fully_connected_1960_without_lat_long.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(4747)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introducing, A Fully-Connected NN Model to Predict Temperature-anomaly based off Latitude, Longitude, Elevation, Species, Year, and d18O."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>elevation</th>\n",
       "      <th>species</th>\n",
       "      <th>year</th>\n",
       "      <th>d18o</th>\n",
       "      <th>temp_anom</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastaea faveolata</td>\n",
       "      <td>1927</td>\n",
       "      <td>-4.067500</td>\n",
       "      <td>-0.405418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastaea faveolata</td>\n",
       "      <td>1928</td>\n",
       "      <td>-3.865000</td>\n",
       "      <td>-0.503512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastaea faveolata</td>\n",
       "      <td>1929</td>\n",
       "      <td>-3.865000</td>\n",
       "      <td>-0.819141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastaea faveolata</td>\n",
       "      <td>1930</td>\n",
       "      <td>-4.007500</td>\n",
       "      <td>-0.337399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastaea faveolata</td>\n",
       "      <td>1931</td>\n",
       "      <td>-4.040000</td>\n",
       "      <td>-0.099585</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9693</th>\n",
       "      <td>16816</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2008</td>\n",
       "      <td>-5.474000</td>\n",
       "      <td>0.210579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9694</th>\n",
       "      <td>16817</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2009</td>\n",
       "      <td>-5.249917</td>\n",
       "      <td>0.288014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9695</th>\n",
       "      <td>16818</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2010</td>\n",
       "      <td>-5.299417</td>\n",
       "      <td>0.664223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9696</th>\n",
       "      <td>16819</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2011</td>\n",
       "      <td>-5.397417</td>\n",
       "      <td>0.425456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9697</th>\n",
       "      <td>16820</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2012</td>\n",
       "      <td>-5.337556</td>\n",
       "      <td>0.466424</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>9698 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Unnamed: 0        lat         lon elevation               species  year  \\\n",
       "0              0  11.170000  119.150000        -4  Montastaea faveolata  1927   \n",
       "1              1  11.170000  119.150000        -4  Montastaea faveolata  1928   \n",
       "2              2  11.170000  119.150000        -4  Montastaea faveolata  1929   \n",
       "3              3  11.170000  119.150000        -4  Montastaea faveolata  1930   \n",
       "4              4  11.170000  119.150000        -4  Montastaea faveolata  1931   \n",
       "...          ...        ...         ...       ...                   ...   ...   \n",
       "9693       16816  20.666667  296.833333        -5           Porites sp.  2008   \n",
       "9694       16817  20.666667  296.833333        -5           Porites sp.  2009   \n",
       "9695       16818  20.666667  296.833333        -5           Porites sp.  2010   \n",
       "9696       16819  20.666667  296.833333        -5           Porites sp.  2011   \n",
       "9697       16820  20.666667  296.833333        -5           Porites sp.  2012   \n",
       "\n",
       "          d18o  temp_anom  \n",
       "0    -4.067500  -0.405418  \n",
       "1    -3.865000  -0.503512  \n",
       "2    -3.865000  -0.819141  \n",
       "3    -4.007500  -0.337399  \n",
       "4    -4.040000  -0.099585  \n",
       "...        ...        ...  \n",
       "9693 -5.474000   0.210579  \n",
       "9694 -5.249917   0.288014  \n",
       "9695 -5.299417   0.664223  \n",
       "9696 -5.397417   0.425456  \n",
       "9697 -5.337556   0.466424  \n",
       "\n",
       "[9698 rows x 8 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# df = pd.read_csv('all_coral_with_ocean_anom.csv')\n",
    "df = pd.read_csv('all_coral_with_temp_fixed_1880-2012.csv')\n",
    "df = df.drop('Unnamed: 0.1', 1)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Fixing typos acquired from the original text scraping."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.replace(to_replace =\"Porites.sp\",\n",
    "                 value =\"Porites sp.\")\n",
    "df = df.replace(to_replace =\"Porites.sp\",\n",
    "                 value =\"Porites sp.\")\n",
    "df = df.replace(to_replace =\"Montastaea faveolata\",\n",
    "                 value =\"Montastraea faveolata\")\n",
    "df = df.replace(to_replace =\"Montastraea faviolata\",\n",
    "                 value =\"Montastraea faveolata\")         \n",
    "df = df.replace(to_replace =\"Porites australiensis\",\n",
    "                 value =\"Porites australiensus\") \n",
    "         "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One-hot encoding of the species."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>elevation</th>\n",
       "      <th>species</th>\n",
       "      <th>year</th>\n",
       "      <th>d18o</th>\n",
       "      <th>temp_anom</th>\n",
       "      <th>species_Ceratoporella nicholsoni</th>\n",
       "      <th>species_Diploastrea heliopora</th>\n",
       "      <th>...</th>\n",
       "      <th>species_Platygyra lamellina</th>\n",
       "      <th>species_Porites australiensus</th>\n",
       "      <th>species_Porites lobata</th>\n",
       "      <th>species_Porites lutea</th>\n",
       "      <th>species_Porites solida</th>\n",
       "      <th>species_Porites sp.</th>\n",
       "      <th>species_Siderastrea radians</th>\n",
       "      <th>species_Siderastrea spp.</th>\n",
       "      <th>species_Sidereastrea siderea</th>\n",
       "      <th>species_Solenastrea bournoni</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastraea faveolata</td>\n",
       "      <td>1927</td>\n",
       "      <td>-4.067500</td>\n",
       "      <td>-0.405418</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastraea faveolata</td>\n",
       "      <td>1928</td>\n",
       "      <td>-3.865000</td>\n",
       "      <td>-0.503512</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastraea faveolata</td>\n",
       "      <td>1929</td>\n",
       "      <td>-3.865000</td>\n",
       "      <td>-0.819141</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastraea faveolata</td>\n",
       "      <td>1930</td>\n",
       "      <td>-4.007500</td>\n",
       "      <td>-0.337399</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>11.170000</td>\n",
       "      <td>119.150000</td>\n",
       "      <td>-4</td>\n",
       "      <td>Montastraea faveolata</td>\n",
       "      <td>1931</td>\n",
       "      <td>-4.040000</td>\n",
       "      <td>-0.099585</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>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>9693</th>\n",
       "      <td>16816</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2008</td>\n",
       "      <td>-5.474000</td>\n",
       "      <td>0.210579</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9694</th>\n",
       "      <td>16817</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2009</td>\n",
       "      <td>-5.249917</td>\n",
       "      <td>0.288014</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9695</th>\n",
       "      <td>16818</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2010</td>\n",
       "      <td>-5.299417</td>\n",
       "      <td>0.664223</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9696</th>\n",
       "      <td>16819</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2011</td>\n",
       "      <td>-5.397417</td>\n",
       "      <td>0.425456</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9697</th>\n",
       "      <td>16820</td>\n",
       "      <td>20.666667</td>\n",
       "      <td>296.833333</td>\n",
       "      <td>-5</td>\n",
       "      <td>Porites sp.</td>\n",
       "      <td>2012</td>\n",
       "      <td>-5.337556</td>\n",
       "      <td>0.466424</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>9698 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Unnamed: 0        lat         lon elevation                species  \\\n",
       "0              0  11.170000  119.150000        -4  Montastraea faveolata   \n",
       "1              1  11.170000  119.150000        -4  Montastraea faveolata   \n",
       "2              2  11.170000  119.150000        -4  Montastraea faveolata   \n",
       "3              3  11.170000  119.150000        -4  Montastraea faveolata   \n",
       "4              4  11.170000  119.150000        -4  Montastraea faveolata   \n",
       "...          ...        ...         ...       ...                    ...   \n",
       "9693       16816  20.666667  296.833333        -5            Porites sp.   \n",
       "9694       16817  20.666667  296.833333        -5            Porites sp.   \n",
       "9695       16818  20.666667  296.833333        -5            Porites sp.   \n",
       "9696       16819  20.666667  296.833333        -5            Porites sp.   \n",
       "9697       16820  20.666667  296.833333        -5            Porites sp.   \n",
       "\n",
       "      year      d18o  temp_anom  species_Ceratoporella nicholsoni  \\\n",
       "0     1927 -4.067500  -0.405418                                 0   \n",
       "1     1928 -3.865000  -0.503512                                 0   \n",
       "2     1929 -3.865000  -0.819141                                 0   \n",
       "3     1930 -4.007500  -0.337399                                 0   \n",
       "4     1931 -4.040000  -0.099585                                 0   \n",
       "...    ...       ...        ...                               ...   \n",
       "9693  2008 -5.474000   0.210579                                 0   \n",
       "9694  2009 -5.249917   0.288014                                 0   \n",
       "9695  2010 -5.299417   0.664223                                 0   \n",
       "9696  2011 -5.397417   0.425456                                 0   \n",
       "9697  2012 -5.337556   0.466424                                 0   \n",
       "\n",
       "      species_Diploastrea heliopora  ...  species_Platygyra lamellina  \\\n",
       "0                                 0  ...                            0   \n",
       "1                                 0  ...                            0   \n",
       "2                                 0  ...                            0   \n",
       "3                                 0  ...                            0   \n",
       "4                                 0  ...                            0   \n",
       "...                             ...  ...                          ...   \n",
       "9693                              0  ...                            0   \n",
       "9694                              0  ...                            0   \n",
       "9695                              0  ...                            0   \n",
       "9696                              0  ...                            0   \n",
       "9697                              0  ...                            0   \n",
       "\n",
       "      species_Porites australiensus  species_Porites lobata  \\\n",
       "0                                 0                       0   \n",
       "1                                 0                       0   \n",
       "2                                 0                       0   \n",
       "3                                 0                       0   \n",
       "4                                 0                       0   \n",
       "...                             ...                     ...   \n",
       "9693                              0                       0   \n",
       "9694                              0                       0   \n",
       "9695                              0                       0   \n",
       "9696                              0                       0   \n",
       "9697                              0                       0   \n",
       "\n",
       "      species_Porites lutea  species_Porites solida  species_Porites sp.  \\\n",
       "0                         0                       0                    0   \n",
       "1                         0                       0                    0   \n",
       "2                         0                       0                    0   \n",
       "3                         0                       0                    0   \n",
       "4                         0                       0                    0   \n",
       "...                     ...                     ...                  ...   \n",
       "9693                      0                       0                    1   \n",
       "9694                      0                       0                    1   \n",
       "9695                      0                       0                    1   \n",
       "9696                      0                       0                    1   \n",
       "9697                      0                       0                    1   \n",
       "\n",
       "      species_Siderastrea radians  species_Siderastrea spp.  \\\n",
       "0                               0                         0   \n",
       "1                               0                         0   \n",
       "2                               0                         0   \n",
       "3                               0                         0   \n",
       "4                               0                         0   \n",
       "...                           ...                       ...   \n",
       "9693                            0                         0   \n",
       "9694                            0                         0   \n",
       "9695                            0                         0   \n",
       "9696                            0                         0   \n",
       "9697                            0                         0   \n",
       "\n",
       "      species_Sidereastrea siderea  species_Solenastrea bournoni  \n",
       "0                                0                             0  \n",
       "1                                0                             0  \n",
       "2                                0                             0  \n",
       "3                                0                             0  \n",
       "4                                0                             0  \n",
       "...                            ...                           ...  \n",
       "9693                             0                             0  \n",
       "9694                             0                             0  \n",
       "9695                             0                             0  \n",
       "9696                             0                             0  \n",
       "9697                             0                             0  \n",
       "\n",
       "[9698 rows x 28 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encoder_df = pd.get_dummies(df[['species']], prefix='species')\n",
    "df = df.join(encoder_df)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we want to predict `temp`, given `species`, `year`, `d18O`, `lat`, and `long`.\n",
    "\n",
    "First, let's configure all the data into one big dataframe. But only for years after 1960 (those are the temperature years we're pretty sure about). We will want to save the data with years before 1960 as part of our test data.\n",
    "\n",
    "Then, let's scale everything between 1 and 0 to make it more suitable for the neural network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_use = ['elevation', 'd18o', 'year']\n",
    "cols_to_use.extend(list(encoder_df.columns))\n",
    "\n",
    "df_1960 = df[df['year']<2021]\n",
    "df_1960 = df_1960[df_1960['year']>1959]\n",
    "\n",
    "df_1960 = df_1960.dropna() # this removes a TON of data.... we should find a way to not have remove all the NaNs and instead fill them with something\n",
    "df_1960 = df_1960.reset_index()\n",
    "df_1960 = df_1960.drop(columns = ['index'])\n",
    "df_1960 \n",
    "\n",
    "X = df_1960[cols_to_use]\n",
    "Y = df_1960['temp_anom']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>elevation</th>\n",
       "      <th>d18o</th>\n",
       "      <th>year</th>\n",
       "      <th>species_Ceratoporella nicholsoni</th>\n",
       "      <th>species_Diploastrea heliopora</th>\n",
       "      <th>species_Diploria labyrinthiformis</th>\n",
       "      <th>species_Diploria strigosa</th>\n",
       "      <th>species_Isopora sp</th>\n",
       "      <th>species_Montastraea faveolata</th>\n",
       "      <th>species_Orbicella faveolata</th>\n",
       "      <th>...</th>\n",
       "      <th>species_Platygyra lamellina</th>\n",
       "      <th>species_Porites australiensus</th>\n",
       "      <th>species_Porites lobata</th>\n",
       "      <th>species_Porites lutea</th>\n",
       "      <th>species_Porites solida</th>\n",
       "      <th>species_Porites sp.</th>\n",
       "      <th>species_Siderastrea radians</th>\n",
       "      <th>species_Siderastrea spp.</th>\n",
       "      <th>species_Sidereastrea siderea</th>\n",
       "      <th>species_Solenastrea bournoni</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-4</td>\n",
       "      <td>-3.777500</td>\n",
       "      <td>1960</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-4</td>\n",
       "      <td>-4.152500</td>\n",
       "      <td>1961</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-4</td>\n",
       "      <td>-4.022500</td>\n",
       "      <td>1962</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-4</td>\n",
       "      <td>-3.672500</td>\n",
       "      <td>1963</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-4</td>\n",
       "      <td>-3.912500</td>\n",
       "      <td>1964</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>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>2864</th>\n",
       "      <td>-5</td>\n",
       "      <td>-5.474000</td>\n",
       "      <td>2008</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2865</th>\n",
       "      <td>-5</td>\n",
       "      <td>-5.249917</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2866</th>\n",
       "      <td>-5</td>\n",
       "      <td>-5.299417</td>\n",
       "      <td>2010</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2867</th>\n",
       "      <td>-5</td>\n",
       "      <td>-5.397417</td>\n",
       "      <td>2011</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2868</th>\n",
       "      <td>-5</td>\n",
       "      <td>-5.337556</td>\n",
       "      <td>2012</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2869 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     elevation      d18o  year  species_Ceratoporella nicholsoni  \\\n",
       "0           -4 -3.777500  1960                                 0   \n",
       "1           -4 -4.152500  1961                                 0   \n",
       "2           -4 -4.022500  1962                                 0   \n",
       "3           -4 -3.672500  1963                                 0   \n",
       "4           -4 -3.912500  1964                                 0   \n",
       "...        ...       ...   ...                               ...   \n",
       "2864        -5 -5.474000  2008                                 0   \n",
       "2865        -5 -5.249917  2009                                 0   \n",
       "2866        -5 -5.299417  2010                                 0   \n",
       "2867        -5 -5.397417  2011                                 0   \n",
       "2868        -5 -5.337556  2012                                 0   \n",
       "\n",
       "      species_Diploastrea heliopora  species_Diploria labyrinthiformis  \\\n",
       "0                                 0                                  0   \n",
       "1                                 0                                  0   \n",
       "2                                 0                                  0   \n",
       "3                                 0                                  0   \n",
       "4                                 0                                  0   \n",
       "...                             ...                                ...   \n",
       "2864                              0                                  0   \n",
       "2865                              0                                  0   \n",
       "2866                              0                                  0   \n",
       "2867                              0                                  0   \n",
       "2868                              0                                  0   \n",
       "\n",
       "      species_Diploria strigosa  species_Isopora sp  \\\n",
       "0                             0                   0   \n",
       "1                             0                   0   \n",
       "2                             0                   0   \n",
       "3                             0                   0   \n",
       "4                             0                   0   \n",
       "...                         ...                 ...   \n",
       "2864                          0                   0   \n",
       "2865                          0                   0   \n",
       "2866                          0                   0   \n",
       "2867                          0                   0   \n",
       "2868                          0                   0   \n",
       "\n",
       "      species_Montastraea faveolata  species_Orbicella faveolata  ...  \\\n",
       "0                                 1                            0  ...   \n",
       "1                                 1                            0  ...   \n",
       "2                                 1                            0  ...   \n",
       "3                                 1                            0  ...   \n",
       "4                                 1                            0  ...   \n",
       "...                             ...                          ...  ...   \n",
       "2864                              0                            0  ...   \n",
       "2865                              0                            0  ...   \n",
       "2866                              0                            0  ...   \n",
       "2867                              0                            0  ...   \n",
       "2868                              0                            0  ...   \n",
       "\n",
       "      species_Platygyra lamellina  species_Porites australiensus  \\\n",
       "0                               0                              0   \n",
       "1                               0                              0   \n",
       "2                               0                              0   \n",
       "3                               0                              0   \n",
       "4                               0                              0   \n",
       "...                           ...                            ...   \n",
       "2864                            0                              0   \n",
       "2865                            0                              0   \n",
       "2866                            0                              0   \n",
       "2867                            0                              0   \n",
       "2868                            0                              0   \n",
       "\n",
       "      species_Porites lobata  species_Porites lutea  species_Porites solida  \\\n",
       "0                          0                      0                       0   \n",
       "1                          0                      0                       0   \n",
       "2                          0                      0                       0   \n",
       "3                          0                      0                       0   \n",
       "4                          0                      0                       0   \n",
       "...                      ...                    ...                     ...   \n",
       "2864                       0                      0                       0   \n",
       "2865                       0                      0                       0   \n",
       "2866                       0                      0                       0   \n",
       "2867                       0                      0                       0   \n",
       "2868                       0                      0                       0   \n",
       "\n",
       "      species_Porites sp.  species_Siderastrea radians  \\\n",
       "0                       0                            0   \n",
       "1                       0                            0   \n",
       "2                       0                            0   \n",
       "3                       0                            0   \n",
       "4                       0                            0   \n",
       "...                   ...                          ...   \n",
       "2864                    1                            0   \n",
       "2865                    1                            0   \n",
       "2866                    1                            0   \n",
       "2867                    1                            0   \n",
       "2868                    1                            0   \n",
       "\n",
       "      species_Siderastrea spp.  species_Sidereastrea siderea  \\\n",
       "0                            0                             0   \n",
       "1                            0                             0   \n",
       "2                            0                             0   \n",
       "3                            0                             0   \n",
       "4                            0                             0   \n",
       "...                        ...                           ...   \n",
       "2864                         0                             0   \n",
       "2865                         0                             0   \n",
       "2866                         0                             0   \n",
       "2867                         0                             0   \n",
       "2868                         0                             0   \n",
       "\n",
       "      species_Solenastrea bournoni  \n",
       "0                                0  \n",
       "1                                0  \n",
       "2                                0  \n",
       "3                                0  \n",
       "4                                0  \n",
       "...                            ...  \n",
       "2864                             0  \n",
       "2865                             0  \n",
       "2866                             0  \n",
       "2867                             0  \n",
       "2868                             0  \n",
       "\n",
       "[2869 rows x 23 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = X.reset_index()\n",
    "X = X.drop(columns = ['index'])\n",
    "X "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = X[X['year'].notna()]\n",
    "X = X[X['year']!='-999']\n",
    "X = X[X['year']!='NAN']\n",
    "\n",
    "X = X[X['elevation'].notna()]\n",
    "X = X[X['elevation']!='-999']\n",
    "X = X[X['elevation']!='NAN']\n",
    "X['elevation'] = X['elevation'].apply(lambda x: x.strip('m'))\n",
    "\n",
    "X['elevation'] = X['elevation'].apply(lambda x: float(x))\n",
    "X['year'] = X['year'].apply(lambda x: float(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Scale all the input variables between -1 and 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "ename": "ModuleNotFoundError",
     "evalue": "No module named 'sklearn'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
      "\u001b[0;32m/var/folders/09/7brpk1nd59s810dlp5fw7gt00000gn/T/ipykernel_23863/2540234040.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0msklearn\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mpreprocessing\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'sklearn'"
     ]
    }
   ],
   "source": [
    "from sklearn import preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>elevation</th>\n",
       "      <th>d18o</th>\n",
       "      <th>year</th>\n",
       "      <th>species_Ceratoporella nicholsoni</th>\n",
       "      <th>species_Diploastrea heliopora</th>\n",
       "      <th>species_Diploria labyrinthiformis</th>\n",
       "      <th>species_Diploria strigosa</th>\n",
       "      <th>species_Isopora sp</th>\n",
       "      <th>...</th>\n",
       "      <th>species_Platygyra lamellina</th>\n",
       "      <th>species_Porites australiensus</th>\n",
       "      <th>species_Porites lobata</th>\n",
       "      <th>species_Porites lutea</th>\n",
       "      <th>species_Porites solida</th>\n",
       "      <th>species_Porites sp.</th>\n",
       "      <th>species_Siderastrea radians</th>\n",
       "      <th>species_Siderastrea spp.</th>\n",
       "      <th>species_Sidereastrea siderea</th>\n",
       "      <th>species_Solenastrea bournoni</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.026103</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.107760</td>\n",
       "      <td>-0.962264</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.079452</td>\n",
       "      <td>-0.924528</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.003239</td>\n",
       "      <td>-0.886792</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.055500</td>\n",
       "      <td>-0.849057</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        lat       lon  elevation      d18o      year  \\\n",
       "0  0.300928 -0.406596   0.819549 -0.026103 -1.000000   \n",
       "1  0.300928 -0.406596   0.819549 -0.107760 -0.962264   \n",
       "2  0.300928 -0.406596   0.819549 -0.079452 -0.924528   \n",
       "3  0.300928 -0.406596   0.819549 -0.003239 -0.886792   \n",
       "4  0.300928 -0.406596   0.819549 -0.055500 -0.849057   \n",
       "\n",
       "   species_Ceratoporella nicholsoni  species_Diploastrea heliopora  \\\n",
       "0                              -1.0                           -1.0   \n",
       "1                              -1.0                           -1.0   \n",
       "2                              -1.0                           -1.0   \n",
       "3                              -1.0                           -1.0   \n",
       "4                              -1.0                           -1.0   \n",
       "\n",
       "   species_Diploria labyrinthiformis  species_Diploria strigosa  \\\n",
       "0                               -1.0                       -1.0   \n",
       "1                               -1.0                       -1.0   \n",
       "2                               -1.0                       -1.0   \n",
       "3                               -1.0                       -1.0   \n",
       "4                               -1.0                       -1.0   \n",
       "\n",
       "   species_Isopora sp  ...  species_Platygyra lamellina  \\\n",
       "0                -1.0  ...                         -1.0   \n",
       "1                -1.0  ...                         -1.0   \n",
       "2                -1.0  ...                         -1.0   \n",
       "3                -1.0  ...                         -1.0   \n",
       "4                -1.0  ...                         -1.0   \n",
       "\n",
       "   species_Porites australiensus  species_Porites lobata  \\\n",
       "0                           -1.0                    -1.0   \n",
       "1                           -1.0                    -1.0   \n",
       "2                           -1.0                    -1.0   \n",
       "3                           -1.0                    -1.0   \n",
       "4                           -1.0                    -1.0   \n",
       "\n",
       "   species_Porites lutea  species_Porites solida  species_Porites sp.  \\\n",
       "0                   -1.0                    -1.0                 -1.0   \n",
       "1                   -1.0                    -1.0                 -1.0   \n",
       "2                   -1.0                    -1.0                 -1.0   \n",
       "3                   -1.0                    -1.0                 -1.0   \n",
       "4                   -1.0                    -1.0                 -1.0   \n",
       "\n",
       "   species_Siderastrea radians  species_Siderastrea spp.  \\\n",
       "0                         -1.0                      -1.0   \n",
       "1                         -1.0                      -1.0   \n",
       "2                         -1.0                      -1.0   \n",
       "3                         -1.0                      -1.0   \n",
       "4                         -1.0                      -1.0   \n",
       "\n",
       "   species_Sidereastrea siderea  species_Solenastrea bournoni  \n",
       "0                          -1.0                          -1.0  \n",
       "1                          -1.0                          -1.0  \n",
       "2                          -1.0                          -1.0  \n",
       "3                          -1.0                          -1.0  \n",
       "4                          -1.0                          -1.0  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 244,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = preprocessing.MinMaxScaler(feature_range=(-1, 1))\n",
    "\n",
    "cols = X.columns\n",
    "d = scaler.fit_transform(X)\n",
    "X_scaled = pd.DataFrame(d, columns=cols)\n",
    "X_scaled.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>temp_anom</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.284105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.504643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.531974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.755829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.417459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2864</th>\n",
       "      <td>0.210579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2865</th>\n",
       "      <td>0.288014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2866</th>\n",
       "      <td>0.664223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2867</th>\n",
       "      <td>0.425456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2868</th>\n",
       "      <td>0.466424</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2869 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      temp_anom\n",
       "0     -0.284105\n",
       "1     -0.504643\n",
       "2     -0.531974\n",
       "3     -0.755829\n",
       "4     -0.417459\n",
       "...         ...\n",
       "2864   0.210579\n",
       "2865   0.288014\n",
       "2866   0.664223\n",
       "2867   0.425456\n",
       "2868   0.466424\n",
       "\n",
       "[2869 rows x 1 columns]"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y = Y.reset_index()\n",
    "Y = Y.drop(columns = ['index'])\n",
    "Y "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.model_selection import train_test_split\n",
    "# X_train, X_val_and_test, Y_train, Y_val_and_test = train_test_split(\n",
    "#     X_scaled, Y, test_size=0.2, random_state=47)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X_val_and_test.shape, Y_val_and_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X_val_and_test.index, Y_val_and_test.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X_train.shape, Y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout\n",
    "from keras.layers.advanced_activations import LeakyReLU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy.random import seed\n",
    "seed(47)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sequential Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.1312 - mse: 0.1312\n",
      "Epoch 2/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.1236 - mse: 0.1236\n",
      "Epoch 3/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.1183 - mse: 0.1183\n",
      "Epoch 4/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.1162 - mse: 0.1162\n",
      "Epoch 5/50\n",
      "90/90 [==============================] - 0s 858us/step - loss: 0.1092 - mse: 0.1092\n",
      "Epoch 6/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.1068 - mse: 0.1068\n",
      "Epoch 7/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.1028 - mse: 0.1028\n",
      "Epoch 8/50\n",
      "90/90 [==============================] - 0s 851us/step - loss: 0.1013 - mse: 0.1013\n",
      "Epoch 9/50\n",
      "90/90 [==============================] - 0s 884us/step - loss: 0.0989 - mse: 0.0989\n",
      "Epoch 10/50\n",
      "90/90 [==============================] - 0s 849us/step - loss: 0.0988 - mse: 0.0988\n",
      "Epoch 11/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0972 - mse: 0.0972\n",
      "Epoch 12/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0975 - mse: 0.0975\n",
      "Epoch 13/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0972 - mse: 0.0972\n",
      "Epoch 14/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0976 - mse: 0.0976\n",
      "Epoch 15/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0965 - mse: 0.0965\n",
      "Epoch 16/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0952 - mse: 0.0952\n",
      "Epoch 17/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0964 - mse: 0.0964\n",
      "Epoch 18/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0943 - mse: 0.0943\n",
      "Epoch 19/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0950 - mse: 0.0950\n",
      "Epoch 20/50\n",
      "90/90 [==============================] - 0s 1ms/step - loss: 0.0947 - mse: 0.0947\n",
      "Epoch 21/50\n",
      "90/90 [==============================] - 0s 951us/step - loss: 0.0953 - mse: 0.0953\n",
      "Epoch 22/50\n",
      "90/90 [==============================] - 0s 975us/step - loss: 0.0924 - mse: 0.0924\n",
      "Epoch 23/50\n",
      "90/90 [==============================] - 0s 850us/step - loss: 0.0939 - mse: 0.0939\n",
      "Epoch 24/50\n",
      "90/90 [==============================] - 0s 824us/step - loss: 0.0927 - mse: 0.0927\n",
      "Epoch 25/50\n",
      "90/90 [==============================] - 0s 875us/step - loss: 0.0924 - mse: 0.0924\n",
      "Epoch 26/50\n",
      "90/90 [==============================] - 0s 880us/step - loss: 0.0946 - mse: 0.0946\n",
      "Epoch 27/50\n",
      "90/90 [==============================] - 0s 892us/step - loss: 0.0929 - mse: 0.0929\n",
      "Epoch 28/50\n",
      "90/90 [==============================] - 0s 908us/step - loss: 0.0923 - mse: 0.0923\n",
      "Epoch 29/50\n",
      "90/90 [==============================] - 0s 850us/step - loss: 0.0930 - mse: 0.0930\n",
      "Epoch 30/50\n",
      "90/90 [==============================] - 0s 841us/step - loss: 0.0935 - mse: 0.0935\n",
      "Epoch 31/50\n",
      "90/90 [==============================] - 0s 903us/step - loss: 0.0923 - mse: 0.0923\n",
      "Epoch 32/50\n",
      "90/90 [==============================] - 0s 849us/step - loss: 0.0918 - mse: 0.0918\n",
      "Epoch 33/50\n",
      "90/90 [==============================] - 0s 866us/step - loss: 0.0912 - mse: 0.0912\n",
      "Epoch 34/50\n",
      "90/90 [==============================] - 0s 874us/step - loss: 0.0917 - mse: 0.0917\n",
      "Epoch 35/50\n",
      "90/90 [==============================] - 0s 819us/step - loss: 0.0914 - mse: 0.0914\n",
      "Epoch 36/50\n",
      "90/90 [==============================] - 0s 804us/step - loss: 0.0919 - mse: 0.0919\n",
      "Epoch 37/50\n",
      "90/90 [==============================] - 0s 848us/step - loss: 0.0915 - mse: 0.0915\n",
      "Epoch 38/50\n",
      "90/90 [==============================] - 0s 860us/step - loss: 0.0918 - mse: 0.0918\n",
      "Epoch 39/50\n",
      "90/90 [==============================] - 0s 830us/step - loss: 0.0908 - mse: 0.0908\n",
      "Epoch 40/50\n",
      "90/90 [==============================] - 0s 871us/step - loss: 0.0920 - mse: 0.0920\n",
      "Epoch 41/50\n",
      "90/90 [==============================] - 0s 913us/step - loss: 0.0907 - mse: 0.0907\n",
      "Epoch 42/50\n",
      "90/90 [==============================] - 0s 820us/step - loss: 0.0909 - mse: 0.0909\n",
      "Epoch 43/50\n",
      "90/90 [==============================] - 0s 884us/step - loss: 0.0921 - mse: 0.0921\n",
      "Epoch 44/50\n",
      "90/90 [==============================] - 0s 849us/step - loss: 0.0915 - mse: 0.0915\n",
      "Epoch 45/50\n",
      "90/90 [==============================] - 0s 828us/step - loss: 0.0921 - mse: 0.0921\n",
      "Epoch 46/50\n",
      "90/90 [==============================] - 0s 860us/step - loss: 0.0908 - mse: 0.0908\n",
      "Epoch 47/50\n",
      "90/90 [==============================] - 0s 856us/step - loss: 0.0895 - mse: 0.0895\n",
      "Epoch 48/50\n",
      "90/90 [==============================] - 0s 870us/step - loss: 0.0914 - mse: 0.0914\n",
      "Epoch 49/50\n",
      "90/90 [==============================] - 0s 2ms/step - loss: 0.0921 - mse: 0.0921\n",
      "Epoch 50/50\n",
      "90/90 [==============================] - 0s 793us/step - loss: 0.0910 - mse: 0.0910\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7fea29a4b520>"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "network = Sequential()\n",
    "\n",
    "#input layer\n",
    "network.add(Dense(8, input_shape=(23,), activation='LeakyReLU'))\n",
    "\n",
    "#middle layers\n",
    "network.add(Dense(6, activation='LeakyReLU'))\n",
    "network.add(Dropout(0.2))\n",
    "\n",
    "network.add(Dense(6, activation='LeakyReLU'))\n",
    "\n",
    "network.add(Dense(4, activation='LeakyReLU'))\n",
    "network.add(Dropout(0.2))\n",
    "\n",
    "#output layer\n",
    "network.add(Dense(1, activation='LeakyReLU'))\n",
    "\n",
    "network.compile('adam', loss='mse', metrics=['mse'])\n",
    "#network.fit(X_train, Y_train, epochs=200)\n",
    "network.fit(X_scaled, Y, epochs=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 [==============================] - 0s 663us/step - loss: 0.0837 - mse: 0.0837\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.08372066915035248, 0.08372066915035248]"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# network.evaluate(X_val_and_test, Y_val_and_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This network with a dropout layer produces a teeny loss, maybe we're overfitting?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [],
   "source": [
    "# actual = pd.DataFrame(Y_val_and_test, columns=[\"temp_anom\"])\n",
    "# predicted = pd.DataFrame(network.predict(X_val_and_test), columns=[\"pred\"])\n",
    "# actual.head(), predicted.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [],
   "source": [
    "# predicted['index'] = actual.index\n",
    "# predicted = predicted.set_index(['index'])\n",
    "# predicted['index'] = actual.index\n",
    "# predicted = predicted.set_index(['index'])\n",
    "# predicted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [],
   "source": [
    "# actual_v_pred = actual.join(predicted, how = \"inner\")\n",
    "# actual_v_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.metrics import mean_squared_error\n",
    "# mean_squared_error(Y_val_and_test, predicted)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test_data_with_preds = actual_v_pred.join(X_val_and_test, how = \"inner\")\n",
    "# test_data_with_preds['diff'] = test_data_with_preds['temp_anom'] - test_data_with_preds['pred']\n",
    "# test_data_with_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "# x = test_data_with_preds['temp_anom']\n",
    "# y = test_data_with_preds['pred']\n",
    "# plt.scatter(test_data_with_preds['temp_anom'], test_data_with_preds['pred'], color=\"pink\")\n",
    "# plt.plot(np.unique(x), np.poly1d(np.polyfit(x, y, 1))(np.unique(x)))\n",
    "# plt.xlim([-2, 1.5])\n",
    "# plt.ylim([-2, 1.5])\n",
    "# plt.title('Actual v. Predicted Gridded Temperature Anomaly')\n",
    "# plt.xlabel('Actual')\n",
    "# plt.ylabel('Predicted')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [],
   "source": [
    "# x = test_data_with_preds['d18o']\n",
    "# y = test_data_with_preds['diff']\n",
    "# plt.scatter(test_data_with_preds['d18o'], test_data_with_preds['diff'], color=\"purple\")\n",
    "# plt.plot(np.unique(x), np.poly1d(np.polyfit(x, y, 1))(np.unique(x)))\n",
    "# plt.xlim([-1, 1])\n",
    "# plt.ylim([-1.5, 1.5])\n",
    "# plt.title('d18O v. Gridded Temperature Anomaly Error')\n",
    "# plt.xlabel('d18O')\n",
    "# plt.ylabel('Temp Error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plt.scatter(test_data_with_preds['lat'], test_data_with_preds['diff'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plt.scatter(test_data_with_preds['lon'], test_data_with_preds['diff'], color=\"orange\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's now take all of our predictions, average them by year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test_data_with_preds.boxplot(column = ['pred', 'temp_anom'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [],
   "source": [
    "# annual_temp_and_preds = test_data_with_preds[['temp_anom', 'pred']].join(X, how=\"inner\")\n",
    "# annual_compare = annual_temp_and_preds.groupby('year', as_index=False)['temp_anom', 'pred'].mean()\n",
    "# annual_compare.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plt.plot(annual_compare['year'], annual_compare['temp_anom']), plt.plot(annual_compare['year'], annual_compare['pred'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sweet --- looks like we've at least followed the general trend of global warming :)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [],
   "source": [
    "# np.corrcoef(annual_compare['temp_anom'], annual_compare['pred'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [],
   "source": [
    "# np.var(annual_compare['temp_anom']), np.var(annual_compare['pred'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We should note that our predictions do not vary as much as the actual values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [],
   "source": [
    "# annual_compare.to_csv('preds_1960-2020.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see how our model performs on the years prior to 1960."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_use = ['elevation', 'd18o', 'year']\n",
    "cols_to_use.extend(list(encoder_df.columns))\n",
    "\n",
    "df_past = df[df['year']<1960]\n",
    "\n",
    "df_past = df_past.dropna()\n",
    "df_past = df_past.reset_index()\n",
    "df_past = df_past.drop(columns = ['index'])\n",
    "df_past \n",
    "\n",
    "X_past = df_past[cols_to_use]\n",
    "Y_past = df_past['temp_anom']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>elevation</th>\n",
       "      <th>d18o</th>\n",
       "      <th>year</th>\n",
       "      <th>species_Ceratoporella nicholsoni</th>\n",
       "      <th>species_Diploastrea heliopora</th>\n",
       "      <th>species_Diploria labyrinthiformis</th>\n",
       "      <th>species_Diploria strigosa</th>\n",
       "      <th>species_Isopora sp</th>\n",
       "      <th>...</th>\n",
       "      <th>species_Platygyra lamellina</th>\n",
       "      <th>species_Porites australiensus</th>\n",
       "      <th>species_Porites lobata</th>\n",
       "      <th>species_Porites lutea</th>\n",
       "      <th>species_Porites solida</th>\n",
       "      <th>species_Porites sp.</th>\n",
       "      <th>species_Siderastrea radians</th>\n",
       "      <th>species_Siderastrea spp.</th>\n",
       "      <th>species_Sidereastrea siderea</th>\n",
       "      <th>species_Solenastrea bournoni</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11.17</td>\n",
       "      <td>119.15</td>\n",
       "      <td>-4</td>\n",
       "      <td>-4.067500</td>\n",
       "      <td>1927</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11.17</td>\n",
       "      <td>119.15</td>\n",
       "      <td>-4</td>\n",
       "      <td>-3.865000</td>\n",
       "      <td>1928</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11.17</td>\n",
       "      <td>119.15</td>\n",
       "      <td>-4</td>\n",
       "      <td>-3.865000</td>\n",
       "      <td>1929</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11.17</td>\n",
       "      <td>119.15</td>\n",
       "      <td>-4</td>\n",
       "      <td>-4.007500</td>\n",
       "      <td>1930</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11.17</td>\n",
       "      <td>119.15</td>\n",
       "      <td>-4</td>\n",
       "      <td>-4.040000</td>\n",
       "      <td>1931</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>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>4025</th>\n",
       "      <td>5.87</td>\n",
       "      <td>17.87</td>\n",
       "      <td>-10</td>\n",
       "      <td>-4.753333</td>\n",
       "      <td>1955</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4026</th>\n",
       "      <td>5.87</td>\n",
       "      <td>17.87</td>\n",
       "      <td>-10</td>\n",
       "      <td>-4.785000</td>\n",
       "      <td>1956</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4027</th>\n",
       "      <td>5.87</td>\n",
       "      <td>17.87</td>\n",
       "      <td>-10</td>\n",
       "      <td>-4.924167</td>\n",
       "      <td>1957</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4028</th>\n",
       "      <td>5.87</td>\n",
       "      <td>17.87</td>\n",
       "      <td>-10</td>\n",
       "      <td>-5.017500</td>\n",
       "      <td>1958</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4029</th>\n",
       "      <td>5.87</td>\n",
       "      <td>17.87</td>\n",
       "      <td>-10</td>\n",
       "      <td>-4.970833</td>\n",
       "      <td>1959</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4030 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        lat     lon elevation      d18o  year  \\\n",
       "0     11.17  119.15        -4 -4.067500  1927   \n",
       "1     11.17  119.15        -4 -3.865000  1928   \n",
       "2     11.17  119.15        -4 -3.865000  1929   \n",
       "3     11.17  119.15        -4 -4.007500  1930   \n",
       "4     11.17  119.15        -4 -4.040000  1931   \n",
       "...     ...     ...       ...       ...   ...   \n",
       "4025   5.87   17.87       -10 -4.753333  1955   \n",
       "4026   5.87   17.87       -10 -4.785000  1956   \n",
       "4027   5.87   17.87       -10 -4.924167  1957   \n",
       "4028   5.87   17.87       -10 -5.017500  1958   \n",
       "4029   5.87   17.87       -10 -4.970833  1959   \n",
       "\n",
       "      species_Ceratoporella nicholsoni  species_Diploastrea heliopora  \\\n",
       "0                                    0                              0   \n",
       "1                                    0                              0   \n",
       "2                                    0                              0   \n",
       "3                                    0                              0   \n",
       "4                                    0                              0   \n",
       "...                                ...                            ...   \n",
       "4025                                 0                              0   \n",
       "4026                                 0                              0   \n",
       "4027                                 0                              0   \n",
       "4028                                 0                              0   \n",
       "4029                                 0                              0   \n",
       "\n",
       "      species_Diploria labyrinthiformis  species_Diploria strigosa  \\\n",
       "0                                     0                          0   \n",
       "1                                     0                          0   \n",
       "2                                     0                          0   \n",
       "3                                     0                          0   \n",
       "4                                     0                          0   \n",
       "...                                 ...                        ...   \n",
       "4025                                  0                          0   \n",
       "4026                                  0                          0   \n",
       "4027                                  0                          0   \n",
       "4028                                  0                          0   \n",
       "4029                                  0                          0   \n",
       "\n",
       "      species_Isopora sp  ...  species_Platygyra lamellina  \\\n",
       "0                      0  ...                            0   \n",
       "1                      0  ...                            0   \n",
       "2                      0  ...                            0   \n",
       "3                      0  ...                            0   \n",
       "4                      0  ...                            0   \n",
       "...                  ...  ...                          ...   \n",
       "4025                   0  ...                            0   \n",
       "4026                   0  ...                            0   \n",
       "4027                   0  ...                            0   \n",
       "4028                   0  ...                            0   \n",
       "4029                   0  ...                            0   \n",
       "\n",
       "      species_Porites australiensus  species_Porites lobata  \\\n",
       "0                                 0                       0   \n",
       "1                                 0                       0   \n",
       "2                                 0                       0   \n",
       "3                                 0                       0   \n",
       "4                                 0                       0   \n",
       "...                             ...                     ...   \n",
       "4025                              0                       0   \n",
       "4026                              0                       0   \n",
       "4027                              0                       0   \n",
       "4028                              0                       0   \n",
       "4029                              0                       0   \n",
       "\n",
       "      species_Porites lutea  species_Porites solida  species_Porites sp.  \\\n",
       "0                         0                       0                    0   \n",
       "1                         0                       0                    0   \n",
       "2                         0                       0                    0   \n",
       "3                         0                       0                    0   \n",
       "4                         0                       0                    0   \n",
       "...                     ...                     ...                  ...   \n",
       "4025                      1                       0                    0   \n",
       "4026                      1                       0                    0   \n",
       "4027                      1                       0                    0   \n",
       "4028                      1                       0                    0   \n",
       "4029                      1                       0                    0   \n",
       "\n",
       "      species_Siderastrea radians  species_Siderastrea spp.  \\\n",
       "0                               0                         0   \n",
       "1                               0                         0   \n",
       "2                               0                         0   \n",
       "3                               0                         0   \n",
       "4                               0                         0   \n",
       "...                           ...                       ...   \n",
       "4025                            0                         0   \n",
       "4026                            0                         0   \n",
       "4027                            0                         0   \n",
       "4028                            0                         0   \n",
       "4029                            0                         0   \n",
       "\n",
       "      species_Sidereastrea siderea  species_Solenastrea bournoni  \n",
       "0                                0                             0  \n",
       "1                                0                             0  \n",
       "2                                0                             0  \n",
       "3                                0                             0  \n",
       "4                                0                             0  \n",
       "...                            ...                           ...  \n",
       "4025                             0                             0  \n",
       "4026                             0                             0  \n",
       "4027                             0                             0  \n",
       "4028                             0                             0  \n",
       "4029                             0                             0  \n",
       "\n",
       "[4030 rows x 25 columns]"
      ]
     },
     "execution_count": 218,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_past = X_past.reset_index()\n",
    "X_past = X_past.drop(columns = ['index'])\n",
    "X_past"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_past = X_past[X_past['year'].notna()]\n",
    "X_past = X_past[X_past['year']!='-999']\n",
    "X_past = X_past[X_past['year']!='NAN']\n",
    "\n",
    "X_past = X_past[X_past['elevation'].notna()]\n",
    "X_past = X_past[X_past['elevation']!='-999']\n",
    "X_past = X_past[X_past['elevation']!='NAN']\n",
    "X_past['elevation'] = X_past['elevation'].apply(lambda x: x.strip('m'))\n",
    "\n",
    "X_past['elevation'] = X_past['elevation'].apply(lambda x: float(x))\n",
    "X_past['year'] = X_past['year'].apply(lambda x: float(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>elevation</th>\n",
       "      <th>d18o</th>\n",
       "      <th>year</th>\n",
       "      <th>species_Ceratoporella nicholsoni</th>\n",
       "      <th>species_Diploastrea heliopora</th>\n",
       "      <th>species_Diploria labyrinthiformis</th>\n",
       "      <th>species_Diploria strigosa</th>\n",
       "      <th>species_Isopora sp</th>\n",
       "      <th>...</th>\n",
       "      <th>species_Platygyra lamellina</th>\n",
       "      <th>species_Porites australiensus</th>\n",
       "      <th>species_Porites lobata</th>\n",
       "      <th>species_Porites lutea</th>\n",
       "      <th>species_Porites solida</th>\n",
       "      <th>species_Porites sp.</th>\n",
       "      <th>species_Siderastrea radians</th>\n",
       "      <th>species_Siderastrea spp.</th>\n",
       "      <th>species_Sidereastrea siderea</th>\n",
       "      <th>species_Solenastrea bournoni</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.445162</td>\n",
       "      <td>0.189873</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.397605</td>\n",
       "      <td>0.215190</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.397605</td>\n",
       "      <td>0.240506</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.431071</td>\n",
       "      <td>0.265823</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.438704</td>\n",
       "      <td>0.291139</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        lat       lon  elevation      d18o      year  \\\n",
       "0  0.300928 -0.406596   0.819549 -0.445162  0.189873   \n",
       "1  0.300928 -0.406596   0.819549 -0.397605  0.215190   \n",
       "2  0.300928 -0.406596   0.819549 -0.397605  0.240506   \n",
       "3  0.300928 -0.406596   0.819549 -0.431071  0.265823   \n",
       "4  0.300928 -0.406596   0.819549 -0.438704  0.291139   \n",
       "\n",
       "   species_Ceratoporella nicholsoni  species_Diploastrea heliopora  \\\n",
       "0                              -1.0                           -1.0   \n",
       "1                              -1.0                           -1.0   \n",
       "2                              -1.0                           -1.0   \n",
       "3                              -1.0                           -1.0   \n",
       "4                              -1.0                           -1.0   \n",
       "\n",
       "   species_Diploria labyrinthiformis  species_Diploria strigosa  \\\n",
       "0                               -1.0                       -1.0   \n",
       "1                               -1.0                       -1.0   \n",
       "2                               -1.0                       -1.0   \n",
       "3                               -1.0                       -1.0   \n",
       "4                               -1.0                       -1.0   \n",
       "\n",
       "   species_Isopora sp  ...  species_Platygyra lamellina  \\\n",
       "0                -1.0  ...                         -1.0   \n",
       "1                -1.0  ...                         -1.0   \n",
       "2                -1.0  ...                         -1.0   \n",
       "3                -1.0  ...                         -1.0   \n",
       "4                -1.0  ...                         -1.0   \n",
       "\n",
       "   species_Porites australiensus  species_Porites lobata  \\\n",
       "0                           -1.0                    -1.0   \n",
       "1                           -1.0                    -1.0   \n",
       "2                           -1.0                    -1.0   \n",
       "3                           -1.0                    -1.0   \n",
       "4                           -1.0                    -1.0   \n",
       "\n",
       "   species_Porites lutea  species_Porites solida  species_Porites sp.  \\\n",
       "0                   -1.0                    -1.0                 -1.0   \n",
       "1                   -1.0                    -1.0                 -1.0   \n",
       "2                   -1.0                    -1.0                 -1.0   \n",
       "3                   -1.0                    -1.0                 -1.0   \n",
       "4                   -1.0                    -1.0                 -1.0   \n",
       "\n",
       "   species_Siderastrea radians  species_Siderastrea spp.  \\\n",
       "0                         -1.0                      -1.0   \n",
       "1                         -1.0                      -1.0   \n",
       "2                         -1.0                      -1.0   \n",
       "3                         -1.0                      -1.0   \n",
       "4                         -1.0                      -1.0   \n",
       "\n",
       "   species_Sidereastrea siderea  species_Solenastrea bournoni  \n",
       "0                          -1.0                          -1.0  \n",
       "1                          -1.0                          -1.0  \n",
       "2                          -1.0                          -1.0  \n",
       "3                          -1.0                          -1.0  \n",
       "4                          -1.0                          -1.0  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = preprocessing.MinMaxScaler(feature_range=(-1, 1))\n",
    "\n",
    "cols = X_past.columns\n",
    "d = scaler.fit_transform(X_past)\n",
    "X_past = pd.DataFrame(d, columns=cols)\n",
    "X_past.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>temp_anom</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.405418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.503512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.819141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.337399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.099585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4025</th>\n",
       "      <td>-0.679975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4026</th>\n",
       "      <td>-0.700354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4027</th>\n",
       "      <td>-0.415288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4028</th>\n",
       "      <td>-0.047501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4029</th>\n",
       "      <td>-0.290741</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4030 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      temp_anom\n",
       "0     -0.405418\n",
       "1     -0.503512\n",
       "2     -0.819141\n",
       "3     -0.337399\n",
       "4     -0.099585\n",
       "...         ...\n",
       "4025  -0.679975\n",
       "4026  -0.700354\n",
       "4027  -0.415288\n",
       "4028  -0.047501\n",
       "4029  -0.290741\n",
       "\n",
       "[4030 rows x 1 columns]"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_past = Y_past.reset_index()\n",
    "Y_past = Y_past.drop(columns = ['index'])\n",
    "Y_past "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use this functionality to convert `X_past` and `Y_past` from `nparrays` to `tensors`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(   temp_anom\n",
       " 0  -0.405418\n",
       " 1  -0.503512\n",
       " 2  -0.819141\n",
       " 3  -0.337399\n",
       " 4  -0.099585,\n",
       "        pred\n",
       " 0 -0.007055\n",
       " 1 -0.000950\n",
       " 2  0.008539\n",
       " 3  0.016193\n",
       " 4  0.025254)"
      ]
     },
     "execution_count": 222,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "actual_past = pd.DataFrame(Y_past, columns=[\"temp_anom\"])\n",
    "predicted_past = pd.DataFrame(network.predict(X_past), columns=[\"pred\"])\n",
    "actual_past.head(), predicted_past.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "126/126 [==============================] - 0s 1ms/step - loss: 0.4693 - mse: 0.4693\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.46934279799461365, 0.46934279799461365]"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "network.evaluate(X_past, Y_past)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have a much larger `test` loss than we did within the ranges of our training bounds. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>temp_anom</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.405418</td>\n",
       "      <td>-0.007055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.503512</td>\n",
       "      <td>-0.000950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.819141</td>\n",
       "      <td>0.008539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.337399</td>\n",
       "      <td>0.016193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.099585</td>\n",
       "      <td>0.025254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4025</th>\n",
       "      <td>-0.679975</td>\n",
       "      <td>0.777254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4026</th>\n",
       "      <td>-0.700354</td>\n",
       "      <td>0.801943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4027</th>\n",
       "      <td>-0.415288</td>\n",
       "      <td>0.824702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4028</th>\n",
       "      <td>-0.047501</td>\n",
       "      <td>0.848284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4029</th>\n",
       "      <td>-0.290741</td>\n",
       "      <td>0.874381</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4030 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      temp_anom      pred\n",
       "0     -0.405418 -0.007055\n",
       "1     -0.503512 -0.000950\n",
       "2     -0.819141  0.008539\n",
       "3     -0.337399  0.016193\n",
       "4     -0.099585  0.025254\n",
       "...         ...       ...\n",
       "4025  -0.679975  0.777254\n",
       "4026  -0.700354  0.801943\n",
       "4027  -0.415288  0.824702\n",
       "4028  -0.047501  0.848284\n",
       "4029  -0.290741  0.874381\n",
       "\n",
       "[4030 rows x 2 columns]"
      ]
     },
     "execution_count": 224,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "actual_v_pred_past = actual_past.join(predicted_past, how = \"inner\")\n",
    "actual_v_pred_past"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>temp_anom</th>\n",
       "      <th>pred</th>\n",
       "      <th>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>elevation</th>\n",
       "      <th>d18o</th>\n",
       "      <th>year</th>\n",
       "      <th>species_Ceratoporella nicholsoni</th>\n",
       "      <th>species_Diploastrea heliopora</th>\n",
       "      <th>species_Diploria labyrinthiformis</th>\n",
       "      <th>...</th>\n",
       "      <th>species_Porites australiensus</th>\n",
       "      <th>species_Porites lobata</th>\n",
       "      <th>species_Porites lutea</th>\n",
       "      <th>species_Porites solida</th>\n",
       "      <th>species_Porites sp.</th>\n",
       "      <th>species_Siderastrea radians</th>\n",
       "      <th>species_Siderastrea spp.</th>\n",
       "      <th>species_Sidereastrea siderea</th>\n",
       "      <th>species_Solenastrea bournoni</th>\n",
       "      <th>diff</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.405418</td>\n",
       "      <td>-0.007055</td>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.445162</td>\n",
       "      <td>0.189873</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-0.398363</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.503512</td>\n",
       "      <td>-0.000950</td>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.397605</td>\n",
       "      <td>0.215190</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-0.502562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.819141</td>\n",
       "      <td>0.008539</td>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.397605</td>\n",
       "      <td>0.240506</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-0.827680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.337399</td>\n",
       "      <td>0.016193</td>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.431071</td>\n",
       "      <td>0.265823</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-0.353592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.099585</td>\n",
       "      <td>0.025254</td>\n",
       "      <td>0.300928</td>\n",
       "      <td>-0.406596</td>\n",
       "      <td>0.819549</td>\n",
       "      <td>-0.438704</td>\n",
       "      <td>0.291139</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-0.124839</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>4025</th>\n",
       "      <td>-0.679975</td>\n",
       "      <td>0.777254</td>\n",
       "      <td>0.126953</td>\n",
       "      <td>-0.999982</td>\n",
       "      <td>0.729323</td>\n",
       "      <td>-0.606231</td>\n",
       "      <td>0.898734</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.457229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4026</th>\n",
       "      <td>-0.700354</td>\n",
       "      <td>0.801943</td>\n",
       "      <td>0.126953</td>\n",
       "      <td>-0.999982</td>\n",
       "      <td>0.729323</td>\n",
       "      <td>-0.613668</td>\n",
       "      <td>0.924051</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.502298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4027</th>\n",
       "      <td>-0.415288</td>\n",
       "      <td>0.824702</td>\n",
       "      <td>0.126953</td>\n",
       "      <td>-0.999982</td>\n",
       "      <td>0.729323</td>\n",
       "      <td>-0.646352</td>\n",
       "      <td>0.949367</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.239990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4028</th>\n",
       "      <td>-0.047501</td>\n",
       "      <td>0.848284</td>\n",
       "      <td>0.126953</td>\n",
       "      <td>-0.999982</td>\n",
       "      <td>0.729323</td>\n",
       "      <td>-0.668271</td>\n",
       "      <td>0.974684</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-0.895785</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4029</th>\n",
       "      <td>-0.290741</td>\n",
       "      <td>0.874381</td>\n",
       "      <td>0.126953</td>\n",
       "      <td>-0.999982</td>\n",
       "      <td>0.729323</td>\n",
       "      <td>-0.657312</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.165122</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4030 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      temp_anom      pred       lat       lon  elevation      d18o      year  \\\n",
       "0     -0.405418 -0.007055  0.300928 -0.406596   0.819549 -0.445162  0.189873   \n",
       "1     -0.503512 -0.000950  0.300928 -0.406596   0.819549 -0.397605  0.215190   \n",
       "2     -0.819141  0.008539  0.300928 -0.406596   0.819549 -0.397605  0.240506   \n",
       "3     -0.337399  0.016193  0.300928 -0.406596   0.819549 -0.431071  0.265823   \n",
       "4     -0.099585  0.025254  0.300928 -0.406596   0.819549 -0.438704  0.291139   \n",
       "...         ...       ...       ...       ...        ...       ...       ...   \n",
       "4025  -0.679975  0.777254  0.126953 -0.999982   0.729323 -0.606231  0.898734   \n",
       "4026  -0.700354  0.801943  0.126953 -0.999982   0.729323 -0.613668  0.924051   \n",
       "4027  -0.415288  0.824702  0.126953 -0.999982   0.729323 -0.646352  0.949367   \n",
       "4028  -0.047501  0.848284  0.126953 -0.999982   0.729323 -0.668271  0.974684   \n",
       "4029  -0.290741  0.874381  0.126953 -0.999982   0.729323 -0.657312  1.000000   \n",
       "\n",
       "      species_Ceratoporella nicholsoni  species_Diploastrea heliopora  \\\n",
       "0                                 -1.0                           -1.0   \n",
       "1                                 -1.0                           -1.0   \n",
       "2                                 -1.0                           -1.0   \n",
       "3                                 -1.0                           -1.0   \n",
       "4                                 -1.0                           -1.0   \n",
       "...                                ...                            ...   \n",
       "4025                              -1.0                           -1.0   \n",
       "4026                              -1.0                           -1.0   \n",
       "4027                              -1.0                           -1.0   \n",
       "4028                              -1.0                           -1.0   \n",
       "4029                              -1.0                           -1.0   \n",
       "\n",
       "      species_Diploria labyrinthiformis  ...  species_Porites australiensus  \\\n",
       "0                                  -1.0  ...                           -1.0   \n",
       "1                                  -1.0  ...                           -1.0   \n",
       "2                                  -1.0  ...                           -1.0   \n",
       "3                                  -1.0  ...                           -1.0   \n",
       "4                                  -1.0  ...                           -1.0   \n",
       "...                                 ...  ...                            ...   \n",
       "4025                               -1.0  ...                           -1.0   \n",
       "4026                               -1.0  ...                           -1.0   \n",
       "4027                               -1.0  ...                           -1.0   \n",
       "4028                               -1.0  ...                           -1.0   \n",
       "4029                               -1.0  ...                           -1.0   \n",
       "\n",
       "      species_Porites lobata  species_Porites lutea  species_Porites solida  \\\n",
       "0                       -1.0                   -1.0                    -1.0   \n",
       "1                       -1.0                   -1.0                    -1.0   \n",
       "2                       -1.0                   -1.0                    -1.0   \n",
       "3                       -1.0                   -1.0                    -1.0   \n",
       "4                       -1.0                   -1.0                    -1.0   \n",
       "...                      ...                    ...                     ...   \n",
       "4025                    -1.0                    1.0                    -1.0   \n",
       "4026                    -1.0                    1.0                    -1.0   \n",
       "4027                    -1.0                    1.0                    -1.0   \n",
       "4028                    -1.0                    1.0                    -1.0   \n",
       "4029                    -1.0                    1.0                    -1.0   \n",
       "\n",
       "      species_Porites sp.  species_Siderastrea radians  \\\n",
       "0                    -1.0                         -1.0   \n",
       "1                    -1.0                         -1.0   \n",
       "2                    -1.0                         -1.0   \n",
       "3                    -1.0                         -1.0   \n",
       "4                    -1.0                         -1.0   \n",
       "...                   ...                          ...   \n",
       "4025                 -1.0                         -1.0   \n",
       "4026                 -1.0                         -1.0   \n",
       "4027                 -1.0                         -1.0   \n",
       "4028                 -1.0                         -1.0   \n",
       "4029                 -1.0                         -1.0   \n",
       "\n",
       "      species_Siderastrea spp.  species_Sidereastrea siderea  \\\n",
       "0                         -1.0                          -1.0   \n",
       "1                         -1.0                          -1.0   \n",
       "2                         -1.0                          -1.0   \n",
       "3                         -1.0                          -1.0   \n",
       "4                         -1.0                          -1.0   \n",
       "...                        ...                           ...   \n",
       "4025                      -1.0                          -1.0   \n",
       "4026                      -1.0                          -1.0   \n",
       "4027                      -1.0                          -1.0   \n",
       "4028                      -1.0                          -1.0   \n",
       "4029                      -1.0                          -1.0   \n",
       "\n",
       "      species_Solenastrea bournoni      diff  \n",
       "0                             -1.0 -0.398363  \n",
       "1                             -1.0 -0.502562  \n",
       "2                             -1.0 -0.827680  \n",
       "3                             -1.0 -0.353592  \n",
       "4                             -1.0 -0.124839  \n",
       "...                            ...       ...  \n",
       "4025                          -1.0 -1.457229  \n",
       "4026                          -1.0 -1.502298  \n",
       "4027                          -1.0 -1.239990  \n",
       "4028                          -1.0 -0.895785  \n",
       "4029                          -1.0 -1.165122  \n",
       "\n",
       "[4030 rows x 28 columns]"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "past_data_with_preds = actual_v_pred_past.join(X_past, how = \"inner\")\n",
    "past_data_with_preds['diff'] = actual_v_pred_past['temp_anom'] - actual_v_pred_past['pred']\n",
    "past_data_with_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([<matplotlib.lines.Line2D at 0x7fea40bd0bb0>],\n",
       " [<matplotlib.lines.Line2D at 0x7fea41879a30>])"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(past_data_with_preds['year'], past_data_with_preds['temp_anom']), plt.plot(past_data_with_preds['year'], past_data_with_preds['pred'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-227-2651521bf6e3>:2: FutureWarning: Indexing with multiple keys (implicitly converted to a tuple of keys) will be deprecated, use a list instead.\n",
      "  annual_compare = past_data_with_preds.groupby('year', as_index=False)['temp_anom', 'pred'].mean()\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<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>year</th>\n",
       "      <th>temp_anom</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-0.580696</td>\n",
       "      <td>-0.319498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.974684</td>\n",
       "      <td>-0.437284</td>\n",
       "      <td>-0.322879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.949367</td>\n",
       "      <td>-0.448660</td>\n",
       "      <td>-0.311360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.924051</td>\n",
       "      <td>-0.535872</td>\n",
       "      <td>-0.312129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.898734</td>\n",
       "      <td>-0.577859</td>\n",
       "      <td>-0.296652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>0.898734</td>\n",
       "      <td>-0.628892</td>\n",
       "      <td>0.479466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>0.924051</td>\n",
       "      <td>-0.655931</td>\n",
       "      <td>0.500699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>0.949367</td>\n",
       "      <td>-0.272213</td>\n",
       "      <td>0.523098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>0.974684</td>\n",
       "      <td>-0.251039</td>\n",
       "      <td>0.544479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.227458</td>\n",
       "      <td>0.569019</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>80 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        year  temp_anom      pred\n",
       "0  -1.000000  -0.580696 -0.319498\n",
       "1  -0.974684  -0.437284 -0.322879\n",
       "2  -0.949367  -0.448660 -0.311360\n",
       "3  -0.924051  -0.535872 -0.312129\n",
       "4  -0.898734  -0.577859 -0.296652\n",
       "..       ...        ...       ...\n",
       "75  0.898734  -0.628892  0.479466\n",
       "76  0.924051  -0.655931  0.500699\n",
       "77  0.949367  -0.272213  0.523098\n",
       "78  0.974684  -0.251039  0.544479\n",
       "79  1.000000  -0.227458  0.569019\n",
       "\n",
       "[80 rows x 3 columns]"
      ]
     },
     "execution_count": 227,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "annual_temp_and_preds = past_data_with_preds[['temp_anom', 'pred']].join(X, how=\"inner\")\n",
    "annual_compare = past_data_with_preds.groupby('year', as_index=False)['temp_anom', 'pred'].mean()\n",
    "annual_compare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([<matplotlib.lines.Line2D at 0x7fea409758e0>],\n",
       " [<matplotlib.lines.Line2D at 0x7fea4095a8b0>])"
      ]
     },
     "execution_count": 228,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(annual_compare['year'], annual_compare['temp_anom']), plt.plot(annual_compare['year'], annual_compare['pred'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our network very much over-predicted the `temp_anom`. Probalby because it over-fit to the traiign data which had higher temp anomalies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>year</th>\n",
       "      <th>temp_anom</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1880</td>\n",
       "      <td>-0.580696</td>\n",
       "      <td>-0.319498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1881</td>\n",
       "      <td>-0.437284</td>\n",
       "      <td>-0.322879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1882</td>\n",
       "      <td>-0.448660</td>\n",
       "      <td>-0.311360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1883</td>\n",
       "      <td>-0.535872</td>\n",
       "      <td>-0.312129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1884</td>\n",
       "      <td>-0.577859</td>\n",
       "      <td>-0.296652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>1955</td>\n",
       "      <td>-0.628892</td>\n",
       "      <td>0.479466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>1956</td>\n",
       "      <td>-0.655931</td>\n",
       "      <td>0.500699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>1957</td>\n",
       "      <td>-0.272213</td>\n",
       "      <td>0.523098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>1958</td>\n",
       "      <td>-0.251039</td>\n",
       "      <td>0.544479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>1959</td>\n",
       "      <td>-0.227458</td>\n",
       "      <td>0.569019</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>80 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    year  temp_anom      pred\n",
       "0   1880  -0.580696 -0.319498\n",
       "1   1881  -0.437284 -0.322879\n",
       "2   1882  -0.448660 -0.311360\n",
       "3   1883  -0.535872 -0.312129\n",
       "4   1884  -0.577859 -0.296652\n",
       "..   ...        ...       ...\n",
       "75  1955  -0.628892  0.479466\n",
       "76  1956  -0.655931  0.500699\n",
       "77  1957  -0.272213  0.523098\n",
       "78  1958  -0.251039  0.544479\n",
       "79  1959  -0.227458  0.569019\n",
       "\n",
       "[80 rows x 3 columns]"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "annual_compare['year'] = np.arange(1880, 1960, 1)\n",
    "annual_compare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [],
   "source": [
    "annual_compare.to_csv('compare_1880-1960.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "da0f4b2f549526c8eb110fe99ce30953b583bbcf13bb8926e556c8ad6926646c"
  },
  "kernelspec": {
   "display_name": "Python 3.8.5 64-bit ('base': conda)",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}