/9+vWl/WlpaZg4cSKqVKkCGxsbBAQEIDpaPYEJDw9H//79YWVlBRcXF0ydOhVZWVkvXorKkWyRO174sx3XEHwzp4qCNqOIk9Oz8CQpvdjXvBuTDACYteUypm26BACwsZCrlTkrzI5Lj/Du+nO4GBGHsKcpuPs4CZci4osdR3HIUL5m1CMiIqooKlxSDACNGjXCo0ePpJ8jR3Kn2f3www+xfft2/PPPPzh06BAiIyMxZMgQaX92djb69++PjIwMHDt2DL///jvWrFmDuXPnGuKhkJbs88wUV81eIS3vvlL0UIZG8/ag9ef7kZhnymOlUmDftWi88sMxdPnyIJ49H4tbkFNhsVrHaponL7W3NIOFvEL+mRERERmVCvnfWi6Xw83NTfqpWrUqACA+Ph6//vorli1bBn9/f7Rq1QqrV6/GsWPHcOLECQDA3r17ce3aNaxbtw7NmzdH37598dlnn2HFihXIyCg8MSLDq+VsjYMfd5XWY5Pz1w3OK29FirsxyRBCYOv5h+jzbQjeXnsGZ+4/w/2nKbgVnVjoefo1cQMAuDsoCm0HAD0buaFvYzd80q8Bjs7wh7OtRZHHEBERkWFVyDHFt2/fhru7OxQKBXx9fbFo0SJ4eXnh7NmzyMzMRI8ePaS29evXh5eXF44fP4727dvj+PHjaNKkCVxdcyc16N27N9555x1cvXoVLVq00HjN9PR0pKfnfgWfkFB4iS4qPZpuIMtLCIGYxHQcuPEYPz0vTwYA5qYm2H0lCh9svAAAsLWQI0spkJqZXeQ1V45ohYjYFK2mQ/ZwsMQPb7Qqsh0RERGVHxWup7hdu3ZYs2YNdu/ejR9++AGhoaHw8/NDYmIioqKiYG5uDgcHB7VjXF1dERWVM2lBVFSUWkKs2q/aV5BFixbB3t5e+vH09NTvA6MSExDYfSUKNWftRNuFQZix+TJCnyRL+52szfHk+TCJFl4OODrTX63nVwiBievPIfjmY43n93SygqO1eek+CCIiIjKICpcU9+3bF6+++iqaNm2K3r17Y+fOnYiLi8Pff/9dqtedOXMm4uPjpZ+IiIhSvR5pJ++scZnZAhPWnUWee/Iwu38Djce52ipgpzBT27buxH0EXn6E0atPa339vKXaCvLgWSoA4M9T4chUln7VDCIiIiq+CpcUv8jBwQF169bFnTt34ObmhoyMDMTFxam1iY6OhptbzphQNze3fNUoVOuqNppYWFjAzs5O7YcMZ9+HnTGgaTUsfbUZzE1zX8a2CvXhDW/51YSZadEVGW5FJ2LOf1e1vv7RGf7oXt8FWyd21PqYv888wOZzDwtt0+XLg1iwXfs4iIiISD8qfFKclJSEu3fvolq1amjVqhXMzMwQFBQk7b958ybCw8Ph6+sLAPD19cXly5fx+HHuV+T79u2DnZ0dGjZsmO/8VD7VcbXF98NbwtLcFFVsLLDgpUaYO6Ahjs3wL9Z5VGXXipMQAzk9xL+ObqPXm+iO3nmC+09TsPpoGABg/cn78J4RiKN3nujtGkRERKRZhUuKP/74Yxw6dAhhYWE4duwYBg8eDFNTUwwbNgz29vYYN24cpkyZgoMHD+Ls2bMYM2YMfH190b59ewBAr1690LBhQ4wcORIXL17Enj17MHv2bEycOBEWFqwSUFGN6uCNsZ18YPvCkIi8tLmhrjQ0fD4D3/k5PdGqhmOB7Z6lqFc/+WTLFQDAu+VwWmkiIqLKpsJVn3jw4AGGDRuGp0+fwtnZGZ06dcKJEyfg7OwMAPj6669hYmKCgIAApKeno3fv3li5cqV0vKmpKXbs2IF33nkHvr6+sLa2xqhRo/Dpp58a6iFRKbM0M0VmdhYGfX8EVuaaX/JjO/rg2N0nuBFVeGk2Xex8309antClFt5eewbNPR0AAJ/vuIZfjoTi8LRuamOh80rPMkwyT0REZEwqXFL8119/FbpfoVBgxYoVWLFiRYFtatSogZ07d+o7NCqnVo5ohU93XMWt6CQkpOXMXHj4doxaG0ergnuYS9MvR0IBAN2WBiNLqc38fERERFQaKtzwCTIeEbEp2HX5UYl7SjvVqYpd73fGt0ObS9uSM8pX7ysTYiIiIsNiUkzl1tRNF/HO+nPwX3oIgZeKns65MKYmMgxq7oH6brYa9w9u6aG2nndK6LIyop0XAMBcboKt5/NXqVAqhdoMfURERKQ/FW74BBmP+NScoQ4P41Lxx4n7ejlnNXuF2rjhsMX9pWWHPEMoOi85qJfrFSQzWwkhBJpWt8elB/H46tVm6NPYDQdvPEZkfJo06x4ACJEzsUjNWTlDfu4t7AcTk6LLzBEREZH22FNM5d7AZu7SmF9tplkuzDevt0ATD3v8+45vvn1NPOyl5WcppdNTrDDL+ZO7GpmAIT8cw+WH8QAAuakM1hZy7JvSBVN714NdnnrL6VlKxKfmxhOTlA4iIiLSLybFVO691ro6Dk/3x8LBTbBwSJMSncveygzbJ3dCqxpOhbZb+mozablZdftCWhaPb80qeLdrLSjMTHA+PE6qOJHwPOm1tpBjYrfaODy9ePWWiYiIqGSYFJNBec8IhPeMQGRlFz79sY2FHMPbeaGRe9EJqvz50AJzefFe3pZ5yrW90qo6Dk/rhpHta2Dzu9rPWldkbKYmmNanPkKmdsPoDt7S9iMvTNBhb6leDWPGv5elZd6UR0REpH9MiqlcOHEvVm/n2vNhZ9Ssao3deeoDa+P/OtcEAIzyrQEA8HSywmcvN4ZpKYzfdbFTYP5LjaT1om6g2301SlpOM9AkJERERJUZk2IqF5QFzVyhg1rONjjwcVfUdLYp1nHWFnKELe6PBYMa6y0WbXWp61zo/qYvDOF4nJCGvt8extXI+NIMi4iIyGgwKSYyoFuf98XmdztgpK93oe1WjWyttv7mb6dw/VEC+i8/UorRERERGQ8mxVQuLA+6jci4VEOHUebM5SZo6eVYaBtVxYq8HsWnlVZIRERERolJMZULZ+4/Q9cvgzF/21XEl1I5tIqqirVFofunbLzAccZEREQlxKSYDMra3BQA4G6vQEa2EmuOhWHB9qvYczUK1x8lAAAeJxhnXd6//88XCjMTrHurndr2X4+EqtUt3nz+IRLSsso6PCIiokqFSTGVCxvebo+PetYFkJPk/d8fZ6V9t6ITCzqsUmvr44Qbn/WFT1Vrte0bToarrXfOc5OenSUnqSQiItIF/4NSuSCTAVfyVFKwtZAjMZ29nyp5y8LZWMiRlOd3s3ZsW5wPf4aEtCxUd7QyRHhEREQVHnuKqdx4lpw7JODoTM7ollfeyTxOf9IDBz7qAgD4Y1xbAEALL8ciy7oRERFRwdhTTOVGTWdrnArLmcTDTmFWRGvjYi43QeB7naBUApbmpqjpbIOwxf0NHRYREVGlwaSYyo35LzVCRrYS/9e5lqFDKZe0meKaiIiIdMOkmMoNhZkplr3W3NBhEBERkRHimGIiIiIiMnpMiqncs7fi+GIiIiIqXUyKqdz6uFddmMiAtzrVNHQoREREVMlxTDGVW5P862CSfx1Dh0FERERGgD3FVKaUSoE9V6OQrRSGDoWIiIhIwp5iKlMBPx7D+fA42CnkmORfG6mZ2YYOiYiIiIhJMZWt8+FxAICEtCws3HkDQM6kHW72CgNGRURERMaOSTEZTM2q1pjQpRZebuEBczlH8hAREZHhMCmmMjW4hQe2nH8IVzsL7JvSBaYmMkOHRERERMQb7ahseThYAgD6Nq7GhJiIiIjKDSbFRERERGT0mBRTmclWCkQlpBk6DCIiIqJ8OKaYSpUQAj4zdwIAalSxwv2nKQAAe0tO3UxERETlB5NiKlV3HidJy/efpsDRygyjOnhjQpdaBoyKiIiISB2TYipVT5MzpOU5AxpiWFtPWJnzZUdERETlC8cUk979L/AavGcE4vDtGLXt4zr5MCEmIiKicolJMendz4dDAQBz/7tq4EiIiIiItMOkmEpNZrZSmtaZiIiIqDxjUkyl5sGzVHyx+4ahwyAiIiIqEpNiKlWctI6IiIgqAt71RHpnaiJDtlLAytwUIdO6wVYhh6mM2TERERGVX+wpJr3r36QaAODjXvVQ1cYCFnJTyE35UiMiIqLyi5kKERERERk9JsVEREREZPSYFBMRERGR0WNSTERERERGj0kx6ZUQAsE3HwMA4lMzDRwNERERkXaYFJNeJKVn4b8LD9H328NISMsCAARefmTgqIiIiIi0wzrFVGLhT1PQ+cuD+banZ2UbIBoiIiKi4mNPMZXY9agEafmjnnWl5abVHQwQDREREVHxsaeYdDJ762WsOxGO83N6wsXWAgDgameByd3rYKRvDfxz5gHGdPQ2bJBEREREWmJPMelk3YlwAMDyA7elbebynJeTg5U53u5ck7PYERERUYXBrIVKJC2T44aJiIio4mNSTFq79CAOmdlKQ4dBREREpHdMio3Mwp3X4T0jEDu1KJf25m+n4D0jEEII/HE8DC99fxR1PtmlduyfpyIQl8J6xERERFSxMSk2MqtC7gEA5my9UmTbkFsxAIBRq09j5+Uoafu768+ptRuz5jQAIDmdQymIiIioYmJSbKSUQuTbFhmXCv+lwbgaGa+2PeRWDI7feyqt21uaaTxnbHKGfoMkIiIiKiNMikkydNUJ3HuSjP7LjwAA7BSaK/YdneEvLb/e2hMDm7mXSXxEREREpYVJcSU0ds1peM8IxL2YJADAP2cicOVhvMa2C7ZfxaKd1wEAUQlpavtUJdUC3+uEqjbm0nYbi9xk2cQE6FCril7jJyIiIiprTIormfCnKThw4zEAYN62q7j0IA5TN13CgO+O4Fz4M7W2CWmZWH00DD+F3Cu0tJq5qUmhs9O90qo6AKBHA5eSPwAiIiIiA+CMdpXEjagE/BB8F9svRkrbjt99iptRidL6kJXHpOVG7vY4fOuJtJ6epV5qTWgYc1wQM1MThC3ur0vYREREROUCk+JK4GpkPF76/iiyleqJrABgaiKT1s1MZcjMzmlz5M4THLnzRG2flbkpMp4nxy+vPKZ245yTde7wibxqOdvo62EQERERGQyHT1RQ3jMC4T0jEM+SM7Dz8iMpId4xuZNauzouttLyoandCjzf1H8uqdUbvhgRJy1nZgvMGdAQthZyfDu0OQBg87sdMKytF8Z29NHDoyEiIiIyLPYUV3AtPtuntt7Yw15t3eT5xx5XOwu4O1iq7TvwURf4f3UIABD4fEKOTrWr4sOedRF46RF+OxoKALj/NBkN3e1weUFv6diWXo5o6eWo18dCREREZChMiiu5tMyc4RDRCelq22tWtYaVee7T37WeMz7sURfNPB0AAK1qOEpJsfaji4mIiIgqJg6fqOBaejlgQpdaGvdlKwUCfjimtm3t2LYAgF9Ht4HIk+5+O7SFlBC/qI23k36CJSIiIiqn2FNcQVW1scCTpHR8OqgxGnvYo4WXA9r7FF0vuHNdZ6lSxKP41ELb3l3YD8kZWbBTaJ7BjoiIiKiyYE9xBSXPU1UCAHo3coO9Vf7ktVs9Z52vYWoiY0JMRERERoFJcSXWv2k19GtSzdBhEBEREZV7TIoroeqOOVUm3upUeLk0B8vc2sNW5qalGhMRERFRecYxxZXQken+0vKdx0kFtrM0N8Vf49tDbiKDmSk/HxEREZHxYlJcyTWoZlfo/vY1i745j4iIiKiyY1JcyTX2sMd3w1qgnptt0Y2JiIiIjBSTYiMwsJm7oUMgIiIiKtc4kJSIiIiIjB6TYiIiIiIyekyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIyekyKiYiIiMjoMSmuIK5FJsB7RiC2XYyEUimQLYShQyIiIiKqNDh5RwUxYd1ZAMB7f57Ht/tvISYxHQBgIefnGiIiIqKSMvqMasWKFfD29oZCoUC7du1w6tQpQ4ekUXJ6lrR8NyYZtgo5PupZF7WcbQwYFREREVHlYNRJ8caNGzFlyhTMmzcP586dQ7NmzdC7d288fvzY0KHl4+5gKS3P7Fsfx2b4Y3L3OjAxkRkwKiIiIqLKwaiT4mXLluHtt9/GmDFj0LBhQ/z444+wsrLCb7/9ZujQ8unftBoAYEDTavi/LrVgqzAzcERERERElYfRJsUZGRk4e/YsevToIW0zMTFBjx49cPz48Xzt09PTkZCQoPZTlqrZK9C6hiPqu9mW6XWJiIiIjIHR3mj35MkTZGdnw9XVVW27q6srbty4ka/9okWLsGDBgrIKL59BzT0wqLmHwa5PREREVJkZbU9xcc2cORPx8fHST0REhKFDIiIiIiI9Mdqe4qpVq8LU1BTR0dFq26Ojo+Hm5pavvYWFBSwsLMoqPCIiIiIqQ0bbU2xubo5WrVohKChI2qZUKhEUFARfX18DRkZEREREZc1oe4oBYMqUKRg1ahRat26Ntm3b4ptvvkFycjLGjBlj6NCIiIiIqAwZdVL8+uuvIyYmBnPnzkVUVBSaN2+O3bt357v5joiIiIgqN5kQQhg6iIooISEB9vb2iI+Ph52dnaHDISIiIqIXFCdfM9oxxUREREREKkyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIyekyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIyekyKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjJ7c0AFUVEIIAEBCQoKBIyEiIiIiTVR5mipvKwyTYh0lJiYCADw9PQ0cCREREREVJjExEfb29oW2kQltUmfKR6lUIjIyEra2tpDJZGVyzYSEBHh6eiIiIgJ2dnZlck0qOT5v5Qufj4qLz13FweeqYqqMz5sQAomJiXB3d4eJSeGjhtlTrCMTExNUr17dINe2s7OrNC9WY8LnrXzh81Fx8bmrOPhcVUyV7XkrqodYhTfaEREREZHRY1JMREREREaPSXEFYmFhgXnz5sHCwsLQoVAx8HkrX/h8VFx87ioOPlcVk7E/b7zRjoiIiIiMHnuKiYiIiMjoMSkmIiIiIqPHpJiIiIiIjB6TYiIiIiIyekyKNQgJCcHAgQPh7u4OmUyGrVu3qu0XQmDu3LmoVq0aLC0t0aNHD9y+fTvfeVJTU2FtbY07d+7g0aNHGD58OOrWrQsTExN88MEHGq/9zTffoF69erC0tISnpyc+/PBDpKWl5Ws3ZswYzJ49G2FhYRg3bhx8fHxgaWmJWrVqYd68ecjIyFBrf+nSJfj5+UGhUMDT0xNLlixR2//zzz/Dz88Pjo6OcHR0RI8ePXDq1Clpf2ZmJqZPn44mTZrA2toa7u7uePPNNxEZGanlb7VsFPXcbd68Gb169UKVKlUgk8lw4cKFAs/l4+OD/fv3Iy0tDaNHj0aTJk0gl8vx8ssva2y/YsUKNGjQAJaWlqhXrx7Wrl2rsd2CBQvwxhtvIDY2FpMnT5aeby8vL7z33nuIj49Xax8eHo7+/fvDysoK/9/emcdFVb1//DMsMwzDMiiLqIQbuCAJlqI2goSypDZqimIRmqlZamVpUhBoLw21zN1M/FGRgkuUpgguLJIhKo4IKpCIGorgEgioyPL8/vA7N65zBwbF1Djv12ter5lznvPcc85zzr3Pfebcc62trTF37lzU1tby2jRs2DBYWVnBzMwMAwcORGJiIk/Hl19+iX79+sHU1BTW1tYYNWoU8vLytHdkC9GYPZozpthcevI0NbfUeHp6IjIyEgAwe/ZsvPDCC5BIJHBxcRGU37ZtG1xcXGBsbAx7e3ssW7ZMUO6HH36AQqHQuf9u3ryJ119/HWZmZpDL5ZgyZQoqKyu5/JSUFCiVStja2kImk8HFxQWbN2/m6YiLi8OLL74IuVzOyURHR+vYY0+G5sx19TlOl75oSGxsLEQikdZzoXoMZGVlISAgAHZ2dpBKpejZsydWrlypIZ+SkoK+fftCIpGgW7du+P7775vdpunTp6Nr166QSqWwsrKCUqlEbm5u4531lLF+/Xo8//zz3IszBg4ciL1792rINcduQ4YMgUgk0vgMHz5cQy+zGwBiaBAfH0+fffYZxcXFEQD65ZdfePkRERFkbm5Ov/76K2VlZdGrr75KnTt3pjt37vDkdu7cST179iQiosLCQpo9ezb98MMP5OLiQu+//77GcTdv3kwSiYQ2b95MhYWFlJiYSLa2tvThhx/y5Gpra8nS0pIyMjJo7969NGnSJEpMTKSCggLauXMnWVtb00cffcTJl5eXk42NDb3++uuUk5NDMTExJJVKacOGDZzMxIkTae3ataRSqejs2bM0adIkMjc3p6KiIiIiKisro6FDh9LWrVspNzeX0tPTqX///vTCCy88Sle3OE3Z7scff6QFCxbQxo0bCQCpVCpBPVlZWWRubk737t2jyspKeuedd+i7774jHx8fUiqVGvLr1q0jU1NTio2NpYKCAoqJiSETExPatWuXhqyrqytt3bqVsrOzacyYMbRr1y46d+4cHTx4kBwcHOi1117jZGtra6l37940dOhQUqlUFB8fT5aWlhQcHMzJvP/++7RkyRI6evQo5efnU3BwMBkaGtKJEyc4GR8fH4qKiqKcnBw6efIkvfLKK/Tcc89RZWVl8zq4mTRmj+aMKTaXnjxNzS0iohs3bpChoSFdvXqViIhmzZpFa9asocDAQOrTp4+gTgMDA1q/fj0VFBTQ7t27ydbWllavXq0hO3r0aFqyZInO/efr60t9+vShI0eOUFpaGnXr1o0CAgK4/EWLFlFISAgdPnyYzp07RytWrCA9PT367bffOJnk5GSKi4ujM2fOcDL6+vqUkJDwkL34+NF1rjc8x+nSF2oKCwupQ4cONHjwYMFzYcMxsGnTJpo9ezalpKRQQUEBRUdHk1Qq5dn3/PnzZGxsTHPmzKEzZ87Q6tWrNfpYlzZt2LCBUlNTqbCwkDIzM2nkyJFkZ2dHtbW1LdCr/w67du2iPXv2UH5+PuXl5dGnn35KhoaGlJOTw8k01243btyg4uJi7pOTk0P6+voUFRXFOzaz232YU9wED5786+vrqV27drRs2TIuraysjCQSCcXExPDKvvXWW/TJJ59o6PTw8BC8kL/33nv08ssv89LmzJlDL730Ei/t0KFDZGtrS/X19YJ1Xrp0KXXu3Jn7vW7dOrKwsKDq6mou7ZNPPqHu3bsLlie67yyYmprSDz/8oFXm6NGjBIAuXryoVeZJou3CTXT/xN6YU7xw4UIaP368RnpQUJDghWDgwIH08ccf89KEbHfp0iUSi8VUXl4ueNxt27aRWCymmpoaIrrvNOjp6XFOBhHR+vXryczMjGfPB+nVqxctWLBAa35paSkBoNTUVK0yLU1j9lCjbUyxufR0oc2WP/74I7m5uWmkh4WFCTrFAQEBNHbsWF7aqlWrqGPHjjyb3Llzh2QyGZ09e1awPg/235kzZwgAHTt2jJPZu3cviUQiunz5stZ2vfLKKzR58mSt+UT3b2pDQkIalXma0DbXtZ3j1Aj1RW1tLQ0aNIgiIyO1ngu1jQE17777Lnl6enK/582bR05OTjyZ8ePHk4+PT7Pb1JCsrCwCQOfOndMq8yxgYWFBkZGR3O+HsVtDvvnmGzI1NdW4SWJ2uw9bPtFMCgsLcfXqVQwdOpRLMzc3h5ubG9LT07m0+vp67N69G0qlUmfdgwYNQmZmJvdX6/nz5xEfH49XXnmFJ7dr1y6MHDkSIpFIUE95eTnatGnD/U5PT4e7uzvEYjGX5uPjg7y8PPz999+COm7fvo2amhqeHqHjiEQiyOVyXZv4zLBr165m2a66uhpGRka8NKlUiqNHj6Kmpoand8iQIVrfKV9eXg4zMzMYGBgAuG87Z2dn2NjYcDI+Pj64desWTp8+Laijvr4eFRUVTdoOQKMyTwKhMcXm0rNDS82boqIiXLx4kUs7ePAgOnTogB49egjqebD/0tPTIZfL8eKLL3IyQ4cOhZ6eHjIyMrTW50F7N4SIcPDgQeTl5cHd3V3XJj5xtM31pmwl1BcLFy6EtbU1pkyZorVcc/Wmp6fzrqfA/TnV8HoqpAPQfv6qqqpCVFQUOnfuDDs7O616nmbq6uoQGxuLqqoqDBw4kEt/GLs1ZNOmTZgwYQJkMhkvndntfzwRV/wZAg9ERA4fPkwA6MqVKzy5cePGkb+/P0/O2tqa6urqNHRqi24REa1cuZIMDQ3JwMCAANA777yjIePg4EC7d+8WLP/nn3+SmZkZfffdd1zasGHDaNq0aTy506dPEwA6c+aMoJ4ZM2ZQly5dNJaEqLlz5w717duXJk6cKJj/NPCg7RrSWKS4qKiIxGIx/f333xp52qIjwcHB1K5dOzp+/DjV19fTsWPHyMbGRmOsDBs2jNasWSNYp2vXrtFzzz1Hn376KZc2depU8vb25slVVVURAIqPjxfUs2TJErKwsKCSkhLB/Lq6Oho+fLhG1PRx05g9iLSPKTaXnj6EbHn37l0yMTHh/dWrRlukeMOGDWRsbEwHDhyguro6ysvLox49ehAA+uOPPzi5qVOnavwTo0ao/xYtWkSOjo4aslZWVrRu3TpBPVu3biWxWKxR/7KyMpLJZGRgYEASiYQ2bdokWP5pRNtcb+wcRyTcF2lpadShQwe6du0aEQmfCxsbA0T357KBgQElJiZyaQ4ODrR48WKe3J49ewgA3b59W+c2ERGtXbuWZDIZAaDu3bs/k1HiU6dOkUwmI319fTI3N6c9e/ZweQ9jt4ZkZGQQAMrIyOClM7v9A4sUPyZ27tyJESNGQE9P9y5OSUnB4sWLsW7dOpw4cQJxcXHYs2cPvvjiC07m7NmzuHLlCry8vDTKX758Gb6+vhg3bhymTp360HWPiIhAbGwsfvnlF40oDnD/QSF/f38QEdavX//Qx3la2bVrFxQKRbOidqGhofDz88OAAQNgaGgIpVKJoKAgAODGwK1bt5CamopXX31Vo/ytW7cwfPhw9OrVC+Hh4Q9d9y1btmDBggXYtm0brK2tBWXee+895OTkIDY29qGP09I0NqbYXHo2SEpKgrW1NZycnHQuM3XqVMycORMjRoyAWCzGgAEDMGHCBAD/zBsiwm+//SY4b1qq/5KTkzF58mRs3LhRo/6mpqY4efIkjh07hkWLFmHOnDlISUl56GP9m2ib642d44T6oqKiAoGBgdi4cSMsLS21Hq+xMZCTkwOlUomwsDB4e3u3eJsA4PXXX4dKpUJqaiocHR3h7+8v+HDt00z37t1x8uRJZGRkYMaMGQgKCsKZM2cANN9uD7Jp0yY4Ozujf//+vHRmtwY8MXf8GQEPREQKCgoEI4zu7u40e/Zs7nePHj3o119/FdSpLbqlUCg0oiHqBe7qKFlERASNGjVKo+zly5fJwcGBAgMDNSJqgYGBGnf0SUlJBIBu3rzJS1+2bBmZm5vz1uI15N69ezRq1Ch6/vnn6fr164IyTwsP2q4hjUWKfX19acWKFYLltEWK1dy7d4/++usvqq2t5R6+U9sjNjaWXFxcNMrcunWLBg4cSF5eXhrRxNDQUI0I2/nz5wkA70E6IuIe+tIW+SS6v9a2Y8eOdP78ea0yjwtt9mhqTLG59PQhZMt33nmHPvjgA0F5bZFiNbW1tVRUVETV1dUUHx9PAKi0tJSIiI4cOUKWlpYaD9401n+bNm0iuVzOS6upqSF9fX2Ki4vjpaekpJBMJuM9LNkYU6ZM0fj35mmksbmu7RynrS9UKhUBIH19fe4jEolIJBKRvr4+F9nTNgZOnz5N1tbWvH/B1AwePFhjDv/f//0fmZmZNatND1JdXU3Gxsa0ZcuWJmWfZry8vLh/p5prt4ZUVlaSmZmZYHlmt39gTnETPHjyVz9o99VXX3Fp5eXlvAft8vPzSSqVUlVVlaBObRfyvn370rx583hpW7ZsIalUyl0QBg4cqPHUaFFRETk4ONCECRMEn9hUPxx07949Li04OFjj4aAlS5aQmZkZpaenC9ZbfRFycnLiLlhPMw/jFFdUVJBEIqHCwkLBck05xQ1xd3fnPe0eEBBAYWFhPJny8nIaMGAAeXh4CI4X9YN2DZdCbNiwgczMzOju3btc2pYtW8jIyEir81hfX0/vvfcetW/fnvLz83Wqf0sjZI+mxhSbS08nQufFDh06UHJysqB8U05xQwIDA2ngwIHc7+DgYAoKCuLJNNV/6gftjh8/zqUlJiZqPGiXnJxMMplM65ImISZPnkweHh46y//bNDXXtZ3jGuuLO3fuUHZ2Nu+jVCrp5ZdfpuzsbKqurtY6BnJycsja2prmzp0rWN958+ZR7969eWkBAQG8B7Ye5vx19+5dkkqlGnP8WcPT05OCgoIeym4NiYqKIolEonEDyezGhznFAlRUVJBKpeLujpcvX04qlYp7sjkiIoLkcjnt3LmTTp06RUqlkrcl27Jly2jkyJEaetU6X3jhBZo4cSKpVCo6ffo0lx8WFkampqYUExND58+fp3379lHXrl25tcolJSVkaGjIrekiun8R79atG3l5eVFRURFv6xU1ZWVlZGNjQ4GBgZSTk0OxsbFkbGzMu6uMiIggsVhMO3bs4OmoqKggovsXoVdffZU6duxIJ0+e5Mk0tgvCv01Ttrtx4wapVCpu7VNsbCypVCquv7Zv307Ozs4aek+fPk0qlYpGjhxJQ4YM4Y6hJi8vj6Kjoyk/P58yMjJo/Pjx1KZNG+4EVlNTQ3K5nDIzM7ky5eXl5ObmRs7OznTu3Dlen6odMvWWbN7e3nTy5ElKSEggKysr3pZsmzdvJgMDA1q7di1PR1lZGSczY8YMMjc3p5SUFJ6M0NqvlqQxe+gypthcenpozJbHjh0jCwsLbtcUNX/++SepVCqaPn06OTo6cuXV7bx27RqtX7+ezp49SyqVimbPnk1GRka8NY9OTk70888/c7917T9fX19ydXWljIwM+v3338nBwYF3k5qUlETGxsYUHBzM03Hjxg1OZvHixbRv3z4qKCigM2fO0FdffUUGBga0cePGFu/flqKpuS50jtOlLx7kwQCB0BjIzs4mKysreuONN3h6G97IqLf2mjt3Lp09e5bWrl2rsbVXU20qKCigxYsX0/Hjx+nixYt0+PBhGjlyJLVp00brsxVPI/Pnz+e2Jzt16hTNnz+fRCIR7du375HtplAoBHetYHbjw5xiAZKTkwmAxkcdraivr6fQ0FCysbEhiURCXl5elJeXx5VXKBSCJ00hnfb29lx+TU0NhYeHU9euXcnIyIjs7Ozo3Xff5RbVR0ZGaixSj4qKEtT74MqYrKwsUigUJJFIqEOHDhQREcHLt7e3F9ShjmyqI6tCH23RoSdBU7bT1l/qdr7xxhv02WefaejV1j9qzpw5Qy4uLiSVSsnMzIyUSiXl5uZy+QcOHKCOHTvqVFcAvGjAhQsXyM/Pj6RSKVlaWtJHH33EO4F5eHg02mYi4bEH4LHfjTdmD13GFJtLTw+N2TIkJIRef/11jTLaxqZ6fF+7do0GDBhAMpmMjI2NycvLi44cOcKVP3fuHEkkEt72Ubr2340bNyggIIBMTEzIzMyMJk+ezN2YEN136oR0NIwCf/bZZ9StWzcyMjIiCwsLGjhwIMXGxrZcpz4GmprrQuc4XfriQR50ioXGQFhYWJNzlej+2HJxcSGxWExdunTROC811abLly+Tn58fWVtbk6GhIXXs2JEmTpzIOwc/C7z11ltkb29PYrGYrKysyMvLi/bt20dEj2a33NxcAsDpagizGx/R/yrOaCGuX78OW1tbFBUV8bbRagleffVVKBQKzJs3r0X1Mu5TW1sLGxsb7N27V+NBhEdl9uzZqK2txbp161pU738ZNpeeHZ5//nmEhITA39+/RfUuX74cBw4cQHx8fIvqba08znPc4xoDDGa3fxODJ12B/xo3b97E8uXLW/wiDgAKhQIBAQEtrpdxn5s3b+LDDz9Ev379Wlx37969eXtNMpqGzaVng3v37uG1116Dn59fi+vu2LEjgoODW1xva+VxneMe5xhgMLv9m7BIMYPBYDAYDAaj1cP2KWYwGAwGg8FgtHqYU8xgMBgMBoPBaPUwp5jBYDAYDAaD0ephTjGDwWAwGAwGo9XDnGIGg8FgMBgMRquHOcUMBoPxBElJSYFIJEJ4ePiTrgqDwWC0aphTzGAwGuXChQsQiUS8j7GxMdq3bw8vLy98/vnnKCgoaJFjhYeHQyQSISUlpUX0NXUc9UdfXx9yuRyOjo4YN24coqKiUFVV9Vjr8LRSVVWFn376Cf7+/nB0dIRUKoVcLoeHhwdiYmK0lquvr8fq1avh7OwMqVQKKysrBAQE4Pz58xqyf/75JxYvXgx3d3e0b98eYrEYdnZ2ePPNN5Gbmyuof9WqVRg+fDg6deoEmUwGuVyOPn36IDw8HDdv3mxWG2/cuIH58+fDyckJxsbGMDY2hr29Pby8vLBgwQKUlJQ0Sx+DwfhvwPYpZjAYjXLhwgV07twZXbt2xRtvvAEAqK6uRmlpKY4ePYqcnBzo6+tj3rx5WLRoEUQi0UMfKzw8HAsWLEBycjKGDBnSQi3QfpzXXnsNvXv3BgDcunULFy5cQFpaGq5fv46OHTsiOjr6sdYDAG7fvo1Lly7B0tISlpaWj/VYupCQkAA/Pz+0bdsWXl5e6NKlC0pLSxEXF4eysjK89957WLNmjUa5qVOnIjIyEk5OThg+fDiuXLmCbdu2wcTEBEeOHIGDgwMnO2HCBGzduhW9e/eGQqGAmZkZsrOzsXfvXkilUiQkJMDd3Z2nv0+fPgAAFxcXtGvXDnfv3kVGRgYyMjJgZ2eHo0ePol27dk22r6ioCIMGDcJff/0FFxcXDBo0CHK5HMXFxfjjjz+Ql5eH/fv3Y+jQoY/YkwwG45njib1gmsFgPBMUFhYSAPLx8RHMT0tLo06dOhEACgkJeaRjhYWFEQBKTk5+JD26HicmJkYj7+7duxQREUF6enokk8koKyvrsdblaUOlUlF0dDRVV1fz0q9evUr29vYEgDIyMnh5SUlJBIDc3d155eLj4wkAeXt78+SjoqLoxIkTGseOiYkhANSrVy+NvDt37gjWNyQkhADQxx9/rFP73nrrLQJACxcuFMw/deoUXbp0SSddDAbjvwVzihkMRqM05RQTEeXm5pJEIiGxWMxzKMrKyigiIoLc3d3J1taWDA0NydbWlgIDA+ncuXM8HR4eHgRA42Nvb8+TKykpoQ8++IC6du1KYrGY2rZtS2PGjKHs7Gyd29SYU6wmPDycAJCfnx8v3d7eXqNOD7ZB6FjJyckUFRVFrq6uJJVKycPDg4iIkpOTCQCFhYUJHqeiooJmz55Ntra2JBaLydnZmbZv3y54/MLCQvL39ycLCwuSyWTk7u5OqampLXazsXjxYgJAy5Yt46UHBAQQAEpNTdUoM2TIEAJAFy9e1OkYjo6OBICuXbumk3xWVhYBoFGjRukk37NnTwJAf//9t07yas6fP09TpkwhOzs7EovF1K5dOwoKCqILFy5oyMbFxdGECROoa9euJJVKyczMjBQKBe3YsUNQd1JSEvn6+nI2tra2JoVCQRs2bNCQ/f333+mVV14hCwsLkkgk1L17d/r888+pqqpKQxYAeXh40NWrV+nNN9+ktm3bkpGREbm5uT32G08G41nE4F8JRzMYjP803bt3h7+/P6Kjo/Hrr79i1qxZAICzZ8/i888/h6enJ0aPHg2ZTIbc3Fxs2bIFe/bswYkTJ2Bvbw8AmDRpEgAgNTUVQUFB6NSpEwBALpdzxykoKMCQIUNQVFQEb29vjBo1CqWlpfj555+RmJiIgwcPws3NrUXa9NFHH2Hp0qVITExEeXk5zM3NH0nfsmXLkJycDKVSCW9vb+jr6zdZpqamBt7e3vj777/x2muv4fbt24iNjYW/vz8SEhLg7e3NyV6+fBmDBg1CcXExfH194erqiry8PAwbNgwvv/zyI9VdjaGhIQDAwIB/6UhJSYFMJsNLL72kUcbHxwcpKSlITU1FYGDgQx9DG3v27AEAbhlMU7Rt2xYAkJ+fj/79++tUJiMjAz4+PqiqqsKIESPg4OCACxcuYPPmzdi7dy/S09PRpUsXTj44OBhisRgKhQK2tra4du0adu3ahbFjx2LVqlXc/FDXf+TIkZDL5VAqlZx8VlYWoqOjMW3aNE52+/btCAgIgEQiwfjx42FtbY19+/Zh4cKFSExMREpKCoyMjHh1Lysrg0KhgLm5OQIDA1FaWoqtW7fCx8cHmZmZOvcbg9EqeNJeOYPBeLrRJVJMRLRp0yYCQIGBgVxaWVkZ3bhxQ0M2KSmJ9PT06O233+alNxXRHDRoEOnr61NCQgIvPS8vj0xNTcnZ2VmnNukSKSYiGjx4MAGggwcPcmkPGymWyWR06tQpjTKNRYoBkFKp5C1JOHDggKA93njjDQJAixYt4qWr7dJYv+pCbW0tOTs7k0gk4kXlKysrCQD17t1bsNyOHTsIAIWGhjZ5jIyMDAJA/fr10yqzYcMGCgsLozlz5nBRaFdXV7p586ZO7Vi1ahUBIGtra/r8888pOTmZysvLtcrfu3ePOnXqRKamphpLPtLS0khfX59GjBjBSy8oKNDQU1FRQc7OzmRubs6L6o4ZM4YA0MmTJzXKXL9+nfteXl5O5ubmJJFIeEt66urqaPz48YJLQtR2f/fdd6muro5Lj4yMJAA0ffp0re1mMFojbPcJBoPRIrRv3x4AcP36dS7N3Nwcbdq00ZD19PSEk5MTDhw4oLN+lUqFP/74A0FBQfDx8eHlOTo6YurUqcjOzkZOTs5DtkAToTY9LNOmTYOzs3Ozy33zzTcQi8Xcby8vL9jb2+PYsWNcWnV1NbZv3w5ra2t89NFHvPKTJ09G9+7dH77i/yM0NBTZ2dmYPHkyL7pYXl4OAFoj6WZmZjw5bZSXlyMoKAh6enpYunSpVrnvvvsOCxYswPLly5GSkgJvb28kJCTAwsJCp3bMnDkTc+fORVlZGRYuXAhPT0/I5XI4OTlh/vz5KC4u5snv3r0bFy5cwNy5c+Hq6srLUygUUCqViI+Px61bt7j0hlFjNSYmJpg0aRLKy8t5tlMjlUo10tRRbQDYuXMnysvL8dZbb+H555/n0tX9ZWBggO+//15Dh0wmw5IlS6Cn98/lPigoCAYGBoL1YDBaM2z5BIPBeKykpKRgxYoVyMjIwPXr11FbW8vlNXT2muLIkSMAgJKSEsE9fdVbeeXm5j6Vfwnr+ld9Q+RyOTp37qyR3rFjR6Snp3O/8/LyUF1djRdffBESiYQnKxKJMGjQIOTl5TW/0v/j22+/xZdffglXV1esXLnyofVo486dOxg9ejRyc3OxaNGiRnf8OH78OID7Nyrp6emYP38++vbti/j4eJ6zqA2RSISlS5di3rx5iI+Px5EjR3D8+HFkZmbizJkz2LBhAxISErhlOOpxl5eXJzjurl69ivr6euTn5+PFF18EAJSWliIiIgJ79+7FxYsXcefOHV6ZK1eucN8nTJiAuLg4DBgwABMnToSXlxcGDx6ssROJSqUCAMG+ee6559ClSxfk5+ejoqICpqamXJ6joyNMTEx48gYGBrCxsUFZWVmT/cVgtCaYU8xgMFoE9YXeysqKS9u+fTvGjx8PExMT+Pj4oFOnTjA2NoZIJML333+Pixcv6qxfvRftnj17uHWkQrTk/sJCbXpYbGxsml1GW/TVwMAA9fX13G91lNLa2rrFjq0mMjIS7777LpydnbF//34NB0tdR22RYHXdtLXl7t27UCqVSE5ORnBwMD799FOd6mVpaYmRI0fCxcUFDg4OmDp1KjIyMnRtFiwtLfHmm2/izTffBHDfuZ05cyZ+/vlnTJs2DVlZWQD+GXebN29uVJ963N28eRP9+vXDpUuX8NJLL2Ho0KGQy+XQ19fHyZMnsXPnTlRXV3Plxo0bh19//RXLly/Ht99+i7Vr10IkEsHT0xNff/01XFxcAPzTj9psaWtri/z8fNy6dYvnFKsj9Q9iYGCAurq6prqJwWhVMKeYwWC0COoXbvTr149LCw8Ph5GRETIzM3n71AJAbGxss/SrL+6rV6/GzJkzH62yOlBZWYnMzEzo6+ujb9++XLqenh7u3bsnWKaxJQKPsn9zU6j7prS0VDD/YV9GsXHjRkyfPh29evXCwYMHeX/nq5HJZLC1tUVhYSHq6uo0HiD8888/AUDD/sD9CLFSqcT+/fsxb948LF68uNl1tLOzQ8+ePXHs2DHcvn0bxsbGzdYBAO3atUN0dDR2796NU6dO4caNG2jbti3Xt7/99htGjBjRpJ5Nmzbh0qVL+OKLLxASEsLLi4iIwM6dOzXKKJVKKJVKVFRU4PDhw4iLi8OmTZvg6+uL3NxcyOVyrh7abHn16lUA2p1gBoPRNGxNMYPBeGTy8/Oxbds2SCQSjB49mksvKChAz549NRyi4uJiwTedqR0qoQiW+u/shssGHidff/01bt++DT8/P16U08LCAqWlpbxlIMD9SKHaAfy36d69OyQSCTIzM3lRSAAgoofqM7VD3LNnTyQlJTUaLffw8EBVVRUOHz6skZeYmAgAGi/jaOgQf/zxx1iyZEmz66imuLiYezPhoyCRSLjdL9Q0d9yp3+6oVCo18tLS0hota2pqCl9fX3z33XeYNGkSSkpKuOi3ej2z0Nse//rrLxQUFKBLly68KDGDwWgezClmMBiPxOHDh+Hj44Pq6mrMnz8fHTp04PLs7e1x7tw5XnTr7t27mDFjBmpqajR0qR/K++uvvzTy+vfvDzc3N8TExGDr1q0a+fX19UhNTX3k9lRXV2Pp0qVYuHAhTExM8OWXX/Ly+/Xrh5qaGt7f6USE4ODgJ/ZqaIlEgrFjx6KkpAQrVqzg5f34449aX52sjcjISEyfPh09evRAUlKS1mUZatTbhoWGhvKi6Hv37uUehlNvvQf8s2Ri//79mDNnDpYtW9ao/uLiYly+fFkjnYgQHh6OkpISeHl5aaynFuLrr7/W2h9r1qxBZWUlevTowUXFlUolnnvuOSxfvhyHDh3SKFNTU4Pff/+d+61uZ8M0ANiyZQvi4+M1yh86dEjwJlAd9VdvsaZUKmFubo6oqCicPn2akyMifPLJJ6itreW2NWQwGA8HWz7BYDB04ty5c9yDRvfu3eNe85ydnQ19fX2EhIQgLCyMV2bWrFmYNWsWXF1dMXbsWNTW1mL//v0gIvTp04dbt6nG09MTIpEIn376KU6fPg1zc3PI5XJuuURMTAw8PT0xYcIErFixAn379oVUKsWlS5eQnp6Oa9eu4e7duzq3aceOHZyDVFlZicLCQhw6dAjXr1+HnZ0dfvrpJ42H9mbOnImoqCi8/fbb2L9/P6ysrJCWloaysjLBNv1bfPnllzhw4ADmz5+P1NRUbp/i3bt3w9fXFwkJCbwdCLSRlJSEadOmgYjg7u6O9evXa8i4uLhg1KhR3G9PT0+8/fbbiIyMRN++fTF8+HAUFxdj69ataNOmDVavXs0r/84772D//v1o164dTE1NBR9gmzRpErdXtXq/5QEDBsDBwQE2Nja4fv060tLSkJeXh/bt22Pt2rU69VN0dDQ+/vhjODs7w83NDdbW1igrK8ORI0dw4sQJSKVSXpslEgl27NgBPz8/eHh44OWXX4azszNEIhEuXryItLQ0tG3blhtHgYGBWLJkCWbNmoXk5GTY29sjKysLBw8exJgxYxAXF8erz+zZs3HlyhUoFAp06tQJIpEIv//+O44ePYoBAwZAoVAAuL8sYuPGjQgICICbmxvGjx8PKysrHDhwAJmZmejfvz/mzp2rUx8wGAwtPMn94BgMxtOPep/ihh+pVEq2trbk6elJoaGhGm+nU1NfX0/ffvstOTk5kZGREbVr146mTJlCpaWlgnv6EhF9//335OzsTBKJRPCNdjdv3qSQkBDq3bs3SaVSMjExIQcHB5o4cSLFxcXp1Cb13sHqj56eHpmZmVG3bt1o7NixFBUVJfiGMDVJSUnk5uZGEomE2rZtS4GBgVRSUtLkG+2EaOqNdkJo67vz58/TuHHjyNzcnIyNjWnw4MGUmppKM2fOJACkUqka6xYiuv8K5gft/eAnKChIo1xdXR2tXLmSnJycuH4ZP3684NjQ9vbChp+G/VVcXEzz5s0jNzc3srKyIgMDAzI1NaW+fftSaGio4F7Y2jhx4gQtWLCAPDw8uLfTSaVS6tGjB82YMYPy8/MFyxUVFdH7779PDg4OJJFIyMzMjHr27Elvv/02bx9rIqKTJ0+St7c3WVhYkKmpKXl4eNCBAwe4vo2KiuJkY2Njyd/fn7p27UrGxsZkbm5Offr0oSVLllBFRYVGPQ4dOkR+fn4kl8tJLBaTo6MjhYaGUmVlpYYs/vdGOyEaG18MRmtFRET0+F1vBoPBYDwpFAoF0tPTUV5errF7BIPBYDDuw9YUMxgMxn+EB188AQA//fQTDh8+jKFDhzKHmMFgMBqBRYoZDAbjP0Lbtm3h6uqKXr16cfvipqSkwNTUFIcPH36oN+oxGAxGa4E5xQwGg/Ef4bPPPsNvv/2GS5cuoaqqClZWVvD09ERoaCh69OjxpKvHYDAYTzXMKWYwGAwGg8FgtHrYmmIGg8FgMBgMRquHOcUMBoPBYDAYjFYPc4oZDAaDwWAwGK0e5hQzGAwGg8FgMFo9zClmMBgMBoPBYLR6mFPMYDAYDAaDwWj1MKeYwWAwGAwGg9HqYU4xg8FgMBgMBqPVw5xiBoPBYDAYDEar5/8BJKqVsJ54FaYAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from xgboost import XGBClassifier\n",
        "\n",
        "# Coded by Tyler Wulff\n",
        "#XGBoost to check feature importance \n",
        "# Training data will be from 2010-2018 and test data will be from the 2023 season\n",
        "df4 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv')\n",
        "df5 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2023Season.csv')\n",
        "\n",
        "y_train = df4[['Covered']]\n",
        "X_train = df4[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "y_test = df5[['Covered']]\n",
        "X_test = df5[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "\n",
        "model = XGBClassifier()\n",
        "model.fit(X_train, y_train)\n",
        "y_pred = model.predict(X_test)\n",
        "predictions = [round(value) for value in y_pred]\n",
        "# evaluate predictions\n",
        "accuracy = accuracy_score(y_test, predictions)\n",
        "accuracy"
      ],
      "metadata": {
        "id": "htcVGIpPBic_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9693e293-3a3f-4524-f3ad-72adc2a6d01b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.573715248525695"
            ]
          },
          "metadata": {},
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "from xgboost import plot_importance\n",
        "#Coded by Tyler Wulff\n",
        "#Plot the feature importance\n",
        "plot_importance(model)\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "4uQ-pJB4Bwlc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "outputId": "b5794b61-b31b-4d8c-e5fe-55e300abb52f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# O/U Prediction"
      ],
      "metadata": {
        "id": "5k1Ps-HpW7BK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "\n",
        "#Coded by Davin Stephens\n",
        "def spread_data():\n",
        "  # access file from github\n",
        "  nba_file = 'https://raw.githubusercontent.com/dcsteph7/Data-Capstone/main/nba_games1.csv'\n",
        "  df = pd.read_csv(nba_file, index_col = 0)\n",
        "  df = df.sort_values(\"date\")\n",
        "  df = df.reset_index(drop=True)\n",
        "\n",
        "  #Calculate point difference\n",
        "  df[\"point_total\"] = df[\"total_opp\"] + df[\"total\"]\n",
        "\n",
        "  return df\n",
        "\n",
        "dF2 = spread_data()\n",
        "dF2"
      ],
      "metadata": {
        "id": "AA92A3IHXGsq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 653
        },
        "outputId": "7a7fd383-98d3-4640-dd4d-5046cf3fafc9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "          mp   mp.1    fg   fga    fg%    3p   3pa    3p%    ft   fta  ...  \\\n",
              "0      240.0  240.0  32.0  90.0  0.356   8.0  32.0  0.250  12.0  20.0  ...   \n",
              "1      240.0  240.0  36.0  79.0  0.456   7.0  20.0  0.350  15.0  22.0  ...   \n",
              "2      240.0  240.0  40.0  85.0  0.471   5.0  15.0  0.333  14.0  18.0  ...   \n",
              "3      240.0  240.0  38.0  77.0  0.494   3.0  13.0  0.231  12.0  31.0  ...   \n",
              "4      240.0  240.0  39.0  75.0  0.520   6.0  13.0  0.462  23.0  28.0  ...   \n",
              "...      ...    ...   ...   ...    ...   ...   ...    ...   ...   ...  ...   \n",
              "25653  240.0  240.0  34.0  85.0  0.400  15.0  38.0  0.395  14.0  19.0  ...   \n",
              "25654  240.0  240.0  31.0  75.0  0.413  11.0  32.0  0.344  21.0  31.0  ...   \n",
              "25655  240.0  240.0  41.0  88.0  0.466   9.0  40.0  0.225  13.0  15.0  ...   \n",
              "25656  240.0  240.0  38.0  92.0  0.413  19.0  46.0  0.413   8.0   8.0  ...   \n",
              "25657  240.0  240.0  34.0  80.0  0.425  11.0  28.0  0.393  11.0  12.0  ...   \n",
              "\n",
              "       usg%_max_opp  ortg_max_opp  drtg_max_opp  team_opp  total_opp  \\\n",
              "0              33.5         164.0         108.0       CLE         94   \n",
              "1              33.6         149.0         118.0       WAS         84   \n",
              "2              31.0         140.0         114.0       LAL         91   \n",
              "3              24.5         146.0         108.0       DAL         99   \n",
              "4              41.7         219.0         121.0       MIA        120   \n",
              "...             ...           ...           ...       ...        ...   \n",
              "25653          36.3         133.0         112.0       GSW        107   \n",
              "25654          36.2         222.0         107.0       GSW        104   \n",
              "25655          94.4         300.0         112.0       BOS         94   \n",
              "25656          42.6         141.0         126.0       BOS         90   \n",
              "25657          31.5         186.0         111.0       GSW        103   \n",
              "\n",
              "       home_opp  season        date    won  point_total  \n",
              "0             1    2013  2012-10-30  False          178  \n",
              "1             0    2013  2012-10-30   True          178  \n",
              "2             1    2013  2012-10-30   True          190  \n",
              "3             0    2013  2012-10-30  False          190  \n",
              "4             1    2013  2012-10-30  False          227  \n",
              "...         ...     ...         ...    ...          ...  \n",
              "25653         0    2022  2022-06-10  False          204  \n",
              "25654         1    2022  2022-06-13  False          198  \n",
              "25655         0    2022  2022-06-13   True          198  \n",
              "25656         1    2022  2022-06-16   True          193  \n",
              "25657         0    2022  2022-06-16  False          193  \n",
              "\n",
              "[25658 rows x 151 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b1713100-77e0-4041-b819-a951c205a1e7\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>mp</th>\n",
              "      <th>mp.1</th>\n",
              "      <th>fg</th>\n",
              "      <th>fga</th>\n",
              "      <th>fg%</th>\n",
              "      <th>3p</th>\n",
              "      <th>3pa</th>\n",
              "      <th>3p%</th>\n",
              "      <th>ft</th>\n",
              "      <th>fta</th>\n",
              "      <th>...</th>\n",
              "      <th>usg%_max_opp</th>\n",
              "      <th>ortg_max_opp</th>\n",
              "      <th>drtg_max_opp</th>\n",
              "      <th>team_opp</th>\n",
              "      <th>total_opp</th>\n",
              "      <th>home_opp</th>\n",
              "      <th>season</th>\n",
              "      <th>date</th>\n",
              "      <th>won</th>\n",
              "      <th>point_total</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>32.0</td>\n",
              "      <td>90.0</td>\n",
              "      <td>0.356</td>\n",
              "      <td>8.0</td>\n",
              "      <td>32.0</td>\n",
              "      <td>0.250</td>\n",
              "      <td>12.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>...</td>\n",
              "      <td>33.5</td>\n",
              "      <td>164.0</td>\n",
              "      <td>108.0</td>\n",
              "      <td>CLE</td>\n",
              "      <td>94</td>\n",
              "      <td>1</td>\n",
              "      <td>2013</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>False</td>\n",
              "      <td>178</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>36.0</td>\n",
              "      <td>79.0</td>\n",
              "      <td>0.456</td>\n",
              "      <td>7.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>0.350</td>\n",
              "      <td>15.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>...</td>\n",
              "      <td>33.6</td>\n",
              "      <td>149.0</td>\n",
              "      <td>118.0</td>\n",
              "      <td>WAS</td>\n",
              "      <td>84</td>\n",
              "      <td>0</td>\n",
              "      <td>2013</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>True</td>\n",
              "      <td>178</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>85.0</td>\n",
              "      <td>0.471</td>\n",
              "      <td>5.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>0.333</td>\n",
              "      <td>14.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>...</td>\n",
              "      <td>31.0</td>\n",
              "      <td>140.0</td>\n",
              "      <td>114.0</td>\n",
              "      <td>LAL</td>\n",
              "      <td>91</td>\n",
              "      <td>1</td>\n",
              "      <td>2013</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>True</td>\n",
              "      <td>190</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>77.0</td>\n",
              "      <td>0.494</td>\n",
              "      <td>3.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>0.231</td>\n",
              "      <td>12.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>...</td>\n",
              "      <td>24.5</td>\n",
              "      <td>146.0</td>\n",
              "      <td>108.0</td>\n",
              "      <td>DAL</td>\n",
              "      <td>99</td>\n",
              "      <td>0</td>\n",
              "      <td>2013</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>False</td>\n",
              "      <td>190</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>39.0</td>\n",
              "      <td>75.0</td>\n",
              "      <td>0.520</td>\n",
              "      <td>6.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>0.462</td>\n",
              "      <td>23.0</td>\n",
              "      <td>28.0</td>\n",
              "      <td>...</td>\n",
              "      <td>41.7</td>\n",
              "      <td>219.0</td>\n",
              "      <td>121.0</td>\n",
              "      <td>MIA</td>\n",
              "      <td>120</td>\n",
              "      <td>1</td>\n",
              "      <td>2013</td>\n",
              "      <td>2012-10-30</td>\n",
              "      <td>False</td>\n",
              "      <td>227</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25653</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>85.0</td>\n",
              "      <td>0.400</td>\n",
              "      <td>15.0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>0.395</td>\n",
              "      <td>14.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>...</td>\n",
              "      <td>36.3</td>\n",
              "      <td>133.0</td>\n",
              "      <td>112.0</td>\n",
              "      <td>GSW</td>\n",
              "      <td>107</td>\n",
              "      <td>0</td>\n",
              "      <td>2022</td>\n",
              "      <td>2022-06-10</td>\n",
              "      <td>False</td>\n",
              "      <td>204</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25654</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>75.0</td>\n",
              "      <td>0.413</td>\n",
              "      <td>11.0</td>\n",
              "      <td>32.0</td>\n",
              "      <td>0.344</td>\n",
              "      <td>21.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>...</td>\n",
              "      <td>36.2</td>\n",
              "      <td>222.0</td>\n",
              "      <td>107.0</td>\n",
              "      <td>GSW</td>\n",
              "      <td>104</td>\n",
              "      <td>1</td>\n",
              "      <td>2022</td>\n",
              "      <td>2022-06-13</td>\n",
              "      <td>False</td>\n",
              "      <td>198</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25655</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>41.0</td>\n",
              "      <td>88.0</td>\n",
              "      <td>0.466</td>\n",
              "      <td>9.0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>0.225</td>\n",
              "      <td>13.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>...</td>\n",
              "      <td>94.4</td>\n",
              "      <td>300.0</td>\n",
              "      <td>112.0</td>\n",
              "      <td>BOS</td>\n",
              "      <td>94</td>\n",
              "      <td>0</td>\n",
              "      <td>2022</td>\n",
              "      <td>2022-06-13</td>\n",
              "      <td>True</td>\n",
              "      <td>198</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25656</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>92.0</td>\n",
              "      <td>0.413</td>\n",
              "      <td>19.0</td>\n",
              "      <td>46.0</td>\n",
              "      <td>0.413</td>\n",
              "      <td>8.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>...</td>\n",
              "      <td>42.6</td>\n",
              "      <td>141.0</td>\n",
              "      <td>126.0</td>\n",
              "      <td>BOS</td>\n",
              "      <td>90</td>\n",
              "      <td>1</td>\n",
              "      <td>2022</td>\n",
              "      <td>2022-06-16</td>\n",
              "      <td>True</td>\n",
              "      <td>193</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25657</th>\n",
              "      <td>240.0</td>\n",
              "      <td>240.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>80.0</td>\n",
              "      <td>0.425</td>\n",
              "      <td>11.0</td>\n",
              "      <td>28.0</td>\n",
              "      <td>0.393</td>\n",
              "      <td>11.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>...</td>\n",
              "      <td>31.5</td>\n",
              "      <td>186.0</td>\n",
              "      <td>111.0</td>\n",
              "      <td>GSW</td>\n",
              "      <td>103</td>\n",
              "      <td>0</td>\n",
              "      <td>2022</td>\n",
              "      <td>2022-06-16</td>\n",
              "      <td>False</td>\n",
              "      <td>193</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>25658 rows × 151 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b1713100-77e0-4041-b819-a951c205a1e7')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-b1713100-77e0-4041-b819-a951c205a1e7 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-b1713100-77e0-4041-b819-a951c205a1e7');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#We had to pull multiple datasets from kaggle to find spreads over the last 10 years\n",
        "#These data sets consisted of different game ids that had to be mapped to merge with other dataset\n",
        "#Coded by Tyler Wulff\n",
        "\n",
        "## this function maps the team ids with the team name\n",
        "def team_ids():\n",
        "    teams = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_teams_all.csv')\n",
        "   \n",
        "    team_dict = {}\n",
        "    ##Loop through the dataframe and convert ids to names\n",
        "    for i in range(len(teams.index)):\n",
        "        if isinstance(teams['abbreviation'][i], str):\n",
        "            team_dict[teams['team_id'][i]] = teams['abbreviation'][i]\n",
        "\n",
        "    return team_dict\n",
        "\n",
        "## This function maps the game ids to a date\n",
        "def game_ids():\n",
        "    teams = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_games_all.csv')\n",
        "    \n",
        "    team_dict = {}\n",
        "    #Loop through dataframe and convert ids to dates\n",
        "    for i in range(len(teams.index)):\n",
        "        if isinstance(teams['game_date'][i], str):\n",
        "            team_dict[teams['game_id'][i]] = teams['game_date'][i]\n",
        "\n",
        "    return team_dict\n",
        "\n",
        "#Add spreads into dataframe using previous two functions to get spread using team name and date\n",
        "def add_teams():\n",
        "    moneylines = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_money_line.csv')\n",
        "    over_under = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_totals.csv')\n",
        "    spreads = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_spread.csv')\n",
        "    games = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_games_all.csv')\n",
        "    \n",
        "    df = pd.DataFrame()\n",
        "    temp_dict = team_ids()\n",
        "    game_dict = game_ids()\n",
        "    #Loop through spreads and add spread along with date, book name, and teams\n",
        "    for i in range(len(over_under.index)):\n",
        "        if (over_under['book_name'][i] == 'Pinnacle Sports'):\n",
        "            temp = pd.DataFrame({\"date\":[game_dict[over_under['game_id'][i]]],\n",
        "                            \"book_name\":[over_under['book_name'][i]],\n",
        "                            \"team\":[temp_dict[over_under['team_id'][i]]],\n",
        "                            \"away_team\": [temp_dict[over_under['a_team_id'][i]]],\n",
        "                            \"total1\": [over_under['total1'][i]],\n",
        "                            \"total2\": [over_under['total2'][i]]})\n",
        "            #df = df.append(temp, ignore_index = True)\n",
        "            df = pd.concat([df, temp])\n",
        "    return df\n",
        "\n",
        "#Add spreads using bovada sportsbook same as previous function\n",
        "def add_teams_2():\n",
        "    moneylines = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_money_line.csv')\n",
        "    over_under = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_totals.csv')\n",
        "    spreads = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_betting_spread.csv')\n",
        "    games = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/nba_games_all.csv')\n",
        "    \n",
        "    df = pd.DataFrame()\n",
        "    temp_dict = team_ids()\n",
        "    game_dict = game_ids()\n",
        "    #Loop through spreads and add spread along with date, book name, and teams\n",
        "    for i in range(len(over_under.index)):\n",
        "        if (over_under['book_name'][i] == 'Bovada'):\n",
        "            temp = pd.DataFrame({\"date\":[game_dict[over_under['game_id'][i]]],\n",
        "                            \"book_name\":[over_under['book_name'][i]],\n",
        "                            \"team\":[temp_dict[over_under['team_id'][i]]],\n",
        "                            \"away_team\": [temp_dict[over_under['a_team_id'][i]]],\n",
        "                            \"total1\": [over_under['total1'][i]],\n",
        "                            \"total2\": [over_under['total2'][i]]})\n",
        "            #df = df.append(temp, ignore_index = True)\n",
        "            df = pd.concat([df, temp])\n",
        "    return df\n",
        "        \n",
        "prev_lines = add_teams()\n",
        "prev_lines = prev_lines.reset_index(drop=True)\n",
        "bovad_lines = add_teams_2()\n",
        "bovad_lines = bovad_lines.reset_index(drop=True)\n",
        "bovad_lines"
      ],
      "metadata": {
        "id": "3Bl9PxX6XaYy",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "658468c5-909c-47b6-b5df-800b7af021e7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "             date book_name team away_team  total1  total2\n",
              "0      2012-01-09    Bovada  NOP       DEN   192.0   192.0\n",
              "1      2012-01-10    Bovada  OKC       MEM   193.0   193.0\n",
              "2      2012-01-11    Bovada  OKC       NOP   184.0   184.0\n",
              "3      2012-01-11    Bovada  HOU       SAS   199.0   199.0\n",
              "4      2012-01-11    Bovada  ORL       POR   185.0   185.0\n",
              "...           ...       ...  ...       ...     ...     ...\n",
              "14537  2018-05-03    Bovada  PHI       BOS   207.0   207.0\n",
              "14538  2018-05-04    Bovada  GSW       NOP   233.0   233.0\n",
              "14539  2018-05-05    Bovada  BOS       PHI   207.0   207.0\n",
              "14540  2018-05-06    Bovada  GSW       NOP   231.0   231.0\n",
              "14541  2018-05-09    Bovada  PHI       BOS   206.0   206.0\n",
              "\n",
              "[14542 rows x 6 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-a411f90f-dec3-4b17-b4c1-1ebef10ded5d\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>date</th>\n",
              "      <th>book_name</th>\n",
              "      <th>team</th>\n",
              "      <th>away_team</th>\n",
              "      <th>total1</th>\n",
              "      <th>total2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2012-01-09</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>NOP</td>\n",
              "      <td>DEN</td>\n",
              "      <td>192.0</td>\n",
              "      <td>192.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2012-01-10</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>OKC</td>\n",
              "      <td>MEM</td>\n",
              "      <td>193.0</td>\n",
              "      <td>193.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2012-01-11</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>OKC</td>\n",
              "      <td>NOP</td>\n",
              "      <td>184.0</td>\n",
              "      <td>184.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2012-01-11</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>HOU</td>\n",
              "      <td>SAS</td>\n",
              "      <td>199.0</td>\n",
              "      <td>199.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2012-01-11</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>ORL</td>\n",
              "      <td>POR</td>\n",
              "      <td>185.0</td>\n",
              "      <td>185.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14537</th>\n",
              "      <td>2018-05-03</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>PHI</td>\n",
              "      <td>BOS</td>\n",
              "      <td>207.0</td>\n",
              "      <td>207.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14538</th>\n",
              "      <td>2018-05-04</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>GSW</td>\n",
              "      <td>NOP</td>\n",
              "      <td>233.0</td>\n",
              "      <td>233.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14539</th>\n",
              "      <td>2018-05-05</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>BOS</td>\n",
              "      <td>PHI</td>\n",
              "      <td>207.0</td>\n",
              "      <td>207.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14540</th>\n",
              "      <td>2018-05-06</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>GSW</td>\n",
              "      <td>NOP</td>\n",
              "      <td>231.0</td>\n",
              "      <td>231.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14541</th>\n",
              "      <td>2018-05-09</td>\n",
              "      <td>Bovada</td>\n",
              "      <td>PHI</td>\n",
              "      <td>BOS</td>\n",
              "      <td>206.0</td>\n",
              "      <td>206.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>14542 rows × 6 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a411f90f-dec3-4b17-b4c1-1ebef10ded5d')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-a411f90f-dec3-4b17-b4c1-1ebef10ded5d button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-a411f90f-dec3-4b17-b4c1-1ebef10ded5d');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Display game line information for presentation\n",
        "game_lines_2023_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2021_2023_Season_Odds.csv'\n",
        "game_lines_2023_data = pd.read_csv(game_lines_2023_file)\n",
        "game_lines_2023_data"
      ],
      "metadata": {
        "id": "fR9Q2il5aBwI",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "ee1d4cdc-bc80-4613-e23f-ff6be4215476"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     Unnamed: 0        Date Away Home Favorite  Line Over/Under\n",
              "0             0  2020-12-22  GSW  BKN      BKN  -7.0      235.0\n",
              "1             1  2020-12-22  LAC  LAL      LAL  -2.0      219.5\n",
              "2             2  2020-12-23  CHA  CLE      CHA  -3.0      217.0\n",
              "3             3  2020-12-23  NYK  IND      IND  -7.5      215.5\n",
              "4             4  2020-12-23  MIA  ORL      MIA  -4.5      218.0\n",
              "...         ...         ...  ...  ...      ...   ...        ...\n",
              "3212        917  2023-02-27  DET  CHA      CHA  -7.0      233.5\n",
              "3213        918  2023-02-27  MIA  PHI      PHI  -5.5      217.0\n",
              "3214        919  2023-02-27  BOS  NYK      BOS  -2.0      224.5\n",
              "3215        920  2023-02-27  ORL  NOP      NOP  -5.0      229.5\n",
              "3216          \u001a         NaN  NaN  NaN      NaN   NaN        NaN\n",
              "\n",
              "[3217 rows x 7 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9e0e5cf3-1bea-4d12-8f3b-d46d6c9115ea\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>Date</th>\n",
              "      <th>Away</th>\n",
              "      <th>Home</th>\n",
              "      <th>Favorite</th>\n",
              "      <th>Line</th>\n",
              "      <th>Over/Under</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>2020-12-22</td>\n",
              "      <td>GSW</td>\n",
              "      <td>BKN</td>\n",
              "      <td>BKN</td>\n",
              "      <td>-7.0</td>\n",
              "      <td>235.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>2020-12-22</td>\n",
              "      <td>LAC</td>\n",
              "      <td>LAL</td>\n",
              "      <td>LAL</td>\n",
              "      <td>-2.0</td>\n",
              "      <td>219.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2020-12-23</td>\n",
              "      <td>CHA</td>\n",
              "      <td>CLE</td>\n",
              "      <td>CHA</td>\n",
              "      <td>-3.0</td>\n",
              "      <td>217.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>2020-12-23</td>\n",
              "      <td>NYK</td>\n",
              "      <td>IND</td>\n",
              "      <td>IND</td>\n",
              "      <td>-7.5</td>\n",
              "      <td>215.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>2020-12-23</td>\n",
              "      <td>MIA</td>\n",
              "      <td>ORL</td>\n",
              "      <td>MIA</td>\n",
              "      <td>-4.5</td>\n",
              "      <td>218.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3212</th>\n",
              "      <td>917</td>\n",
              "      <td>2023-02-27</td>\n",
              "      <td>DET</td>\n",
              "      <td>CHA</td>\n",
              "      <td>CHA</td>\n",
              "      <td>-7.0</td>\n",
              "      <td>233.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3213</th>\n",
              "      <td>918</td>\n",
              "      <td>2023-02-27</td>\n",
              "      <td>MIA</td>\n",
              "      <td>PHI</td>\n",
              "      <td>PHI</td>\n",
              "      <td>-5.5</td>\n",
              "      <td>217.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3214</th>\n",
              "      <td>919</td>\n",
              "      <td>2023-02-27</td>\n",
              "      <td>BOS</td>\n",
              "      <td>NYK</td>\n",
              "      <td>BOS</td>\n",
              "      <td>-2.0</td>\n",
              "      <td>224.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3215</th>\n",
              "      <td>920</td>\n",
              "      <td>2023-02-27</td>\n",
              "      <td>ORL</td>\n",
              "      <td>NOP</td>\n",
              "      <td>NOP</td>\n",
              "      <td>-5.0</td>\n",
              "      <td>229.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3216</th>\n",
              "      <td>\u001a</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3217 rows × 7 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9e0e5cf3-1bea-4d12-8f3b-d46d6c9115ea')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-9e0e5cf3-1bea-4d12-8f3b-d46d6c9115ea button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-9e0e5cf3-1bea-4d12-8f3b-d46d6c9115ea');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "working_spreads_file = 'https://raw.githubusercontent.com/austinfett/ML-Capstone/main/WorkingSpreads.csv'\n",
        "working_spreads_data = pd.read_csv(working_spreads_file)\n",
        "working_spreads_data['O/U_Line'] = 0\n",
        "\n",
        "for i in range(len(working_spreads_data.index)):\n",
        "  if working_spreads_data['date'][i] < '2018-11-06':\n",
        "    for j in range(len(bovad_lines.index)):\n",
        "      if working_spreads_data['date'][i] == bovad_lines['date'][j]:\n",
        "        if working_spreads_data['team'][i] == bovad_lines['team'][j] or working_spreads_data['team'][i] == bovad_lines['away_team'][j]:\n",
        "          working_spreads_data['O/U_Line'][i] = bovad_lines['total1'][j]\n",
        "          break\n",
        "  else:\n",
        "    for j in range(len(game_lines_2023_data.index)):\n",
        "      if working_spreads_data['date'][i] == game_lines_2023_data['Date'][j]:\n",
        "        if working_spreads_data['team'][i] == game_lines_2023_data['Home'][j] or working_spreads_data['team'][i] == game_lines_2023_data['Away'][j]:\n",
        "          working_spreads_data['O/U_Line'][i] = game_lines_2023_data['Over/Under'][j]\n",
        "          break\n",
        "\n",
        "working_spreads_data"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 840
        },
        "id": "5_UWK5iBuu38",
        "outputId": "9c058d3d-af64-47c4-80c1-3410b7c98e56"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-12-8020987f8fa8>:10: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  working_spreads_data['O/U_Line'][i] = bovad_lines['total1'][j]\n",
            "<ipython-input-12-8020987f8fa8>:16: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  working_spreads_data['O/U_Line'][i] = game_lines_2023_data['Over/Under'][j]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Unnamed: 0  index  home       fg_r  fga_r       3p_r  3pa_r  3p_max_r  \\\n",
              "0               0      0     1  40.000000   85.0   7.000000   22.0       4.0   \n",
              "1               1      1     0  40.500000   84.0   7.500000   23.5       3.0   \n",
              "2               2      2     1  39.666667   85.0   7.333333   23.0       3.0   \n",
              "3               3      3     1  38.250000   84.0   7.750000   23.5       3.0   \n",
              "4               4      4     0  36.600000   81.6   8.000000   23.8       3.0   \n",
              "...           ...    ...   ...        ...    ...        ...    ...       ...   \n",
              "19010       19010  25653     0  43.200000   85.4  12.000000   28.8       4.2   \n",
              "19011       19011  25654     0  45.600000   88.8  11.400000   29.2       3.6   \n",
              "19012       19012  25655     0  46.400000   90.2  12.200000   31.0       4.0   \n",
              "19013       19013  25656     1  44.000000   85.6  11.400000   29.4       3.6   \n",
              "19014       19014  25657     0  42.600000   86.4   9.600000   27.4       3.2   \n",
              "\n",
              "            ft_r      fta_r  ...  tov_opp_r  team  team_opp        date  \\\n",
              "0      15.000000  17.000000  ...  21.000000   ATL       HOU  2012-11-02   \n",
              "1      14.500000  18.500000  ...  20.500000   ATL       OKC  2012-11-04   \n",
              "2      11.666667  16.333333  ...  18.666667   ATL       IND  2012-11-07   \n",
              "3      11.750000  16.250000  ...  17.500000   ATL       MIA  2012-11-09   \n",
              "4      10.800000  16.200000  ...  16.600000   ATL       LAC  2012-11-11   \n",
              "...          ...        ...  ...        ...   ...       ...         ...   \n",
              "19010  17.800000  21.400000  ...   9.800000   WAS       BOS  2022-04-03   \n",
              "19011  15.400000  19.800000  ...   9.600000   WAS       MIN  2022-04-05   \n",
              "19012  14.800000  18.600000  ...   9.200000   WAS       ATL  2022-04-06   \n",
              "19013  13.400000  16.800000  ...   9.000000   WAS       NYK  2022-04-08   \n",
              "19014  12.600000  17.600000  ...   9.800000   WAS       CHO  2022-04-10   \n",
              "\n",
              "       PointDifference  Spread_Result  Line  O/U_Result  Covered  O/U_Line  \n",
              "0                    7            NaN   5.0         NaN        0       203  \n",
              "1                   -9            NaN   9.0         NaN        1       199  \n",
              "2                   -3            NaN   4.0         NaN        1       192  \n",
              "3                    6            NaN  -5.0         NaN        0       198  \n",
              "4                   13            NaN   6.5         NaN        0       195  \n",
              "...                ...            ...   ...         ...      ...       ...  \n",
              "19010               42            NaN  13.0         NaN        0     222.5  \n",
              "19011              -18            NaN  12.0         NaN        1     238.5  \n",
              "19012               15            NaN  10.5         NaN        0     235.0  \n",
              "19013               22            NaN   3.5         NaN        0     224.0  \n",
              "19014               16            NaN  13.0         NaN        0     233.0  \n",
              "\n",
              "[19015 rows x 38 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-f904e33a-3802-4b15-aa69-e6078db100f8\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>index</th>\n",
              "      <th>home</th>\n",
              "      <th>fg_r</th>\n",
              "      <th>fga_r</th>\n",
              "      <th>3p_r</th>\n",
              "      <th>3pa_r</th>\n",
              "      <th>3p_max_r</th>\n",
              "      <th>ft_r</th>\n",
              "      <th>fta_r</th>\n",
              "      <th>...</th>\n",
              "      <th>tov_opp_r</th>\n",
              "      <th>team</th>\n",
              "      <th>team_opp</th>\n",
              "      <th>date</th>\n",
              "      <th>PointDifference</th>\n",
              "      <th>Spread_Result</th>\n",
              "      <th>Line</th>\n",
              "      <th>O/U_Result</th>\n",
              "      <th>Covered</th>\n",
              "      <th>O/U_Line</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>40.000000</td>\n",
              "      <td>85.0</td>\n",
              "      <td>7.000000</td>\n",
              "      <td>22.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>15.000000</td>\n",
              "      <td>17.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>21.000000</td>\n",
              "      <td>ATL</td>\n",
              "      <td>HOU</td>\n",
              "      <td>2012-11-02</td>\n",
              "      <td>7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>203</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>40.500000</td>\n",
              "      <td>84.0</td>\n",
              "      <td>7.500000</td>\n",
              "      <td>23.5</td>\n",
              "      <td>3.0</td>\n",
              "      <td>14.500000</td>\n",
              "      <td>18.500000</td>\n",
              "      <td>...</td>\n",
              "      <td>20.500000</td>\n",
              "      <td>ATL</td>\n",
              "      <td>OKC</td>\n",
              "      <td>2012-11-04</td>\n",
              "      <td>-9</td>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1</td>\n",
              "      <td>199</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>39.666667</td>\n",
              "      <td>85.0</td>\n",
              "      <td>7.333333</td>\n",
              "      <td>23.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>11.666667</td>\n",
              "      <td>16.333333</td>\n",
              "      <td>...</td>\n",
              "      <td>18.666667</td>\n",
              "      <td>ATL</td>\n",
              "      <td>IND</td>\n",
              "      <td>2012-11-07</td>\n",
              "      <td>-3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1</td>\n",
              "      <td>192</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>38.250000</td>\n",
              "      <td>84.0</td>\n",
              "      <td>7.750000</td>\n",
              "      <td>23.5</td>\n",
              "      <td>3.0</td>\n",
              "      <td>11.750000</td>\n",
              "      <td>16.250000</td>\n",
              "      <td>...</td>\n",
              "      <td>17.500000</td>\n",
              "      <td>ATL</td>\n",
              "      <td>MIA</td>\n",
              "      <td>2012-11-09</td>\n",
              "      <td>6</td>\n",
              "      <td>NaN</td>\n",
              "      <td>-5.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>198</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>36.600000</td>\n",
              "      <td>81.6</td>\n",
              "      <td>8.000000</td>\n",
              "      <td>23.8</td>\n",
              "      <td>3.0</td>\n",
              "      <td>10.800000</td>\n",
              "      <td>16.200000</td>\n",
              "      <td>...</td>\n",
              "      <td>16.600000</td>\n",
              "      <td>ATL</td>\n",
              "      <td>LAC</td>\n",
              "      <td>2012-11-11</td>\n",
              "      <td>13</td>\n",
              "      <td>NaN</td>\n",
              "      <td>6.5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>195</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19010</th>\n",
              "      <td>19010</td>\n",
              "      <td>25653</td>\n",
              "      <td>0</td>\n",
              "      <td>43.200000</td>\n",
              "      <td>85.4</td>\n",
              "      <td>12.000000</td>\n",
              "      <td>28.8</td>\n",
              "      <td>4.2</td>\n",
              "      <td>17.800000</td>\n",
              "      <td>21.400000</td>\n",
              "      <td>...</td>\n",
              "      <td>9.800000</td>\n",
              "      <td>WAS</td>\n",
              "      <td>BOS</td>\n",
              "      <td>2022-04-03</td>\n",
              "      <td>42</td>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>222.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19011</th>\n",
              "      <td>19011</td>\n",
              "      <td>25654</td>\n",
              "      <td>0</td>\n",
              "      <td>45.600000</td>\n",
              "      <td>88.8</td>\n",
              "      <td>11.400000</td>\n",
              "      <td>29.2</td>\n",
              "      <td>3.6</td>\n",
              "      <td>15.400000</td>\n",
              "      <td>19.800000</td>\n",
              "      <td>...</td>\n",
              "      <td>9.600000</td>\n",
              "      <td>WAS</td>\n",
              "      <td>MIN</td>\n",
              "      <td>2022-04-05</td>\n",
              "      <td>-18</td>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1</td>\n",
              "      <td>238.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19012</th>\n",
              "      <td>19012</td>\n",
              "      <td>25655</td>\n",
              "      <td>0</td>\n",
              "      <td>46.400000</td>\n",
              "      <td>90.2</td>\n",
              "      <td>12.200000</td>\n",
              "      <td>31.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>14.800000</td>\n",
              "      <td>18.600000</td>\n",
              "      <td>...</td>\n",
              "      <td>9.200000</td>\n",
              "      <td>WAS</td>\n",
              "      <td>ATL</td>\n",
              "      <td>2022-04-06</td>\n",
              "      <td>15</td>\n",
              "      <td>NaN</td>\n",
              "      <td>10.5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>235.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19013</th>\n",
              "      <td>19013</td>\n",
              "      <td>25656</td>\n",
              "      <td>1</td>\n",
              "      <td>44.000000</td>\n",
              "      <td>85.6</td>\n",
              "      <td>11.400000</td>\n",
              "      <td>29.4</td>\n",
              "      <td>3.6</td>\n",
              "      <td>13.400000</td>\n",
              "      <td>16.800000</td>\n",
              "      <td>...</td>\n",
              "      <td>9.000000</td>\n",
              "      <td>WAS</td>\n",
              "      <td>NYK</td>\n",
              "      <td>2022-04-08</td>\n",
              "      <td>22</td>\n",
              "      <td>NaN</td>\n",
              "      <td>3.5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>224.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19014</th>\n",
              "      <td>19014</td>\n",
              "      <td>25657</td>\n",
              "      <td>0</td>\n",
              "      <td>42.600000</td>\n",
              "      <td>86.4</td>\n",
              "      <td>9.600000</td>\n",
              "      <td>27.4</td>\n",
              "      <td>3.2</td>\n",
              "      <td>12.600000</td>\n",
              "      <td>17.600000</td>\n",
              "      <td>...</td>\n",
              "      <td>9.800000</td>\n",
              "      <td>WAS</td>\n",
              "      <td>CHO</td>\n",
              "      <td>2022-04-10</td>\n",
              "      <td>16</td>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>233.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>19015 rows × 38 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f904e33a-3802-4b15-aa69-e6078db100f8')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-f904e33a-3802-4b15-aa69-e6078db100f8 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-f904e33a-3802-4b15-aa69-e6078db100f8');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "working_spreads_data['Covered_O/U'] = 0\n",
        "\n",
        "for i in range(len(working_spreads_data.index)):\n",
        "  for j in range(len(dF2.index)):\n",
        "    if working_spreads_data['date'][i] == dF2['date'][j]:\n",
        "      if working_spreads_data['team'][i] == dF2['team'][j] or working_spreads_data['team'][i] == dF2['team_opp'][j]:\n",
        "        working_spreads_data['O/U_Result'][i] = dF2['point_total'][j]\n",
        "        if float(working_spreads_data['O/U_Result'][i]) > float(working_spreads_data['O/U_Line'][i]):\n",
        "          working_spreads_data['Covered_O/U'][i] = 1\n",
        "        elif float(working_spreads_data['O/U_Result'][i]) < float(working_spreads_data['O/U_Line'][i]):\n",
        "          working_spreads_data['Covered_O/U'][i] = -1\n",
        "        break\n",
        "\n",
        "working_spreads_data"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 925
        },
        "id": "xMl4__en5b9t",
        "outputId": "f9c71c20-275d-472f-bec4-973ceef02a01"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-15-47948f937c7b>:7: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  working_spreads_data['O/U_Result'][i] = dF2['point_total'][j]\n",
            "<ipython-input-15-47948f937c7b>:9: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  working_spreads_data['Covered_O/U'][i] = 1\n",
            "<ipython-input-15-47948f937c7b>:11: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  working_spreads_data['Covered_O/U'][i] = -1\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Unnamed: 0  index  home       fg_r  fga_r       3p_r  3pa_r  3p_max_r  \\\n",
              "0               0      0     1  40.000000   85.0   7.000000   22.0       4.0   \n",
              "1               1      1     0  40.500000   84.0   7.500000   23.5       3.0   \n",
              "2               2      2     1  39.666667   85.0   7.333333   23.0       3.0   \n",
              "3               3      3     1  38.250000   84.0   7.750000   23.5       3.0   \n",
              "4               4      4     0  36.600000   81.6   8.000000   23.8       3.0   \n",
              "...           ...    ...   ...        ...    ...        ...    ...       ...   \n",
              "19010       19010  25653     0  43.200000   85.4  12.000000   28.8       4.2   \n",
              "19011       19011  25654     0  45.600000   88.8  11.400000   29.2       3.6   \n",
              "19012       19012  25655     0  46.400000   90.2  12.200000   31.0       4.0   \n",
              "19013       19013  25656     1  44.000000   85.6  11.400000   29.4       3.6   \n",
              "19014       19014  25657     0  42.600000   86.4   9.600000   27.4       3.2   \n",
              "\n",
              "            ft_r      fta_r  ...  team  team_opp        date  PointDifference  \\\n",
              "0      15.000000  17.000000  ...   ATL       HOU  2012-11-02                7   \n",
              "1      14.500000  18.500000  ...   ATL       OKC  2012-11-04               -9   \n",
              "2      11.666667  16.333333  ...   ATL       IND  2012-11-07               -3   \n",
              "3      11.750000  16.250000  ...   ATL       MIA  2012-11-09                6   \n",
              "4      10.800000  16.200000  ...   ATL       LAC  2012-11-11               13   \n",
              "...          ...        ...  ...   ...       ...         ...              ...   \n",
              "19010  17.800000  21.400000  ...   WAS       BOS  2022-04-03               42   \n",
              "19011  15.400000  19.800000  ...   WAS       MIN  2022-04-05              -18   \n",
              "19012  14.800000  18.600000  ...   WAS       ATL  2022-04-06               15   \n",
              "19013  13.400000  16.800000  ...   WAS       NYK  2022-04-08               22   \n",
              "19014  12.600000  17.600000  ...   WAS       CHO  2022-04-10               16   \n",
              "\n",
              "       Spread_Result  Line  O/U_Result  Covered  O/U_Line  Covered_O/U  \n",
              "0                NaN   5.0       211.0        0       203            1  \n",
              "1                NaN   9.0       199.0        1       199            0  \n",
              "2                NaN   4.0       175.0        1       192           -1  \n",
              "3                NaN  -5.0       184.0        0       198           -1  \n",
              "4                NaN   6.5       165.0        0       195           -1  \n",
              "...              ...   ...         ...      ...       ...          ...  \n",
              "19010            NaN  13.0       246.0        0     222.5            1  \n",
              "19011            NaN  12.0       246.0        1     238.5            1  \n",
              "19012            NaN  10.5       221.0        0     235.0           -1  \n",
              "19013            NaN   3.5       206.0        0     224.0           -1  \n",
              "19014            NaN  13.0       232.0        0     233.0           -1  \n",
              "\n",
              "[19015 rows x 39 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-3a157be9-8571-498b-b047-73ef68b9166b\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>index</th>\n",
              "      <th>home</th>\n",
              "      <th>fg_r</th>\n",
              "      <th>fga_r</th>\n",
              "      <th>3p_r</th>\n",
              "      <th>3pa_r</th>\n",
              "      <th>3p_max_r</th>\n",
              "      <th>ft_r</th>\n",
              "      <th>fta_r</th>\n",
              "      <th>...</th>\n",
              "      <th>team</th>\n",
              "      <th>team_opp</th>\n",
              "      <th>date</th>\n",
              "      <th>PointDifference</th>\n",
              "      <th>Spread_Result</th>\n",
              "      <th>Line</th>\n",
              "      <th>O/U_Result</th>\n",
              "      <th>Covered</th>\n",
              "      <th>O/U_Line</th>\n",
              "      <th>Covered_O/U</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>40.000000</td>\n",
              "      <td>85.0</td>\n",
              "      <td>7.000000</td>\n",
              "      <td>22.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>15.000000</td>\n",
              "      <td>17.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>ATL</td>\n",
              "      <td>HOU</td>\n",
              "      <td>2012-11-02</td>\n",
              "      <td>7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>211.0</td>\n",
              "      <td>0</td>\n",
              "      <td>203</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>40.500000</td>\n",
              "      <td>84.0</td>\n",
              "      <td>7.500000</td>\n",
              "      <td>23.5</td>\n",
              "      <td>3.0</td>\n",
              "      <td>14.500000</td>\n",
              "      <td>18.500000</td>\n",
              "      <td>...</td>\n",
              "      <td>ATL</td>\n",
              "      <td>OKC</td>\n",
              "      <td>2012-11-04</td>\n",
              "      <td>-9</td>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>199.0</td>\n",
              "      <td>1</td>\n",
              "      <td>199</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>39.666667</td>\n",
              "      <td>85.0</td>\n",
              "      <td>7.333333</td>\n",
              "      <td>23.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>11.666667</td>\n",
              "      <td>16.333333</td>\n",
              "      <td>...</td>\n",
              "      <td>ATL</td>\n",
              "      <td>IND</td>\n",
              "      <td>2012-11-07</td>\n",
              "      <td>-3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>175.0</td>\n",
              "      <td>1</td>\n",
              "      <td>192</td>\n",
              "      <td>-1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>38.250000</td>\n",
              "      <td>84.0</td>\n",
              "      <td>7.750000</td>\n",
              "      <td>23.5</td>\n",
              "      <td>3.0</td>\n",
              "      <td>11.750000</td>\n",
              "      <td>16.250000</td>\n",
              "      <td>...</td>\n",
              "      <td>ATL</td>\n",
              "      <td>MIA</td>\n",
              "      <td>2012-11-09</td>\n",
              "      <td>6</td>\n",
              "      <td>NaN</td>\n",
              "      <td>-5.0</td>\n",
              "      <td>184.0</td>\n",
              "      <td>0</td>\n",
              "      <td>198</td>\n",
              "      <td>-1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>36.600000</td>\n",
              "      <td>81.6</td>\n",
              "      <td>8.000000</td>\n",
              "      <td>23.8</td>\n",
              "      <td>3.0</td>\n",
              "      <td>10.800000</td>\n",
              "      <td>16.200000</td>\n",
              "      <td>...</td>\n",
              "      <td>ATL</td>\n",
              "      <td>LAC</td>\n",
              "      <td>2012-11-11</td>\n",
              "      <td>13</td>\n",
              "      <td>NaN</td>\n",
              "      <td>6.5</td>\n",
              "      <td>165.0</td>\n",
              "      <td>0</td>\n",
              "      <td>195</td>\n",
              "      <td>-1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19010</th>\n",
              "      <td>19010</td>\n",
              "      <td>25653</td>\n",
              "      <td>0</td>\n",
              "      <td>43.200000</td>\n",
              "      <td>85.4</td>\n",
              "      <td>12.000000</td>\n",
              "      <td>28.8</td>\n",
              "      <td>4.2</td>\n",
              "      <td>17.800000</td>\n",
              "      <td>21.400000</td>\n",
              "      <td>...</td>\n",
              "      <td>WAS</td>\n",
              "      <td>BOS</td>\n",
              "      <td>2022-04-03</td>\n",
              "      <td>42</td>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>246.0</td>\n",
              "      <td>0</td>\n",
              "      <td>222.5</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19011</th>\n",
              "      <td>19011</td>\n",
              "      <td>25654</td>\n",
              "      <td>0</td>\n",
              "      <td>45.600000</td>\n",
              "      <td>88.8</td>\n",
              "      <td>11.400000</td>\n",
              "      <td>29.2</td>\n",
              "      <td>3.6</td>\n",
              "      <td>15.400000</td>\n",
              "      <td>19.800000</td>\n",
              "      <td>...</td>\n",
              "      <td>WAS</td>\n",
              "      <td>MIN</td>\n",
              "      <td>2022-04-05</td>\n",
              "      <td>-18</td>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>246.0</td>\n",
              "      <td>1</td>\n",
              "      <td>238.5</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19012</th>\n",
              "      <td>19012</td>\n",
              "      <td>25655</td>\n",
              "      <td>0</td>\n",
              "      <td>46.400000</td>\n",
              "      <td>90.2</td>\n",
              "      <td>12.200000</td>\n",
              "      <td>31.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>14.800000</td>\n",
              "      <td>18.600000</td>\n",
              "      <td>...</td>\n",
              "      <td>WAS</td>\n",
              "      <td>ATL</td>\n",
              "      <td>2022-04-06</td>\n",
              "      <td>15</td>\n",
              "      <td>NaN</td>\n",
              "      <td>10.5</td>\n",
              "      <td>221.0</td>\n",
              "      <td>0</td>\n",
              "      <td>235.0</td>\n",
              "      <td>-1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19013</th>\n",
              "      <td>19013</td>\n",
              "      <td>25656</td>\n",
              "      <td>1</td>\n",
              "      <td>44.000000</td>\n",
              "      <td>85.6</td>\n",
              "      <td>11.400000</td>\n",
              "      <td>29.4</td>\n",
              "      <td>3.6</td>\n",
              "      <td>13.400000</td>\n",
              "      <td>16.800000</td>\n",
              "      <td>...</td>\n",
              "      <td>WAS</td>\n",
              "      <td>NYK</td>\n",
              "      <td>2022-04-08</td>\n",
              "      <td>22</td>\n",
              "      <td>NaN</td>\n",
              "      <td>3.5</td>\n",
              "      <td>206.0</td>\n",
              "      <td>0</td>\n",
              "      <td>224.0</td>\n",
              "      <td>-1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19014</th>\n",
              "      <td>19014</td>\n",
              "      <td>25657</td>\n",
              "      <td>0</td>\n",
              "      <td>42.600000</td>\n",
              "      <td>86.4</td>\n",
              "      <td>9.600000</td>\n",
              "      <td>27.4</td>\n",
              "      <td>3.2</td>\n",
              "      <td>12.600000</td>\n",
              "      <td>17.600000</td>\n",
              "      <td>...</td>\n",
              "      <td>WAS</td>\n",
              "      <td>CHO</td>\n",
              "      <td>2022-04-10</td>\n",
              "      <td>16</td>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>232.0</td>\n",
              "      <td>0</td>\n",
              "      <td>233.0</td>\n",
              "      <td>-1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>19015 rows × 39 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-3a157be9-8571-498b-b047-73ef68b9166b')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-3a157be9-8571-498b-b047-73ef68b9166b button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-3a157be9-8571-498b-b047-73ef68b9166b');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "working_spreads_data = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/working_spreads_all.csv')\n",
        "\n",
        "for i in range(len(working_spreads_data.index)):\n",
        "  if working_spreads_data['Covered_O/U'][i] != 0 and working_spreads_data['Covered_O/U'][i] != 1:\n",
        "    print('gone')\n",
        "\n",
        "# working_spreads_data.to_csv('working_spreads_all.csv')"
      ],
      "metadata": {
        "id": "bMOAFmpPFMR8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "season_2023 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2023Season.csv')\n",
        "scores_2023 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2021_2023_Box_Scores.csv', on_bad_lines='skip')\n",
        "odds_2023 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/2023_Season_Odds.csv')\n",
        "season_2023['points'] = 0\n",
        "season_2023['points_opp'] = 0\n",
        "season_2023['O/U_Result'] = 0\n",
        "season_2023['O/U_Line'] = 0\n",
        "\n",
        "for i in range(len(season_2023.index)):\n",
        "  found_1 = False\n",
        "  found_2 = False\n",
        "  month, day, year = season_2023['date'][i].split('/')\n",
        "  date = '{}-{}-{}'.format(year, month, day)\n",
        "\n",
        "  for j in range(len(scores_2023.index)):\n",
        "    if date == scores_2023['Date'][j]:\n",
        "      if season_2023['team'][i] == scores_2023['Opponent'][j]:\n",
        "        season_2023['points_opp'][i] = scores_2023['Points'][j]\n",
        "        found_1 = True\n",
        "      elif season_2023['team_opp'][i] == scores_2023['Opponent'][j]:\n",
        "        season_2023['points'][i] = scores_2023['Points'][j]\n",
        "        found_2 = True\n",
        "\n",
        "      if found_1 and found_2:\n",
        "        season_2023['O/U_Result'][i] = season_2023['points'][i] + season_2023['points_opp'][i]\n",
        "        break\n",
        "\n",
        "  for j in range(len(odds_2023.index)):\n",
        "    if date == odds_2023['Date'][j]:\n",
        "      if season_2023['team'][i] == odds_2023['Away'][j] or season_2023['team'][i] == odds_2023['Home'][j]:\n",
        "        season_2023['O/U_Line'][i] = odds_2023['Over/Under'][j]\n",
        "\n",
        "season_2023.to_csv('2023_Season_All.csv')\n",
        "season_2023"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 840
        },
        "id": "1W3-xJmD9EYS",
        "outputId": "303d60a3-98f0-4055-f892-0e50322f0206"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-46-8bfe777d7ec0>:18: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  season_2023['points_opp'][i] = scores_2023['Points'][j]\n",
            "<ipython-input-46-8bfe777d7ec0>:21: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  season_2023['points'][i] = scores_2023['Points'][j]\n",
            "<ipython-input-46-8bfe777d7ec0>:25: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  season_2023['O/U_Result'][i] = season_2023['points'][i] + season_2023['points_opp'][i]\n",
            "<ipython-input-46-8bfe777d7ec0>:31: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  season_2023['O/U_Line'][i] = odds_2023['Over/Under'][j]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Unnamed: 0.1  index  Unnamed: 0  home  fg_r  fga_r  3p_r  3pa_r  \\\n",
              "0                0      0        1744     0  40.0   80.0  13.0   34.0   \n",
              "1                1      1          79     1  46.0   82.0  12.0   35.0   \n",
              "2                2      2        1031     0  40.0   94.0  10.0   40.0   \n",
              "3                3      3         713     1  45.0   99.0  16.0   45.0   \n",
              "4                4      4        2218     1  42.0   83.0  16.0   38.0   \n",
              "...            ...    ...         ...   ...   ...    ...   ...    ...   \n",
              "2369          2369   2369         236     1  41.4   87.4  12.4   34.0   \n",
              "2370          2370   2370        1426     0  40.8   87.0  10.6   29.6   \n",
              "2371          2371   2371        1346     0  47.4   92.2  14.2   36.6   \n",
              "2372          2372   2372         872     1  43.4   94.4  10.4   30.4   \n",
              "2373          2373   2373        2375     1  44.4   93.6  15.0   37.4   \n",
              "\n",
              "      3p_max_r  ft_r  ...  team  team_opp        date  PointDifference  Line  \\\n",
              "0          5.0  24.0  ...   PHI       BOS  10/18/2022                9   3.0   \n",
              "1          4.0  22.0  ...   BOS       PHI  10/18/2022               -9  -3.0   \n",
              "2          3.0  19.0  ...   LAL       GSW  10/18/2022               14   7.5   \n",
              "3          4.0  17.0  ...   GSW       LAL  10/18/2022              -14  -7.5   \n",
              "4          4.0  23.0  ...   UTA       DEN  10/19/2022              -21   7.0   \n",
              "...        ...   ...  ...   ...       ...         ...              ...   ...   \n",
              "2369       4.4  18.0  ...   BKN       MIN    4/4/2023                5   3.0   \n",
              "2370       3.4  16.2  ...   MIN       BKN    4/4/2023               -5  -3.0   \n",
              "2371       4.0  17.2  ...   MIL       WAS    4/4/2023              -12 -14.0   \n",
              "2372       5.0  19.4  ...   HOU       DEN    4/4/2023              -21  11.5   \n",
              "2373       5.2  12.2  ...   WAS       MIL    4/4/2023               12  14.0   \n",
              "\n",
              "      Covered  points  points_opp  O/U_Result  O/U_Line  \n",
              "0           0     117         126         243     216.0  \n",
              "1           1     126         117         243     216.0  \n",
              "2           0     109         123         232     223.5  \n",
              "3           1     123         109         232     223.5  \n",
              "4           1     123         102         225     225.0  \n",
              "...       ...     ...         ...         ...       ...  \n",
              "2369        0       0           0           0       0.0  \n",
              "2370        1       0           0           0       0.0  \n",
              "2371        0       0           0           0       0.0  \n",
              "2372        1       0           0           0       0.0  \n",
              "2373        1       0           0           0       0.0  \n",
              "\n",
              "[2374 rows x 40 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-18df67dc-d8cf-44fd-a1f1-6fd8d1e901d4\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0.1</th>\n",
              "      <th>index</th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>home</th>\n",
              "      <th>fg_r</th>\n",
              "      <th>fga_r</th>\n",
              "      <th>3p_r</th>\n",
              "      <th>3pa_r</th>\n",
              "      <th>3p_max_r</th>\n",
              "      <th>ft_r</th>\n",
              "      <th>...</th>\n",
              "      <th>team</th>\n",
              "      <th>team_opp</th>\n",
              "      <th>date</th>\n",
              "      <th>PointDifference</th>\n",
              "      <th>Line</th>\n",
              "      <th>Covered</th>\n",
              "      <th>points</th>\n",
              "      <th>points_opp</th>\n",
              "      <th>O/U_Result</th>\n",
              "      <th>O/U_Line</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1744</td>\n",
              "      <td>0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>80.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>24.0</td>\n",
              "      <td>...</td>\n",
              "      <td>PHI</td>\n",
              "      <td>BOS</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>9</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0</td>\n",
              "      <td>117</td>\n",
              "      <td>126</td>\n",
              "      <td>243</td>\n",
              "      <td>216.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79</td>\n",
              "      <td>1</td>\n",
              "      <td>46.0</td>\n",
              "      <td>82.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>...</td>\n",
              "      <td>BOS</td>\n",
              "      <td>PHI</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>-9</td>\n",
              "      <td>-3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>126</td>\n",
              "      <td>117</td>\n",
              "      <td>243</td>\n",
              "      <td>216.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1031</td>\n",
              "      <td>0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>94.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>...</td>\n",
              "      <td>LAL</td>\n",
              "      <td>GSW</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>14</td>\n",
              "      <td>7.5</td>\n",
              "      <td>0</td>\n",
              "      <td>109</td>\n",
              "      <td>123</td>\n",
              "      <td>232</td>\n",
              "      <td>223.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>713</td>\n",
              "      <td>1</td>\n",
              "      <td>45.0</td>\n",
              "      <td>99.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>45.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>17.0</td>\n",
              "      <td>...</td>\n",
              "      <td>GSW</td>\n",
              "      <td>LAL</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>-14</td>\n",
              "      <td>-7.5</td>\n",
              "      <td>1</td>\n",
              "      <td>123</td>\n",
              "      <td>109</td>\n",
              "      <td>232</td>\n",
              "      <td>223.5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>2218</td>\n",
              "      <td>1</td>\n",
              "      <td>42.0</td>\n",
              "      <td>83.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>...</td>\n",
              "      <td>UTA</td>\n",
              "      <td>DEN</td>\n",
              "      <td>10/19/2022</td>\n",
              "      <td>-21</td>\n",
              "      <td>7.0</td>\n",
              "      <td>1</td>\n",
              "      <td>123</td>\n",
              "      <td>102</td>\n",
              "      <td>225</td>\n",
              "      <td>225.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2369</th>\n",
              "      <td>2369</td>\n",
              "      <td>2369</td>\n",
              "      <td>236</td>\n",
              "      <td>1</td>\n",
              "      <td>41.4</td>\n",
              "      <td>87.4</td>\n",
              "      <td>12.4</td>\n",
              "      <td>34.0</td>\n",
              "      <td>4.4</td>\n",
              "      <td>18.0</td>\n",
              "      <td>...</td>\n",
              "      <td>BKN</td>\n",
              "      <td>MIN</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>5</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2370</th>\n",
              "      <td>2370</td>\n",
              "      <td>2370</td>\n",
              "      <td>1426</td>\n",
              "      <td>0</td>\n",
              "      <td>40.8</td>\n",
              "      <td>87.0</td>\n",
              "      <td>10.6</td>\n",
              "      <td>29.6</td>\n",
              "      <td>3.4</td>\n",
              "      <td>16.2</td>\n",
              "      <td>...</td>\n",
              "      <td>MIN</td>\n",
              "      <td>BKN</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>-5</td>\n",
              "      <td>-3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2371</th>\n",
              "      <td>2371</td>\n",
              "      <td>2371</td>\n",
              "      <td>1346</td>\n",
              "      <td>0</td>\n",
              "      <td>47.4</td>\n",
              "      <td>92.2</td>\n",
              "      <td>14.2</td>\n",
              "      <td>36.6</td>\n",
              "      <td>4.0</td>\n",
              "      <td>17.2</td>\n",
              "      <td>...</td>\n",
              "      <td>MIL</td>\n",
              "      <td>WAS</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>-12</td>\n",
              "      <td>-14.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2372</th>\n",
              "      <td>2372</td>\n",
              "      <td>2372</td>\n",
              "      <td>872</td>\n",
              "      <td>1</td>\n",
              "      <td>43.4</td>\n",
              "      <td>94.4</td>\n",
              "      <td>10.4</td>\n",
              "      <td>30.4</td>\n",
              "      <td>5.0</td>\n",
              "      <td>19.4</td>\n",
              "      <td>...</td>\n",
              "      <td>HOU</td>\n",
              "      <td>DEN</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>-21</td>\n",
              "      <td>11.5</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2373</th>\n",
              "      <td>2373</td>\n",
              "      <td>2373</td>\n",
              "      <td>2375</td>\n",
              "      <td>1</td>\n",
              "      <td>44.4</td>\n",
              "      <td>93.6</td>\n",
              "      <td>15.0</td>\n",
              "      <td>37.4</td>\n",
              "      <td>5.2</td>\n",
              "      <td>12.2</td>\n",
              "      <td>...</td>\n",
              "      <td>WAS</td>\n",
              "      <td>MIL</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>12</td>\n",
              "      <td>14.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2374 rows × 40 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-18df67dc-d8cf-44fd-a1f1-6fd8d1e901d4')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-18df67dc-d8cf-44fd-a1f1-6fd8d1e901d4 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-18df67dc-d8cf-44fd-a1f1-6fd8d1e901d4');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "season_2023['Covered_O/U'] = 0\n",
        "\n",
        "for i in range(len(season_2023.index)):\n",
        "  if float(season_2023['O/U_Result'][i]) >= float(season_2023['O/U_Line'][i]):\n",
        "    season_2023['Covered_O/U'][i] = 1\n",
        "  else:\n",
        "    season_2023['Covered_O/U'][i] = 0\n",
        "\n",
        "season_2023.to_csv('2023_Season_All.csv')\n",
        "season_2023"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 670
        },
        "id": "gdQ0AOfoEmKv",
        "outputId": "dad1b072-af61-4690-f05e-2604d7b23bfe"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-47-9156ca9557b4>:5: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  season_2023['Covered_O/U'][i] = 1\n",
            "<ipython-input-47-9156ca9557b4>:7: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  season_2023['Covered_O/U'][i] = 0\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Unnamed: 0.1  index  Unnamed: 0  home  fg_r  fga_r  3p_r  3pa_r  \\\n",
              "0                0      0        1744     0  40.0   80.0  13.0   34.0   \n",
              "1                1      1          79     1  46.0   82.0  12.0   35.0   \n",
              "2                2      2        1031     0  40.0   94.0  10.0   40.0   \n",
              "3                3      3         713     1  45.0   99.0  16.0   45.0   \n",
              "4                4      4        2218     1  42.0   83.0  16.0   38.0   \n",
              "...            ...    ...         ...   ...   ...    ...   ...    ...   \n",
              "2369          2369   2369         236     1  41.4   87.4  12.4   34.0   \n",
              "2370          2370   2370        1426     0  40.8   87.0  10.6   29.6   \n",
              "2371          2371   2371        1346     0  47.4   92.2  14.2   36.6   \n",
              "2372          2372   2372         872     1  43.4   94.4  10.4   30.4   \n",
              "2373          2373   2373        2375     1  44.4   93.6  15.0   37.4   \n",
              "\n",
              "      3p_max_r  ft_r  ...  team_opp        date  PointDifference  Line  \\\n",
              "0          5.0  24.0  ...       BOS  10/18/2022                9   3.0   \n",
              "1          4.0  22.0  ...       PHI  10/18/2022               -9  -3.0   \n",
              "2          3.0  19.0  ...       GSW  10/18/2022               14   7.5   \n",
              "3          4.0  17.0  ...       LAL  10/18/2022              -14  -7.5   \n",
              "4          4.0  23.0  ...       DEN  10/19/2022              -21   7.0   \n",
              "...        ...   ...  ...       ...         ...              ...   ...   \n",
              "2369       4.4  18.0  ...       MIN    4/4/2023                5   3.0   \n",
              "2370       3.4  16.2  ...       BKN    4/4/2023               -5  -3.0   \n",
              "2371       4.0  17.2  ...       WAS    4/4/2023              -12 -14.0   \n",
              "2372       5.0  19.4  ...       DEN    4/4/2023              -21  11.5   \n",
              "2373       5.2  12.2  ...       MIL    4/4/2023               12  14.0   \n",
              "\n",
              "      Covered  points  points_opp  O/U_Result  O/U_Line  Covered_O/U  \n",
              "0           0     117         126         243     216.0            1  \n",
              "1           1     126         117         243     216.0            1  \n",
              "2           0     109         123         232     223.5            1  \n",
              "3           1     123         109         232     223.5            1  \n",
              "4           1     123         102         225     225.0            1  \n",
              "...       ...     ...         ...         ...       ...          ...  \n",
              "2369        0       0           0           0       0.0            1  \n",
              "2370        1       0           0           0       0.0            1  \n",
              "2371        0       0           0           0       0.0            1  \n",
              "2372        1       0           0           0       0.0            1  \n",
              "2373        1       0           0           0       0.0            1  \n",
              "\n",
              "[2374 rows x 41 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-92d88ffd-de0e-41b5-a360-9fdbe68208b1\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0.1</th>\n",
              "      <th>index</th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>home</th>\n",
              "      <th>fg_r</th>\n",
              "      <th>fga_r</th>\n",
              "      <th>3p_r</th>\n",
              "      <th>3pa_r</th>\n",
              "      <th>3p_max_r</th>\n",
              "      <th>ft_r</th>\n",
              "      <th>...</th>\n",
              "      <th>team_opp</th>\n",
              "      <th>date</th>\n",
              "      <th>PointDifference</th>\n",
              "      <th>Line</th>\n",
              "      <th>Covered</th>\n",
              "      <th>points</th>\n",
              "      <th>points_opp</th>\n",
              "      <th>O/U_Result</th>\n",
              "      <th>O/U_Line</th>\n",
              "      <th>Covered_O/U</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1744</td>\n",
              "      <td>0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>80.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>24.0</td>\n",
              "      <td>...</td>\n",
              "      <td>BOS</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>9</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0</td>\n",
              "      <td>117</td>\n",
              "      <td>126</td>\n",
              "      <td>243</td>\n",
              "      <td>216.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>79</td>\n",
              "      <td>1</td>\n",
              "      <td>46.0</td>\n",
              "      <td>82.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>...</td>\n",
              "      <td>PHI</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>-9</td>\n",
              "      <td>-3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>126</td>\n",
              "      <td>117</td>\n",
              "      <td>243</td>\n",
              "      <td>216.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1031</td>\n",
              "      <td>0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>94.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>...</td>\n",
              "      <td>GSW</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>14</td>\n",
              "      <td>7.5</td>\n",
              "      <td>0</td>\n",
              "      <td>109</td>\n",
              "      <td>123</td>\n",
              "      <td>232</td>\n",
              "      <td>223.5</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>713</td>\n",
              "      <td>1</td>\n",
              "      <td>45.0</td>\n",
              "      <td>99.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>45.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>17.0</td>\n",
              "      <td>...</td>\n",
              "      <td>LAL</td>\n",
              "      <td>10/18/2022</td>\n",
              "      <td>-14</td>\n",
              "      <td>-7.5</td>\n",
              "      <td>1</td>\n",
              "      <td>123</td>\n",
              "      <td>109</td>\n",
              "      <td>232</td>\n",
              "      <td>223.5</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>2218</td>\n",
              "      <td>1</td>\n",
              "      <td>42.0</td>\n",
              "      <td>83.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>...</td>\n",
              "      <td>DEN</td>\n",
              "      <td>10/19/2022</td>\n",
              "      <td>-21</td>\n",
              "      <td>7.0</td>\n",
              "      <td>1</td>\n",
              "      <td>123</td>\n",
              "      <td>102</td>\n",
              "      <td>225</td>\n",
              "      <td>225.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2369</th>\n",
              "      <td>2369</td>\n",
              "      <td>2369</td>\n",
              "      <td>236</td>\n",
              "      <td>1</td>\n",
              "      <td>41.4</td>\n",
              "      <td>87.4</td>\n",
              "      <td>12.4</td>\n",
              "      <td>34.0</td>\n",
              "      <td>4.4</td>\n",
              "      <td>18.0</td>\n",
              "      <td>...</td>\n",
              "      <td>MIN</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>5</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2370</th>\n",
              "      <td>2370</td>\n",
              "      <td>2370</td>\n",
              "      <td>1426</td>\n",
              "      <td>0</td>\n",
              "      <td>40.8</td>\n",
              "      <td>87.0</td>\n",
              "      <td>10.6</td>\n",
              "      <td>29.6</td>\n",
              "      <td>3.4</td>\n",
              "      <td>16.2</td>\n",
              "      <td>...</td>\n",
              "      <td>BKN</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>-5</td>\n",
              "      <td>-3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2371</th>\n",
              "      <td>2371</td>\n",
              "      <td>2371</td>\n",
              "      <td>1346</td>\n",
              "      <td>0</td>\n",
              "      <td>47.4</td>\n",
              "      <td>92.2</td>\n",
              "      <td>14.2</td>\n",
              "      <td>36.6</td>\n",
              "      <td>4.0</td>\n",
              "      <td>17.2</td>\n",
              "      <td>...</td>\n",
              "      <td>WAS</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>-12</td>\n",
              "      <td>-14.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2372</th>\n",
              "      <td>2372</td>\n",
              "      <td>2372</td>\n",
              "      <td>872</td>\n",
              "      <td>1</td>\n",
              "      <td>43.4</td>\n",
              "      <td>94.4</td>\n",
              "      <td>10.4</td>\n",
              "      <td>30.4</td>\n",
              "      <td>5.0</td>\n",
              "      <td>19.4</td>\n",
              "      <td>...</td>\n",
              "      <td>DEN</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>-21</td>\n",
              "      <td>11.5</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2373</th>\n",
              "      <td>2373</td>\n",
              "      <td>2373</td>\n",
              "      <td>2375</td>\n",
              "      <td>1</td>\n",
              "      <td>44.4</td>\n",
              "      <td>93.6</td>\n",
              "      <td>15.0</td>\n",
              "      <td>37.4</td>\n",
              "      <td>5.2</td>\n",
              "      <td>12.2</td>\n",
              "      <td>...</td>\n",
              "      <td>MIL</td>\n",
              "      <td>4/4/2023</td>\n",
              "      <td>12</td>\n",
              "      <td>14.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2374 rows × 41 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-92d88ffd-de0e-41b5-a360-9fdbe68208b1')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-92d88ffd-de0e-41b5-a360-9fdbe68208b1 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-92d88ffd-de0e-41b5-a360-9fdbe68208b1');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Neural Network (O/U)"
      ],
      "metadata": {
        "id": "D1NjGUWpuQLd"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "#Coded by Davin Stephens\n",
        "df8 = pd.read_csv('https://raw.githubusercontent.com/austinfett/ML-Capstone/main/working_spreads_all.csv')\n",
        "#Covered is the target label to predict\n",
        "target2 = df8['Covered_O/U']\n",
        "#Features that will be predicted\n",
        "df8 = df8[['home', 'fg_r', 'fga_r', '3p_r', '3pa_r', '3p_max_r', 'ft_r', 'fta_r', 'orb_r', 'drb_r', 'ast_r', 'stl_r', 'blk_r', 'tov_r', 'fg_opp_r', 'fga_opp_r', '3p_opp_r', '3pa_opp_r', '3p_max_opp_r', 'ft_opp_r', 'fta_opp_r', 'orb_opp_r', 'drb_opp_r', 'ast_opp_r', 'stl_opp_r', 'blk_opp_r', 'tov_opp_r']]\n",
        "\n",
        "# normalize and split data into training and testing\n",
        "xx_train2 = tf.keras.utils.normalize(df8, axis=1)\n",
        "X_train2, X_test2, y_train2, y_test2 = train_test_split(xx_train2, target2, test_size=.2)\n",
        "\n",
        "# neural network model, not fully optimized yet \n",
        "model2 = tf.keras.models.Sequential()\n",
        "model2.add(tf.keras.layers.Dense(64, activation=tf.nn.relu6))\n",
        "model2.add(tf.keras.layers.Dense(32, activation=tf.nn.relu6))\n",
        "model2.add(tf.keras.layers.Dense(1, activation=tf.nn.sigmoid))\n",
        "model2.compile(optimizer='adam', loss='binary_crossentropy', metrics='accuracy')\n",
        "model2.fit(X_train2, y_train2, epochs=50, batch_size=32)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZhKs7RWRuYrH",
        "outputId": "36f4b103-3253-4592-d043-a2eb8ea7880b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "476/476 [==============================] - 2s 2ms/step - loss: 0.6875 - accuracy: 0.5543\n",
            "Epoch 2/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6863 - accuracy: 0.5560\n",
            "Epoch 3/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6855 - accuracy: 0.5571\n",
            "Epoch 4/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6833 - accuracy: 0.5606\n",
            "Epoch 5/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6774 - accuracy: 0.5726\n",
            "Epoch 6/50\n",
            "476/476 [==============================] - 2s 3ms/step - loss: 0.6731 - accuracy: 0.5835\n",
            "Epoch 7/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6694 - accuracy: 0.5866\n",
            "Epoch 8/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6691 - accuracy: 0.5884\n",
            "Epoch 9/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6690 - accuracy: 0.5843\n",
            "Epoch 10/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6683 - accuracy: 0.5884\n",
            "Epoch 11/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6667 - accuracy: 0.5943\n",
            "Epoch 12/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6666 - accuracy: 0.5945\n",
            "Epoch 13/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6660 - accuracy: 0.5906\n",
            "Epoch 14/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6651 - accuracy: 0.5937\n",
            "Epoch 15/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6652 - accuracy: 0.5937\n",
            "Epoch 16/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6663 - accuracy: 0.5923\n",
            "Epoch 17/50\n",
            "476/476 [==============================] - 1s 3ms/step - loss: 0.6648 - accuracy: 0.5919\n",
            "Epoch 18/50\n",
            "476/476 [==============================] - 2s 4ms/step - loss: 0.6657 - accuracy: 0.5920\n",
            "Epoch 19/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6643 - accuracy: 0.5916\n",
            "Epoch 20/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6650 - accuracy: 0.5943\n",
            "Epoch 21/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6649 - accuracy: 0.5943\n",
            "Epoch 22/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6648 - accuracy: 0.5938\n",
            "Epoch 23/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6652 - accuracy: 0.5926\n",
            "Epoch 24/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6636 - accuracy: 0.5964\n",
            "Epoch 25/50\n",
            "476/476 [==============================] - 2s 3ms/step - loss: 0.6639 - accuracy: 0.5955\n",
            "Epoch 26/50\n",
            "476/476 [==============================] - 2s 4ms/step - loss: 0.6650 - accuracy: 0.5957\n",
            "Epoch 27/50\n",
            "476/476 [==============================] - 2s 4ms/step - loss: 0.6644 - accuracy: 0.5953\n",
            "Epoch 28/50\n",
            "476/476 [==============================] - 2s 3ms/step - loss: 0.6651 - accuracy: 0.5936\n",
            "Epoch 29/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6648 - accuracy: 0.5926\n",
            "Epoch 30/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6645 - accuracy: 0.5947\n",
            "Epoch 31/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6637 - accuracy: 0.5951\n",
            "Epoch 32/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6656 - accuracy: 0.5959\n",
            "Epoch 33/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6643 - accuracy: 0.5964\n",
            "Epoch 34/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6645 - accuracy: 0.5981\n",
            "Epoch 35/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6642 - accuracy: 0.5928\n",
            "Epoch 36/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6644 - accuracy: 0.5978\n",
            "Epoch 37/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6646 - accuracy: 0.5964\n",
            "Epoch 38/50\n",
            "476/476 [==============================] - 2s 4ms/step - loss: 0.6640 - accuracy: 0.5936\n",
            "Epoch 39/50\n",
            "476/476 [==============================] - 2s 3ms/step - loss: 0.6641 - accuracy: 0.5938\n",
            "Epoch 40/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6648 - accuracy: 0.5948\n",
            "Epoch 41/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6642 - accuracy: 0.5960\n",
            "Epoch 42/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6629 - accuracy: 0.5966\n",
            "Epoch 43/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6640 - accuracy: 0.5966\n",
            "Epoch 44/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6640 - accuracy: 0.5935\n",
            "Epoch 45/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6629 - accuracy: 0.5980\n",
            "Epoch 46/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6634 - accuracy: 0.5955\n",
            "Epoch 47/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6638 - accuracy: 0.5933\n",
            "Epoch 48/50\n",
            "476/476 [==============================] - 1s 2ms/step - loss: 0.6635 - accuracy: 0.5955\n",
            "Epoch 49/50\n",
            "476/476 [==============================] - 2s 3ms/step - loss: 0.6643 - accuracy: 0.5953\n",
            "Epoch 50/50\n",
            "476/476 [==============================] - 2s 3ms/step - loss: 0.6640 - accuracy: 0.5979\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f26d62b9390>"
            ]
          },
          "metadata": {},
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "#Prediction on the test set from the neural network on the spreads\n",
        "y_pred2 = model2.predict(X_test2)\n",
        "y_pred2 = [0 if val < 0.5 else 1 for val in y_pred2]\n",
        "nn_test_accuracy2 = accuracy_score(y_test2, y_pred2)\n",
        "print(nn_test_accuracy2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fU3l2jiOuZAy",
        "outputId": "b1f78939-d37f-46d7-f45a-e9c0b8df766a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "119/119 [==============================] - 0s 1ms/step\n",
            "0.6224033657638707\n"
          ]
        }
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}