Stats for ML - Mini Project Solutions - Priyadharshini K.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "5c9a3cb0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# type your code here\n",
    "# import 'pandas' \n",
    "import pandas as pd \n",
    "\n",
    "# import 'numpy' \n",
    "import numpy as np\n",
    "\n",
    "# import subpackage of matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# import 'seaborn'\n",
    "import seaborn as sns\n",
    "\n",
    "# to suppress warnings \n",
    "from warnings import filterwarnings\n",
    "filterwarnings('ignore')\n",
    "\n",
    "# import 'random' to generate random sample\n",
    "import random\n",
    "\n",
    "# import statistics to perform statistical computation  \n",
    "import statistics\n",
    "\n",
    "# import 'stats' package from scipy library\n",
    "from scipy import stats\n",
    "\n",
    "# import a library to perform Z-test\n",
    "from statsmodels.stats import weightstats as stests\n",
    "\n",
    "# to test the normality \n",
    "from scipy.stats import shapiro\n",
    "\n",
    "# import the function to calculate the power of test\n",
    "from statsmodels.stats import power"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a75303c",
   "metadata": {},
   "source": [
    "#### 1. Perform the following statistical analysis on the fifa world cup dataset. The dataset consists of player information which includes height, weight overall rating, potential etc. Data Cleaning and basic EDA process needs to be performed. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 335,
   "id": "301a42f6",
   "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>Name</th>\n",
       "      <th>Age</th>\n",
       "      <th>Nationality</th>\n",
       "      <th>Overall_rating</th>\n",
       "      <th>Potential</th>\n",
       "      <th>Club</th>\n",
       "      <th>Preferred Foot</th>\n",
       "      <th>International Reputation</th>\n",
       "      <th>Weak Foot</th>\n",
       "      <th>Skill Moves</th>\n",
       "      <th>Work Rate</th>\n",
       "      <th>Position</th>\n",
       "      <th>Jersey Number</th>\n",
       "      <th>Height</th>\n",
       "      <th>Weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>L. Messi</td>\n",
       "      <td>31</td>\n",
       "      <td>Argentina</td>\n",
       "      <td>94</td>\n",
       "      <td>94</td>\n",
       "      <td>FC Barcelona</td>\n",
       "      <td>Left</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Medium/ Medium</td>\n",
       "      <td>RF</td>\n",
       "      <td>10</td>\n",
       "      <td>5.583333</td>\n",
       "      <td>159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Cristiano Ronaldo</td>\n",
       "      <td>33</td>\n",
       "      <td>Portugal</td>\n",
       "      <td>94</td>\n",
       "      <td>94</td>\n",
       "      <td>Juventus</td>\n",
       "      <td>Right</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>High/ Low</td>\n",
       "      <td>ST</td>\n",
       "      <td>7</td>\n",
       "      <td>6.166667</td>\n",
       "      <td>183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Neymar Jr</td>\n",
       "      <td>26</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>92</td>\n",
       "      <td>93</td>\n",
       "      <td>Paris Saint-Germain</td>\n",
       "      <td>Right</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>High/ Medium</td>\n",
       "      <td>LW</td>\n",
       "      <td>10</td>\n",
       "      <td>5.750000</td>\n",
       "      <td>150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>De Gea</td>\n",
       "      <td>27</td>\n",
       "      <td>Spain</td>\n",
       "      <td>91</td>\n",
       "      <td>93</td>\n",
       "      <td>Manchester United</td>\n",
       "      <td>Right</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>Medium/ Medium</td>\n",
       "      <td>GK</td>\n",
       "      <td>1</td>\n",
       "      <td>6.333333</td>\n",
       "      <td>168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>K. De Bruyne</td>\n",
       "      <td>27</td>\n",
       "      <td>Belgium</td>\n",
       "      <td>91</td>\n",
       "      <td>92</td>\n",
       "      <td>Manchester City</td>\n",
       "      <td>Right</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>High/ High</td>\n",
       "      <td>RCM</td>\n",
       "      <td>7</td>\n",
       "      <td>5.083333</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Name  Age Nationality  Overall_rating  Potential  \\\n",
       "0           L. Messi   31   Argentina              94         94   \n",
       "1  Cristiano Ronaldo   33    Portugal              94         94   \n",
       "2          Neymar Jr   26      Brazil              92         93   \n",
       "3             De Gea   27       Spain              91         93   \n",
       "4       K. De Bruyne   27     Belgium              91         92   \n",
       "\n",
       "                  Club Preferred Foot  International Reputation  Weak Foot  \\\n",
       "0         FC Barcelona           Left                         5          4   \n",
       "1             Juventus          Right                         5          4   \n",
       "2  Paris Saint-Germain          Right                         5          5   \n",
       "3    Manchester United          Right                         4          3   \n",
       "4      Manchester City          Right                         4          5   \n",
       "\n",
       "   Skill Moves       Work Rate Position  Jersey Number    Height  Weight  \n",
       "0            4  Medium/ Medium       RF             10  5.583333     159  \n",
       "1            5       High/ Low       ST              7  6.166667     183  \n",
       "2            5    High/ Medium       LW             10  5.750000     150  \n",
       "3            1  Medium/ Medium       GK              1  6.333333     168  \n",
       "4            4      High/ High      RCM              7  5.083333     154  "
      ]
     },
     "execution_count": 335,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fifa_df = pd.read_csv('fifa_stats.csv')\n",
    "fifa_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 336,
   "id": "afc6782e",
   "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>total_missing</th>\n",
       "      <th>perc_missing</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Name</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Nationality</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Overall_rating</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Potential</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Club</th>\n",
       "      <td>19</td>\n",
       "      <td>0.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Preferred Foot</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>International Reputation</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Weak Foot</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Skill Moves</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Work Rate</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Position</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Jersey Number</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Height</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Weight</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          total_missing  perc_missing\n",
       "Name                                  0          0.00\n",
       "Age                                   0          0.00\n",
       "Nationality                           0          0.00\n",
       "Overall_rating                        0          0.00\n",
       "Potential                             0          0.00\n",
       "Club                                 19          0.95\n",
       "Preferred Foot                        0          0.00\n",
       "International Reputation              0          0.00\n",
       "Weak Foot                             0          0.00\n",
       "Skill Moves                           0          0.00\n",
       "Work Rate                             0          0.00\n",
       "Position                              0          0.00\n",
       "Jersey Number                         0          0.00\n",
       "Height                                0          0.00\n",
       "Weight                                0          0.00"
      ]
     },
     "execution_count": 336,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking missing values in each column\n",
    "\n",
    "missing_data = pd.DataFrame({'total_missing': fifa_df.isnull().sum(), \n",
    "                             'perc_missing': (fifa_df.isnull().sum() / len(fifa_df)) * 100})\n",
    "\n",
    "missing_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 337,
   "id": "ee37c7d3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The percentage of data reduction is:  0.95\n",
      "The percentage of data remaining after dropping rows with missing values is: 99.05\n"
     ]
    }
   ],
   "source": [
    "# Dropping rows with missing data as the total percentage of missing data is less than 10% \n",
    "\n",
    "df = fifa_df.dropna()\n",
    "len(df) \n",
    "print(\"The percentage of data reduction is: \",((len(fifa_df) - len(df)) / len(fifa_df)) * 100)\n",
    "print(\"The percentage of data remaining after dropping rows with missing values is:\", 100 - ((len(fifa_df) - len(df)) / len(fifa_df)) * 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 338,
   "id": "81212b7b",
   "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>Name</th>\n",
       "      <th>Age</th>\n",
       "      <th>Nationality</th>\n",
       "      <th>Overall_rating</th>\n",
       "      <th>Potential</th>\n",
       "      <th>Club</th>\n",
       "      <th>Preferred Foot</th>\n",
       "      <th>International Reputation</th>\n",
       "      <th>Weak Foot</th>\n",
       "      <th>Skill Moves</th>\n",
       "      <th>Work Rate</th>\n",
       "      <th>Position</th>\n",
       "      <th>Jersey Number</th>\n",
       "      <th>Height</th>\n",
       "      <th>Weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>L. Messi</td>\n",
       "      <td>31</td>\n",
       "      <td>Argentina</td>\n",
       "      <td>94</td>\n",
       "      <td>94</td>\n",
       "      <td>FC Barcelona</td>\n",
       "      <td>Left</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>Medium/ Medium</td>\n",
       "      <td>RF</td>\n",
       "      <td>10</td>\n",
       "      <td>5.583333</td>\n",
       "      <td>159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Cristiano Ronaldo</td>\n",
       "      <td>33</td>\n",
       "      <td>Portugal</td>\n",
       "      <td>94</td>\n",
       "      <td>94</td>\n",
       "      <td>Juventus</td>\n",
       "      <td>Right</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>High/ Low</td>\n",
       "      <td>ST</td>\n",
       "      <td>7</td>\n",
       "      <td>6.166667</td>\n",
       "      <td>183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Neymar Jr</td>\n",
       "      <td>26</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>92</td>\n",
       "      <td>93</td>\n",
       "      <td>Paris Saint-Germain</td>\n",
       "      <td>Right</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>High/ Medium</td>\n",
       "      <td>LW</td>\n",
       "      <td>10</td>\n",
       "      <td>5.750000</td>\n",
       "      <td>150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>De Gea</td>\n",
       "      <td>27</td>\n",
       "      <td>Spain</td>\n",
       "      <td>91</td>\n",
       "      <td>93</td>\n",
       "      <td>Manchester United</td>\n",
       "      <td>Right</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>Medium/ Medium</td>\n",
       "      <td>GK</td>\n",
       "      <td>1</td>\n",
       "      <td>6.333333</td>\n",
       "      <td>168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>K. De Bruyne</td>\n",
       "      <td>27</td>\n",
       "      <td>Belgium</td>\n",
       "      <td>91</td>\n",
       "      <td>92</td>\n",
       "      <td>Manchester City</td>\n",
       "      <td>Right</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>High/ High</td>\n",
       "      <td>RCM</td>\n",
       "      <td>7</td>\n",
       "      <td>5.083333</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Name  Age Nationality  Overall_rating  Potential  \\\n",
       "0           L. Messi   31   Argentina              94         94   \n",
       "1  Cristiano Ronaldo   33    Portugal              94         94   \n",
       "2          Neymar Jr   26      Brazil              92         93   \n",
       "3             De Gea   27       Spain              91         93   \n",
       "4       K. De Bruyne   27     Belgium              91         92   \n",
       "\n",
       "                  Club Preferred Foot  International Reputation  Weak Foot  \\\n",
       "0         FC Barcelona           Left                         5          4   \n",
       "1             Juventus          Right                         5          4   \n",
       "2  Paris Saint-Germain          Right                         5          5   \n",
       "3    Manchester United          Right                         4          3   \n",
       "4      Manchester City          Right                         4          5   \n",
       "\n",
       "   Skill Moves       Work Rate Position  Jersey Number    Height  Weight  \n",
       "0            4  Medium/ Medium       RF             10  5.583333     159  \n",
       "1            5       High/ Low       ST              7  6.166667     183  \n",
       "2            5    High/ Medium       LW             10  5.750000     150  \n",
       "3            1  Medium/ Medium       GK              1  6.333333     168  \n",
       "4            4      High/ High      RCM              7  5.083333     154  "
      ]
     },
     "execution_count": 338,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8726d2b",
   "metadata": {},
   "source": [
    "##### a) Assuming age as population, perform the central limit theorem with sample size of 200. Find the standard error and the mean of sampling distribution. Compare the results with population mean and standard deviation. Explain in non-technical terms why the estimated standard error of a sample mean tends to decrease with an increase in sample size\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "id": "e1957aae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(df['Age'])\n",
    "plt.show()                       # slightly right skewed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "id": "8e0d5b5e",
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_size= 200\n",
    "no_of_samples= 3000 \n",
    "sample_mean_list= [] \n",
    "\n",
    "for i in range(no_of_samples): \n",
    "    sample= np.random.choice(df['Age'], size= sample_size, replace= True) \n",
    "    sample_mean= np.mean(sample) \n",
    "    sample_mean_list.append(sample_mean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "id": "38efba7e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(sample_mean_list)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "id": "5ebdf6d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Standard error: 0.07033704467533643\n",
      "Mean of sampling distribution: 27.216426666666667\n",
      "\n",
      "Population mean: 27.215547703180214\n",
      "Population standard deviation: 3.8525185996930396\n"
     ]
    }
   ],
   "source": [
    "sigma = df['Age'].std() \n",
    "n = 3000 \n",
    "\n",
    "print(\"Standard error:\", sigma/(n**0.5))\n",
    "print(\"Mean of sampling distribution:\", np.mean(sample_mean_list))\n",
    "print() \n",
    "print(\"Population mean:\", df['Age'].mean())\n",
    "print(\"Population standard deviation:\", df['Age'].std())"
   ]
  },
  {
   "cell_type": "raw",
   "id": "44072928",
   "metadata": {},
   "source": [
    "Inference: \n",
    "    \n",
    "-> Mean of Sampling distribution is approximately equal to the mean of the population. \n",
    "-> Standard deviation of the Sampling distribution is inversely proportional to Sample Size.\n",
    "\n",
    "-> If there are two samples (s1, s2) extracted from a population of 150 data points (mu): \n",
    "    s1 - sample size of 10 data points\n",
    "    s2 - sample size of 100 data points. \n",
    "    \n",
    "    -> If the mean of sample s1 and sample s2 is calculated, based on respective sample size, it is sure and evident that mean of sample s1 (x1_bar) will be less than when compared to mean of sample s2 (x2_bar). \n",
    "    -> Standard error is the difference of sample mean with respect to actual population mean.\n",
    "    -> Here, the mean of sample s2 (x2_bar) will be closer to the population mean (mu) thus having very low standard error as the sample size of sample s2 is larger compared to sample size of sample s2. \n",
    "    -> Hence, sample s2 will reflect most of the properties of population as it's size is closer to population length. \n",
    "    \n",
    "-> Thus, we could say that the estimated standard error of a sample mean tends to decrease with an increase in sample size."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71b69ee9",
   "metadata": {},
   "source": [
    "##### b) Is there any evidence that the players overall rating is greater than 80?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "id": "93f5cba5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shapiro test on Overall rating sample:\n",
      " ShapiroResult(statistic=0.8532959818840027, pvalue=1.721150444000814e-39) \n",
      "\n",
      "\n",
      "Transformed values: [0.07016879 0.07016879 0.07016879 ... 0.07016879 0.07016879 0.07016879]\n",
      "Lambda/power value: -14.251349344860774 \n",
      "\n",
      "\n",
      "After transforming the claimed population mean value, mu: 7.557875971848567e-28 \n",
      "\n",
      "\n",
      "Shapiro test on transformed Overall Rating sample:\n",
      " ShapiroResult(statistic=1.0, pvalue=1.0)\n"
     ]
    }
   ],
   "source": [
    "# Test of Normality of data: \n",
    "\n",
    "# Hypothesis formation for Shapiro test:\n",
    "# Ho: skew = 0 (Data is normal) \n",
    "# Ha: skew != 0 (Data is not normal) \n",
    "print(\"Shapiro test on Overall rating sample:\\n\",stats.shapiro(df['Overall_rating']),'\\n\\n')\n",
    "\n",
    "# Inference: \n",
    "# pValue (0) < 0.05 significance value. \n",
    "# Hence, Ha can be selected and so data is not normal\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Hence, transformation process can be applied on 'Overall Rating' feature to bring the distribution close to normal. \n",
    "# BoxCox transformation: \n",
    "tr, lam = stats.boxcox(df['Overall_rating'])\n",
    "print(\"Transformed values:\", tr)\n",
    "print(\"Lambda/power value:\", lam,'\\n\\n') \n",
    "\n",
    "# Since, we have the transformed values, the claimed mean (mu = 80) is transformed based on the same lambda value.\n",
    "mu = np.power(80, lam)\n",
    "print(\"After transforming the claimed population mean value, mu:\", mu,'\\n\\n') \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Hypothesis formation for Shapiro test after transformation:\n",
    "# Ho: skew = 0 (Data is normal) \n",
    "# Ha: skew != 0 (Data is not normal) \n",
    "print(\"Shapiro test on transformed Overall Rating sample:\\n\",stats.shapiro(tr))\n",
    "\n",
    "# Inference: \n",
    "# pValue (1.0) > 0.05 significance value. \n",
    "# Hence, Ho can be selected and so now data is normal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "id": "a88991ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test statistics value, t:  inf\n",
      "Probability value, pValue:  0.0\n"
     ]
    }
   ],
   "source": [
    "# Hypothesis Formation for One Sample t test:\n",
    "# Ho: mu = 0\n",
    "# Ha: mu != 0 \n",
    "\n",
    "\n",
    "\n",
    "# Choosing the right test:\n",
    "# 1) Data is normal.\n",
    "# 2) Population standard deviation is not known.\n",
    "\n",
    "# Hence, One sample t test can be applied. \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Test Statistics:\n",
    "x_bar = np.mean(tr) \n",
    "s = np.std(tr, ddof= 1)\n",
    "n = len(tr) \n",
    "mu = 0 \n",
    "\n",
    "t = (x_bar - mu) / (s/n**0.5)\n",
    "print(\"Test statistics value, t: \", t)\n",
    "\n",
    "\n",
    "\n",
    "# pValue\n",
    "pval = stats.t.sf(abs(t), df = n-1) * 2 \n",
    "print(\"Probability value, pValue: \", pval)"
   ]
  },
  {
   "cell_type": "raw",
   "id": "fab73eba",
   "metadata": {},
   "source": [
    "Inference: \n",
    "    \n",
    "-> pValue 0 < 0.05 Significance level. \n",
    "-> As the pValue is less than significance level, Ho is rejected and Ha can be selected (mu != 0).\n",
    "-> So we infer that the average overall rating is not equal to 80. \n",
    "-> Test statistics (t) value is positive and so it depicts that the average overall rating is greater than 80 as claimed."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0640bb22",
   "metadata": {},
   "source": [
    "##### c) Perform the statistical test to identify whether preferred foot has any impact on potential. Also, perform the graphical analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "id": "1f9475ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Right    1519\n",
       "Left      462\n",
       "Name: Preferred Foot, dtype: int64"
      ]
     },
     "execution_count": 293,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Preferred Foot'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "id": "a4cb449c",
   "metadata": {},
   "outputs": [],
   "source": [
    "rfoot_pot = df[df['Preferred Foot'] == 'Right']['Potential'] \n",
    "lfoot_pot = df[df['Preferred Foot'] == 'Left']['Potential'] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "id": "f5e95154",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shapiro test on rfoot_pot sample:\n",
      " ShapiroResult(statistic=0.9373535513877869, pvalue=8.425229710834448e-25)\n",
      "Shapiro test on lfoot_pot sample:\n",
      " ShapiroResult(statistic=0.9437844157218933, pvalue=3.187897645515192e-12) \n",
      "\n",
      "\n",
      "Shapiro test on transformed rfoot_pot sample:\n",
      " ShapiroResult(statistic=1.0, pvalue=1.0)\n",
      "Shapiro test on transformed lfoot_pot sample:\n",
      " ShapiroResult(statistic=1.0, pvalue=1.0)\n"
     ]
    }
   ],
   "source": [
    "# Test of Normality of data: \n",
    "\n",
    "# Hypothesis formation for Shapiro test:\n",
    "# Ho: skew = 0 (Data is normal) \n",
    "# Ha: skew != 0 (Data is not normal) \n",
    "print(\"Shapiro test on rfoot_pot sample:\\n\",stats.shapiro(rfoot_pot))\n",
    "print(\"Shapiro test on lfoot_pot sample:\\n\",stats.shapiro(lfoot_pot),'\\n\\n')\n",
    "\n",
    "# Inference: \n",
    "# pValue of both samples (0, 0) < 0.05 significance value. \n",
    "# Hence, Ha can be selected and so data is not normal \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Hence, transformation process can be applied on both samples to bring the distribution close to normal. \n",
    "# BoxCox transformation: \n",
    "tr1, lam1 = stats.boxcox(rfoot_pot)\n",
    "tr2, lam2 = stats.boxcox(lfoot_pot)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Hypothesis formation for Shapiro test after transformation:\n",
    "# Ho: skew = 0 (Data is normal) \n",
    "# Ha: skew != 0 (Data is not normal) \n",
    "print(\"Shapiro test on transformed rfoot_pot sample:\\n\",stats.shapiro(tr1))\n",
    "print(\"Shapiro test on transformed lfoot_pot sample:\\n\",stats.shapiro(tr2))\n",
    "\n",
    "\n",
    "# Inference: \n",
    "# pValue (1.0) > 0.05 significance value. \n",
    "# Hence, Ho can be selected and so now data is normal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "id": "9b7569db",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test statistics value, t:  -29241418708.931744\n",
      "Probability value, pValue:  0.0\n"
     ]
    }
   ],
   "source": [
    "# Hypothesis Formation: \n",
    "# Ho: mu1 (average potential of right foot prefered players) = mu2 (average potential of left foot prefered players) \n",
    "# Ha: mu1 (average potential of right foot prefered players) != mu2 (average potential of left foot prefered players)\n",
    "\n",
    "# Ho: mu1 - m2 = 0\n",
    "# Ha: mu1 - mu2 != 0\n",
    "\n",
    "\n",
    "\n",
    "# Choosing the right test \n",
    "# 1) Data is normal \n",
    "# 2) Population standard deviation is not known \n",
    "# 3) Data is unpaired\n",
    "\n",
    "# Hence, Two sample unpaired  t test can be applied. \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Test Statistics: \n",
    "x1_bar = np.mean(tr1) \n",
    "x2_bar = np.mean(tr2)\n",
    "s1 = np.std(tr1, ddof=1) \n",
    "s2 = np.std(tr2, ddof=1)\n",
    "n1 = len(tr1) \n",
    "n2 = len(tr2) \n",
    "df1 = n1 - 1\n",
    "df2 = n2 - 1 \n",
    "df_tot = df1 + df2 \n",
    "\n",
    "\n",
    "sp_var = ((df1 * s1**2) + (df2 * s2**2)) / df_tot\n",
    "\n",
    "num = (x1_bar - x2_bar) - 0 \n",
    "den = np.sqrt(sp_var * (1/n1 + 1/n2))\n",
    "\n",
    "t = num / den\n",
    "print(\"Test statistics value, t: \", t)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# pValue:\n",
    "pval = stats.t.sf(abs(t), df = df_tot) * 2 \n",
    "print(\"Probability value, pValue: \", pval)"
   ]
  },
  {
   "cell_type": "raw",
   "id": "cec86730",
   "metadata": {},
   "source": [
    "Inference: \n",
    "\n",
    "-> pValue 0 < 0.05 Significance Value\n",
    "-> As the pValue is less than significance level, Ho is rejected and Ha can be selected. \n",
    "-> Test statistics (t) is negative and so average potential of right foot preferred players is less than the average potential of right foot preferred players. \n",
    "-> Hence we can claim that preferred foot does have an impact on the potential of the players."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4670888a",
   "metadata": {},
   "source": [
    "##### d) Perform the statistical test to identify the relation between the overall rating and potential. [Assume both are continuous data points]."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "id": "b7f0984d",
   "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>Overall_rating</th>\n",
       "      <th>Potential</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Overall_rating</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.732086</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Potential</th>\n",
       "      <td>0.732086</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Overall_rating  Potential\n",
       "Overall_rating        1.000000   0.732086\n",
       "Potential             0.732086   1.000000"
      ]
     },
     "execution_count": 309,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[['Overall_rating','Potential']].corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "id": "8076337a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shapiro test on transformed Overall Rating sample:\n",
      " ShapiroResult(statistic=1.0, pvalue=1.0) \n",
      "\n",
      "Shapiro test on transformed Potential sample:\n",
      " ShapiroResult(statistic=1.0, pvalue=1.0)\n"
     ]
    }
   ],
   "source": [
    "# Based on the analysis, we could see that the data of Overall_rating and Potential do not follow normality. \n",
    "# Hence we transform the data to bring the data closer to normal distribution using BoxCox transformation.\n",
    "\n",
    "# BoxCox transformation: \n",
    "tr1, lam1 = stats.boxcox(df['Overall_rating'])\n",
    "tr2, lam2 = stats.boxcox(df['Potential'])\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Hypothesis formation for Shapiro test after transformation:\n",
    "# Ho: skew = 0 (Data is normal) \n",
    "# Ha: skew != 0 (Data is not normal) \n",
    "print(\"Shapiro test on transformed Overall Rating sample:\\n\",stats.shapiro(tr1),'\\n')\n",
    "print(\"Shapiro test on transformed Potential sample:\\n\",stats.shapiro(tr2))\n",
    "\n",
    "# Inference: \n",
    "# pValue (1.0) > 0.05 significance value. \n",
    "# Hence, Ho can be selected and so now data is normal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "id": "715c36b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(nan, nan)"
      ]
     },
     "execution_count": 313,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hypothesis formation for Pearson r Correlation: \n",
    "# Ho : corr = 0 (no relation)\n",
    "# Ha : corr != 0 (relation exists) \n",
    "\n",
    "\n",
    "# Choosing the right test: \n",
    "# 1) Data should follow normal distribution. \n",
    "\n",
    "stats.pearsonr(np.array(tr1), np.array(tr2))"
   ]
  },
  {
   "cell_type": "raw",
   "id": "f45b9638",
   "metadata": {},
   "source": [
    "Inference: \n",
    "\n",
    "-> The NaN, in this case, is interpreted as no correlation between the two variables. \n",
    "-> The correlation describes how much one variable changes as the other variable changes. That requires both variables to change.\n",
    "-> Reason for this is that one of the input must have identical values. When each value of an input is identical, the result is a vector of 0s. When it is divided by zero, the result will be NaN. \n",
    "-> Hence, there is no relation between the Overall Rating and Potential of the players"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f6a7e95",
   "metadata": {},
   "source": [
    "##### e) Does skill move have any effect in overall rating. Perform the graphical analysis and the statistical analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "id": "577960f3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3    818\n",
       "4    540\n",
       "2    392\n",
       "1    192\n",
       "5     39\n",
       "Name: Skill Moves, dtype: int64"
      ]
     },
     "execution_count": 320,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Skill Moves'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "id": "7a0252e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Based on previous analysis, since Overall_rating does not follow normality, BoxCox transformation is applied.\n",
    "\n",
    "tr, lam = stats.boxcox(df['Overall_rating'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "id": "9126826d",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['tr_Overall_Rating'] = pd.Series(tr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "id": "b1ae9179",
   "metadata": {},
   "outputs": [],
   "source": [
    "sk1_rating = df[df['Skill Moves'] == 1]['tr_Overall_Rating']\n",
    "sk2_rating = df[df['Skill Moves'] == 2]['tr_Overall_Rating']\n",
    "sk3_rating = df[df['Skill Moves'] == 3]['tr_Overall_Rating']\n",
    "sk4_rating = df[df['Skill Moves'] == 4]['tr_Overall_Rating']\n",
    "sk5_rating = df[df['Skill Moves'] == 5]['tr_Overall_Rating']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "id": "69f10158",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shapiro test on sk1_rating sample:\n",
      " ShapiroResult(statistic=nan, pvalue=1.0)\n",
      "Shapiro test on sk2_rating sample:\n",
      " ShapiroResult(statistic=nan, pvalue=1.0)\n",
      "Shapiro test on sk3_rating sample:\n",
      " ShapiroResult(statistic=nan, pvalue=1.0)\n",
      "Shapiro test on sk4_rating sample:\n",
      " ShapiroResult(statistic=nan, pvalue=1.0)\n",
      "Shapiro test on sk5_rating sample:\n",
      " ShapiroResult(statistic=nan, pvalue=1.0)\n",
      "\n",
      "Test of variance for all five skill moves samples:\n",
      " LeveneResult(statistic=nan, pvalue=nan)\n"
     ]
    }
   ],
   "source": [
    "# Test of Normality of data: \n",
    "\n",
    "# Hypothesis formation for Shapiro test on trasformed Overall_rating sample:\n",
    "# Ho: skew = 0 (Data is normal) \n",
    "# Ha: skew != 0 (Data is not normal) \n",
    "print(\"Shapiro test on sk1_rating sample:\\n\",stats.shapiro(sk1_rating))\n",
    "print(\"Shapiro test on sk2_rating sample:\\n\",stats.shapiro(sk2_rating))\n",
    "print(\"Shapiro test on sk3_rating sample:\\n\",stats.shapiro(sk3_rating))\n",
    "print(\"Shapiro test on sk4_rating sample:\\n\",stats.shapiro(sk4_rating))\n",
    "print(\"Shapiro test on sk5_rating sample:\\n\",stats.shapiro(sk5_rating)) \n",
    "\n",
    "# Inference: \n",
    "# pValue (1.0) > 0.05 significance value \n",
    "# Hence, Ho can be selected and so now data is normal \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Test for equality of variance: \n",
    "\n",
    "# Hypothesis formation for Levene test: \n",
    "# Ho: All variances are equal\n",
    "# Ha: Atleast one variance is different \n",
    "print()\n",
    "print(\"Test of variance for all five skill moves samples:\\n\", stats.levene(sk1_rating, sk2_rating, sk3_rating, sk4_rating, sk5_rating)) \n",
    "\n",
    "# Inference: \n",
    "# pValue of Levene test of overall rating for each skill move is returned as 'Nan'.\n",
    "# Hence, Ha is selected and so atleast one variance is different."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "id": "21bcee82",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "F_onewayResult(statistic=nan, pvalue=nan)"
      ]
     },
     "execution_count": 361,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hypothesis for ANOVA: \n",
    "# Ho: All means are equal (no relation) \n",
    "# Ha: Atleast one mean is different (relation exists) \n",
    "\n",
    "\n",
    "# Choosing the right test: \n",
    "# 1) Data is normal \n",
    "# 2) All variances are different. \n",
    "\n",
    "# Hence, One Way Anova test cannot be applied."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3cec084",
   "metadata": {},
   "source": [
    "#### 2. Perform the following statistical analysis on the mobile network dataset given. The given data set contains information about the internet, calls, messages of various network. Data Cleaning and basic EDA process needs to be performed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "id": "5816ae71",
   "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>date</th>\n",
       "      <th>duration</th>\n",
       "      <th>item</th>\n",
       "      <th>month</th>\n",
       "      <th>network</th>\n",
       "      <th>network_type</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>15-10-14 6:58</td>\n",
       "      <td>221.321120</td>\n",
       "      <td>data</td>\n",
       "      <td>2014-11</td>\n",
       "      <td>data</td>\n",
       "      <td>data</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>15-10-14 6:58</td>\n",
       "      <td>81.610599</td>\n",
       "      <td>call</td>\n",
       "      <td>2014-11</td>\n",
       "      <td>Vodafone</td>\n",
       "      <td>mobile</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15-10-14 14:46</td>\n",
       "      <td>143.109111</td>\n",
       "      <td>call</td>\n",
       "      <td>2014-11</td>\n",
       "      <td>Meteor</td>\n",
       "      <td>mobile</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>15-10-14 14:48</td>\n",
       "      <td>111.888670</td>\n",
       "      <td>call</td>\n",
       "      <td>2014-11</td>\n",
       "      <td>Tesco</td>\n",
       "      <td>mobile</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15-10-14 17:27</td>\n",
       "      <td>246.695171</td>\n",
       "      <td>call</td>\n",
       "      <td>2014-11</td>\n",
       "      <td>Tesco</td>\n",
       "      <td>mobile</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             date    duration  item    month   network network_type\n",
       "0   15-10-14 6:58  221.321120  data  2014-11      data         data\n",
       "1   15-10-14 6:58   81.610599  call  2014-11  Vodafone       mobile\n",
       "2  15-10-14 14:46  143.109111  call  2014-11    Meteor       mobile\n",
       "3  15-10-14 14:48  111.888670  call  2014-11     Tesco       mobile\n",
       "4  15-10-14 17:27  246.695171  call  2014-11     Tesco       mobile"
      ]
     },
     "execution_count": 227,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "phone_df = pd.read_csv('phone.csv')\n",
    "phone_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "id": "799ca101",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 820 entries, 0 to 819\n",
      "Data columns (total 6 columns):\n",
      " #   Column        Non-Null Count  Dtype  \n",
      "---  ------        --------------  -----  \n",
      " 0   date          820 non-null    object \n",
      " 1   duration      820 non-null    float64\n",
      " 2   item          820 non-null    object \n",
      " 3   month         820 non-null    object \n",
      " 4   network       820 non-null    object \n",
      " 5   network_type  820 non-null    object \n",
      "dtypes: float64(1), object(5)\n",
      "memory usage: 38.6+ KB\n"
     ]
    }
   ],
   "source": [
    "phone_df.info()       # There are no missing values in the given dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "id": "5cd2241f",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Converting datatype of column 'date':\n",
    "\n",
    "phone_df['date'] = pd.to_datetime(phone_df['date'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "id": "fd6d0c0f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 820 entries, 0 to 819\n",
      "Data columns (total 6 columns):\n",
      " #   Column        Non-Null Count  Dtype         \n",
      "---  ------        --------------  -----         \n",
      " 0   date          820 non-null    datetime64[ns]\n",
      " 1   duration      820 non-null    float64       \n",
      " 2   item          820 non-null    object        \n",
      " 3   month         820 non-null    object        \n",
      " 4   network       820 non-null    object        \n",
      " 5   network_type  820 non-null    object        \n",
      "dtypes: datetime64[ns](1), float64(1), object(4)\n",
      "memory usage: 38.6+ KB\n"
     ]
    }
   ],
   "source": [
    "phone_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7dd506b",
   "metadata": {},
   "source": [
    "##### a) With 95% & 99% Confidence Interval find the range estimate of average duration calls made by all the network. Which of the Intervals are wider? Interpret the results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "id": "fc1f7729",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The range estimate of average duration calls made by all the networks with confidence of 95% is:\n",
      " 141.17321160796186 and 147.43665951170905\n"
     ]
    }
   ],
   "source": [
    "n= len(phone_df['duration']) \n",
    "s= phone_df['duration'].std() \n",
    "x_bar= phone_df['duration'].mean()\n",
    "\n",
    "conf_lvl= 0.95         # 95%\n",
    "\n",
    "t= stats.t.isf((1-0.95)/2, df= n-1)\n",
    "l1= x_bar - (t * (s/(n**0.5)))\n",
    "u1= x_bar + (t * (s/(n**0.5)))\n",
    "print(\"The range estimate of average duration calls made by all the networks with confidence of 95% is:\\n\",l1,\"and\",u1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "id": "feadb548",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The range estimate of average duration calls made by all the networks with confidence of 99% is:\n",
      " 140.185636073024 and 148.4242350466469\n"
     ]
    }
   ],
   "source": [
    "n= len(phone_df['duration']) \n",
    "s= phone_df['duration'].std() \n",
    "x_bar= phone_df['duration'].mean()\n",
    "\n",
    "conf_lvl= 0.99         # 95%\n",
    "\n",
    "t= stats.t.isf((1-0.99)/2, df= n-1)\n",
    "l1= x_bar - (t * (s/(n**0.5)))\n",
    "u1= x_bar + (t * (s/(n**0.5)))\n",
    "print(\"The range estimate of average duration calls made by all the networks with confidence of 99% is:\\n\",l1,\"and\",u1)"
   ]
  },
  {
   "cell_type": "raw",
   "id": "08da91b9",
   "metadata": {},
   "source": [
    "Inference: \n",
    "\n",
    "-> The range estimate of average duration calls made by all the networks with confidence interval of 99% is wider when compared to that of the range estimate of average duration calls made by all the networks with confidence interval of 95%. \n",
    "\n",
    "-> The mean interval range increases with increased confidence interval so that it is approximately attains equality with the true mean of the population."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e181df41",
   "metadata": {},
   "source": [
    "##### b) Perform statistical analysis to check whether the networks are equally distributed in the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "id": "274660b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Three        215\n",
       "Vodafone     215\n",
       "data         150\n",
       "Meteor        87\n",
       "Tesco         84\n",
       "landline      42\n",
       "voicemail     27\n",
       "Name: network, dtype: int64"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "phone_df['network'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "id": "55c21190",
   "metadata": {},
   "outputs": [],
   "source": [
    "obs_cnt = np.array([215, 215, 150, 87, 84, 42, 27])\n",
    "\n",
    "exp_cnt = np.array([len(phone_df['network'])/7 for i in phone_df['network'].unique()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "id": "9c50ab67",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Power_divergenceResult(statistic=307.409756097561, pvalue=2.112592760257685e-63)"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hypothesis formation \n",
    "\n",
    "# Ho: observed = expected \n",
    "# Ha: observed != expected  \n",
    "\n",
    "stats.chisquare(f_exp= exp_cnt, f_obs= obs_cnt)"
   ]
  },
  {
   "cell_type": "raw",
   "id": "0768a527",
   "metadata": {},
   "source": [
    "Inference: \n",
    "    \n",
    "-> Given significance level is 0.05 (by default) while the pValue derived is 0 \n",
    "-> Since the pValue is less than significance level, Ho is rejected and Ha can be selected \n",
    "-> So we infer that the observed proportion is not equal to that of expected proportion.\n",
    "-> Hence the proportion distribution of all the departments is not equal."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "602e9d69",
   "metadata": {},
   "source": [
    "##### c) Plot the distribution of duration with respect to each network in the same graph and Interpret the results from the graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "id": "e4ec306e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Three        215\n",
       "Vodafone     215\n",
       "data         150\n",
       "Meteor        87\n",
       "Tesco         84\n",
       "landline      42\n",
       "voicemail     27\n",
       "Name: network, dtype: int64"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "phone_df['network'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "id": "526033b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "dur_three = phone_df[phone_df['network'] == 'Three']['duration']\n",
    "dur_vod = phone_df[phone_df['network'] == 'Vodafone']['duration']\n",
    "dur_data = phone_df[phone_df['network'] == 'data']['duration']\n",
    "dur_met = phone_df[phone_df['network'] == 'Meteor']['duration']\n",
    "dur_tesco = phone_df[phone_df['network'] == 'Tesco']['duration']\n",
    "dur_land = phone_df[phone_df['network'] == 'landline']['duration']\n",
    "dur_voice = phone_df[phone_df['network'] == 'voicemail']['duration']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "id": "54d584e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize= (12,8))\n",
    "sns.distplot(dur_three)\n",
    "sns.distplot(dur_vod)\n",
    "sns.distplot(dur_data)\n",
    "sns.distplot(dur_met)\n",
    "sns.distplot(dur_tesco)\n",
    "sns.distplot(dur_land)\n",
    "sns.distplot(dur_voice) \n",
    "plt.legend(['Three', 'Vodafone', 'data', 'Meteor', 'Tesco', 'landline', 'voicemail'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "raw",
   "id": "d31336b6",
   "metadata": {},
   "source": [
    "Inference: \n",
    "    \n",
    "-> From the graph, we could see that the distribution of duration with respect to each network is more or less similar. \n",
    "-> We could also see that the peaks of all the distribution approximately overlap and so it can also be infered from this as that the mean of call duration for all the networks are more or less equal. \n",
    "-> This can be statistically proved with One-way ANOVA method of hypothesis."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74b7543f",
   "metadata": {},
   "source": [
    "##### d) Perform statistical tests to identify the relation between item and duration. Detailed explanation of relation between the classes is required."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "id": "4e2a5497",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "call    388\n",
       "sms     282\n",
       "data    150\n",
       "Name: item, dtype: int64"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "phone_df['item'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "id": "0cec81a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "duration_call = phone_df[phone_df['item'] == 'call']['duration']\n",
    "duration_sms = phone_df[phone_df['item'] == 'sms']['duration']\n",
    "duration_data = phone_df[phone_df['item'] == 'data']['duration']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "id": "53534033",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shapiro test for sample - duration_call:\n",
      " ShapiroResult(statistic=0.9973042011260986, pvalue=0.7803382277488708) \n",
      "\n",
      "Shapiro test for sample - duration_sms:\n",
      " ShapiroResult(statistic=0.9943404197692871, pvalue=0.37839275598526) \n",
      "\n",
      "Shapiro test for sample - duration_data:\n",
      " ShapiroResult(statistic=0.9912639856338501, pvalue=0.48521795868873596) \n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Test of variance for all three sample items:\n",
      " LeveneResult(statistic=3.4300976447008478, pvalue=0.03285087780533987) \n",
      "\n",
      "\n",
      "\n",
      "Levene test for - duration_call & duration_sms:\n",
      " LeveneResult(statistic=0.1986118328972824, pvalue=0.6559880731498395) \n",
      "\n",
      "Levene test for - duration_sms & duration_data:\n",
      " LeveneResult(statistic=6.5802914126186565, pvalue=0.010649597357420926) \n",
      "\n",
      "Levene test for - duration_call & duration_data:\n",
      " LeveneResult(statistic=4.997795914374391, pvalue=0.02579057825683839) \n",
      "\n",
      "\n",
      "\n",
      "\n",
      "ANOVA result for - duration_call & duration_sms:\n",
      " F_onewayResult(statistic=2.772571885758172, pvalue=0.09636087255994416)\n"
     ]
    }
   ],
   "source": [
    "# Test of Normality:  \n",
    "\n",
    "# Hypothesis formation for Shapiro test: \n",
    "# Ho: skew = 0 (Data is normal)\n",
    "# Ha: skew != 0 (Data is not normal)\n",
    "print(\"Shapiro test for sample - duration_call:\\n\", stats.shapiro(duration_call),'\\n')\n",
    "print(\"Shapiro test for sample - duration_sms:\\n\", stats.shapiro(duration_sms),'\\n')\n",
    "print(\"Shapiro test for sample - duration_data:\\n\", stats.shapiro(duration_data),'\\n\\n\\n')\n",
    "\n",
    "# Inference: \n",
    "# pValue of Shapiro test of duration data for each item (0.78, 0.37, 0.48) > significance value (0.05).\n",
    "# Hence, Ho is selected and so Data is normal.\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Test for equality of variance: \n",
    "\n",
    "# Hypothesis formation for Levene test: \n",
    "# Ho: All variances are equal\n",
    "# Ha: Atleast one variance is different \n",
    "print()\n",
    "print(\"Test of variance for all three sample items:\\n\", stats.levene(duration_call, duration_sms, duration_data),'\\n\\n') \n",
    "\n",
    "# Inference: \n",
    "# pValue of Levene test of duration data for each item (0.032) < significance value (0.05).\n",
    "# Hence, Ha is selected and so atleast one variance is different. \n",
    "\n",
    "#### Determining which variance is different:\n",
    "print()\n",
    "print(\"Levene test for - duration_call & duration_sms:\\n\",stats.levene(duration_call, duration_sms),'\\n')\n",
    "print(\"Levene test for - duration_sms & duration_data:\\n\",stats.levene(duration_sms, duration_data),'\\n')\n",
    "print(\"Levene test for - duration_call & duration_data:\\n\",stats.levene(duration_call, duration_data),'\\n\\n\\n') \n",
    "\n",
    "# Inference: \n",
    "# pValue of Levene test for - duration_call & duration_sms > significance value (0.05). \n",
    "# Hence, in this case we can select null Hypothesis Ho. \n",
    "# Thus, we infer that only the variance for samples - duration_call & duration_sms are equal.\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Hypothesis formation for ANOVA:\n",
    "# Ho: All means are equal (no relation)\n",
    "# Ha: Atleast one mean is different (relation exists)\n",
    "\n",
    "# Conditions for Anova: \n",
    "# 1) Data is normal \n",
    "# 2) Data has equal variance only for samples - duration_call & duration_sms\n",
    "\n",
    "# Inference: \n",
    "# Hence, one - way ANOVA can be applied only on samples - duration_call & duration_sms. \n",
    "\n",
    "print()\n",
    "print(\"ANOVA result for - duration_call & duration_sms:\\n\", stats.f_oneway(duration_call, duration_sms))"
   ]
  },
  {
   "cell_type": "raw",
   "id": "0ec8d325",
   "metadata": {},
   "source": [
    "Inference: \n",
    "\n",
    "-> pValue 0.09 > significance level 0.05 \n",
    "-> Hence, Ho can be selected and means of samples - duration_call & duration_sms are equal.\n",
    "-> So we can infer that there is no relation between duration of the item - call and the duration of item - sms."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bacaeed5",
   "metadata": {},
   "source": [
    "##### e) Perform statistical tests to identify the relation between network and network_type. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "id": "c646bc92",
   "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>network_type</th>\n",
       "      <th>data</th>\n",
       "      <th>landline</th>\n",
       "      <th>mobile</th>\n",
       "      <th>voicemail</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>network</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Meteor</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>87</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Tesco</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>84</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Three</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>215</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Vodafone</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>215</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>data</th>\n",
       "      <td>150</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>landline</th>\n",
       "      <td>0</td>\n",
       "      <td>42</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>voicemail</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "network_type  data  landline  mobile  voicemail\n",
       "network                                        \n",
       "Meteor           0         0      87          0\n",
       "Tesco            0         0      84          0\n",
       "Three            0         0     215          0\n",
       "Vodafone         0         0     215          0\n",
       "data           150         0       0          0\n",
       "landline         0        42       0          0\n",
       "voicemail        0         0       0         27"
      ]
     },
     "execution_count": 203,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hypothesis formation: \n",
    "# Ho: Network and Network-type are not related. \n",
    "# Ha: Network and Network-type are related.    \n",
    "\n",
    "obs = pd.crosstab(phone_df['network'], phone_df['network_type'])\n",
    "obs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "id": "e3465360",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Statistics:  2460.0 \n",
      "\n",
      "pValue:  0.0 \n",
      "\n",
      "Degrees of freedom:  18 \n",
      "\n",
      "Expected:  [[ 15.91463415   4.45609756  63.76463415   2.86463415]\n",
      " [ 15.36585366   4.30243902  61.56585366   2.76585366]\n",
      " [ 39.32926829  11.01219512 157.57926829   7.07926829]\n",
      " [ 39.32926829  11.01219512 157.57926829   7.07926829]\n",
      " [ 27.43902439   7.68292683 109.93902439   4.93902439]\n",
      " [  7.68292683   2.15121951  30.78292683   1.38292683]\n",
      " [  4.93902439   1.38292683  19.78902439   0.88902439]]\n"
     ]
    }
   ],
   "source": [
    "test_stat, pval, dof, exp = stats.chi2_contingency(obs)\n",
    "\n",
    "print(\"Test Statistics: \", test_stat,'\\n')\n",
    "print(\"pValue: \", pval,'\\n') \n",
    "print(\"Degrees of freedom: \", dof,'\\n') \n",
    "print(\"Expected: \", exp)"
   ]
  },
  {
   "cell_type": "raw",
   "id": "c300f10f",
   "metadata": {},
   "source": [
    "Inference: \n",
    "\n",
    "-> Significance lvl is 0.05 while the pValue in this case is 0.\n",
    "-> As the pValue is less than significance lvl, Ho is rejected and Ha can be selected. \n",
    "-> Hence, Network and Network-Type are related."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}