28th July - Simple Linear Regression (SLR).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SIMPLE LINEAR REGRESSION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings \n",
    "warnings.filterwarnings('ignore') \n",
    "# https://docs.python.org/3/library/warnings.html\n",
    "# importing warnings to overcome that we use this code.\n",
    "# ignoring all the warnings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd # pandas \n",
    "import numpy as np # numpy \n",
    "# pd for importing dataset\n",
    "# np to do discriptive"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.set_printoptions(precision=4, linewidth=100) \n",
    "# (precision=4, linewidth=100) is the precaution taking when results are got.\n",
    "# precision = 4 means dont have more than 4 decimals\n",
    "# linewidth = 100 means dont give too many things in the rows.\n",
    "# Setting Pandas print option to print decimal values upto 4 places"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "mba_salary_df = pd.read_csv( r\"C:\\Users\\Home\\Desktop\\MLAP - Getting Started with Python\\MLR and SLR\\MBA Salary.csv\") \n",
    "# Load the dataset. Type 'r' before your normal string it converts normal string to raw string.\n",
    "# r is raw data"
   ]
  },
  {
   "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>S. No.</th>\n",
       "      <th>Percentage in Grade 10</th>\n",
       "      <th>Salary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>62.00</td>\n",
       "      <td>270000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>76.33</td>\n",
       "      <td>200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>72.00</td>\n",
       "      <td>240000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>60.00</td>\n",
       "      <td>250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>61.00</td>\n",
       "      <td>180000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>55.00</td>\n",
       "      <td>300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>70.00</td>\n",
       "      <td>260000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>68.00</td>\n",
       "      <td>235000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>82.80</td>\n",
       "      <td>425000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>59.00</td>\n",
       "      <td>240000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   S. No.  Percentage in Grade 10  Salary\n",
       "0       1                   62.00  270000\n",
       "1       2                   76.33  200000\n",
       "2       3                   72.00  240000\n",
       "3       4                   60.00  250000\n",
       "4       5                   61.00  180000\n",
       "5       6                   55.00  300000\n",
       "6       7                   70.00  260000\n",
       "7       8                   68.00  235000\n",
       "8       9                   82.80  425000\n",
       "9      10                   59.00  240000"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mba_salary_df.head(10)\n",
    "#first 10 rows in dataset\n",
    "# .head() means first n no of rows in the dataset given."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50 entries, 0 to 49\n",
      "Data columns (total 3 columns):\n",
      " #   Column                  Non-Null Count  Dtype  \n",
      "---  ------                  --------------  -----  \n",
      " 0   S. No.                  50 non-null     int64  \n",
      " 1   Percentage in Grade 10  50 non-null     float64\n",
      " 2   Salary                  50 non-null     int64  \n",
      "dtypes: float64(1), int64(2)\n",
      "memory usage: 1.3 KB\n"
     ]
    }
   ],
   "source": [
    "mba_salary_df.info()\n",
    "# More about the Dataset\n",
    "# missing values, data type given in the information data\n",
    "# non null means no missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# no feature extraction \n",
    "# no gtesting for outliers \n",
    "# as we are using only 2 data\n",
    "# 1 is independent data\n",
    "# 1 is dependent data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model building\n",
    "import statsmodels.api as sm\n",
    "# statsmodels is a Python module that provides classes and functions \n",
    "# for the estimation of many different statistical models, \n",
    "# as well as for conducting statistical tests, and statistical data exploration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>const</th>\n",
       "      <th>Percentage in Grade 10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>62.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>76.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>72.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>60.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>61.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   const  Percentage in Grade 10\n",
       "0    1.0                   62.00\n",
       "1    1.0                   76.33\n",
       "2    1.0                   72.00\n",
       "3    1.0                   60.00\n",
       "4    1.0                   61.00"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = sm.add_constant( mba_salary_df['Percentage in Grade 10'] )\n",
    "X.head(5)\n",
    "# x is independent variable which includes % of 10th grade.\n",
    "#Separate the \"Independent Variable\"\n",
    "# beto not = constant \n",
    "# so we added constant to our dataset\n",
    "# when we use ordinary least squared criteria the constant will take various values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y = mba_salary_df['Salary']\n",
    "# Dependent Variable separate.\n",
    "# y is dependent and salary assigned to it\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "# Splitting the dataset into Train and Test.\n",
    "# we divide the dataset in train and test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_X, test_X, train_y, test_y = train_test_split( X ,Y,train_size = 0.8,random_state = 100 )\n",
    "# 80:20 - Train:Test\n",
    "# standard code\n",
    "# random state?\n",
    "#On the other hand if you use random_state=some_number, then you can guarantee that the output of Run 1 will be equal to the output of Run 2, i.e. your split will be always the same. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "mba_salary_lm = sm.OLS( train_y, train_X ).fit()\n",
    "# Model Fit\n",
    "# data building\n",
    "# used ordinary least square\n",
    "# fitting the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "const                     30587.285652\n",
      "Percentage in Grade 10     3560.587383\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print( mba_salary_lm.params )\n",
    "# Estimated Parameter Results\n",
    "# beta not and beta 1 asked\n",
    "# Python attributes extended to have features such as type and range checking, dynamically generated values, documentation strings, default values, etc., each of which is inherited from parent classes if not specified in a subclass."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_X, test_X, train_y, test_y = train_test_split( X ,Y,train_size = 0.8 )\n",
    "# 80:20 - Train:Test\n",
    "# standard code\n",
    "# no random state"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "mba_salary_lm = sm.OLS( train_y, train_X ).fit()\n",
    "# Model Fit \n",
    "#train_X, test_X, train_y, test_y = train_test_split( X ,Y,train_size = 0.8 )\n",
    "# no random state"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "const                     60584.943180\n",
      "Percentage in Grade 10     3148.929116\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print( mba_salary_lm.params )\n",
    "# Estimated Parameter Results\n",
    "#train_X, test_X, train_y, test_y = train_test_split( X ,Y,train_size = 0.8 )\n",
    "# no random state"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "        <td>Model:</td>               <td>OLS</td>         <td>Adj. R-squared:</td>      <td>0.190</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Dependent Variable:</td>      <td>Salary</td>             <td>AIC:</td>          <td>1008.8680</td>\n",
       "</tr>\n",
       "<tr>\n",
       "         <td>Date:</td>        <td>2020-07-29 03:12</td>        <td>BIC:</td>          <td>1012.2458</td>\n",
       "</tr>\n",
       "<tr>\n",
       "   <td>No. Observations:</td>         <td>40</td>          <td>Log-Likelihood:</td>     <td>-502.43</td> \n",
       "</tr>\n",
       "<tr>\n",
       "       <td>Df Model:</td>              <td>1</td>           <td>F-statistic:</td>        <td>10.16</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>Df Residuals:</td>           <td>38</td>        <td>Prob (F-statistic):</td>   <td>0.00287</td> \n",
       "</tr>\n",
       "<tr>\n",
       "      <td>R-squared:</td>            <td>0.211</td>            <td>Scale:</td>        <td>5.0121e+09</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>               <th>Coef.</th>    <th>Std.Err.</th>     <th>t</th>    <th>P>|t|</th>    <th>[0.025</th>      <th>0.975]</th>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>const</th>                  <td>30587.2857</td> <td>71869.4497</td> <td>0.4256</td> <td>0.6728</td> <td>-114904.8089</td> <td>176079.3802</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Percentage in Grade 10</th>  <td>3560.5874</td>  <td>1116.9258</td> <td>3.1878</td> <td>0.0029</td>   <td>1299.4892</td>   <td>5821.6855</td> \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "     <td>Omnibus:</td>    <td>2.048</td>  <td>Durbin-Watson:</td>   <td>2.611</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Prob(Omnibus):</td> <td>0.359</td> <td>Jarque-Bera (JB):</td> <td>1.724</td>\n",
       "</tr>\n",
       "<tr>\n",
       "       <td>Skew:</td>     <td>0.369</td>     <td>Prob(JB):</td>     <td>0.422</td>\n",
       "</tr>\n",
       "<tr>\n",
       "     <td>Kurtosis:</td>   <td>2.300</td>  <td>Condition No.:</td>    <td>413</td> \n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary2.Summary'>\n",
       "\"\"\"\n",
       "                          Results: Ordinary least squares\n",
       "===================================================================================\n",
       "Model:                    OLS                   Adj. R-squared:          0.190     \n",
       "Dependent Variable:       Salary                AIC:                     1008.8680 \n",
       "Date:                     2020-07-29 03:12      BIC:                     1012.2458 \n",
       "No. Observations:         40                    Log-Likelihood:          -502.43   \n",
       "Df Model:                 1                     F-statistic:             10.16     \n",
       "Df Residuals:             38                    Prob (F-statistic):      0.00287   \n",
       "R-squared:                0.211                 Scale:                   5.0121e+09\n",
       "-----------------------------------------------------------------------------------\n",
       "                         Coef.     Std.Err.    t    P>|t|     [0.025       0.975]  \n",
       "-----------------------------------------------------------------------------------\n",
       "const                  30587.2857 71869.4497 0.4256 0.6728 -114904.8089 176079.3802\n",
       "Percentage in Grade 10  3560.5874  1116.9258 3.1878 0.0029    1299.4892   5821.6855\n",
       "-----------------------------------------------------------------------------------\n",
       "Omnibus:                    2.048              Durbin-Watson:                 2.611\n",
       "Prob(Omnibus):              0.359              Jarque-Bera (JB):              1.724\n",
       "Skew:                       0.369              Prob(JB):                      0.422\n",
       "Kurtosis:                   2.300              Condition No.:                 413  \n",
       "===================================================================================\n",
       "\n",
       "\"\"\""
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mba_salary_lm.summary2()\n",
    "# Model Diagostics\n",
    "# giving many criteria\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sn\n",
    "#matplotlib inline means the output of plotting commands is displayed inline within frontends\n",
    "# Python libraries for Visualization\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 576x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "mba_salary_resid = mba_salary_lm.resid\n",
    "probplot = sm.ProbPlot( mba_salary_resid )\n",
    "plt.figure( figsize = (8, 6) )\n",
    "probplot.ppplot( line='45' )\n",
    "plt.title( \"Fig 1 - Normal P-P Plot of Regression Standardized Residuals\" )\n",
    "plt.show()\n",
    "#Residual Analysis\n",
    "# when all variables fall in line 45 then it is normally distributed.\n",
    "# error assumption is that: error terms follow normal distribution, mean is 0 and its variance is uniformm as well as error terms are uncorrelated."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_standardized_values( vals ):\n",
    "    return (vals - vals.mean())/vals.std()\n",
    "# Testing for uniform variance - Homoscedasticity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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": [
    "plt.scatter( get_standardized_values( mba_salary_lm.fittedvalues ),\n",
    "get_standardized_values( mba_salary_resid ) )\n",
    "plt.title( \"Fig 2 - Residual Plot: MBA Salary Prediction\" );\n",
    "plt.xlabel( \"Standardized predicted values\")\n",
    "plt.ylabel( \"Standardized Residuals\");\n",
    "# homoscedacity\n",
    "# uniform variance\n",
    "# error terms are random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats import zscore\n",
    "# Finding Z-Score for Outlier Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "mba_salary_df['z_score_salary'] = zscore( mba_salary_df.Salary )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>S. No.</th>\n",
       "      <th>Percentage in Grade 10</th>\n",
       "      <th>Salary</th>\n",
       "      <th>z_score_salary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [S. No., Percentage in Grade 10, Salary, z_score_salary]\n",
       "Index: []"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mba_salary_df[ (mba_salary_df.z_score_salary > 3.0) | (mba_salary_df.z_score_salary < -3.0) ]\n",
    "# Outlier would lie in plus or minus 3 SD."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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": [
    "import numpy as np\n",
    "mba_influence = mba_salary_lm.get_influence()\n",
    "(c, p) = mba_influence.cooks_distance\n",
    "plt.stem( np.arange( len( train_X) ),\n",
    "np.round( c, 3 ),\n",
    "markerfmt=\",\" );\n",
    "plt.title( \"Figure 3 - Cooks distance for all observations in MBA Salaray dataset\" );\n",
    "plt.xlabel( \"Row index\")\n",
    "plt.ylabel( \"Cooks Distance\");\n",
    "# Cooks Distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from statsmodels.graphics.regressionplots import influence_plot\n",
    "fig, ax = plt.subplots( figsize=(8,6) )\n",
    "influence_plot( mba_salary_lm, ax = ax )\n",
    "plt.title( \"Figure 4 - Leverage Value Vs Residuals\")\n",
    "plt.show();\n",
    "\n",
    "# Leverage Value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_y = mba_salary_lm.predict( test_X )\n",
    "# Prediction on Validation set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# to find my r square\n",
    "from sklearn.metrics import r2_score, mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.15664584974230378"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.abs(r2_score(test_y, pred_y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "73458.04348346894"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sqrt(mean_squared_error(test_y, pred_y))\n",
    "# root mean square error checked"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.sandbox.regression.predstd import wls_prediction_std\n",
    "# Predict the y values\n",
    "pred_y = mba_salary_lm.predict( test_X )\n",
    "# Predict the low and high interval values for y\n",
    "_, pred_y_low, pred_y_high = wls_prediction_std( mba_salary_lm,\n",
    "test_X,\n",
    "alpha = 0.1)\n",
    "# Store all the values in a dataframe\n",
    "pred_y_df = pd.DataFrame( { 'grade_10_perc': test_X['Percentage in Grade 10'],\n",
    "'pred_y': pred_y,\n",
    "'pred_y_left': pred_y_low,\n",
    "'pred_y_right': pred_y_high } )\n",
    "# pred y new predicted values\n",
    "# left hand side prediction\n",
    "# right hand side prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>grade_10_perc</th>\n",
       "      <th>pred_y</th>\n",
       "      <th>pred_y_left</th>\n",
       "      <th>pred_y_right</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>70.0</td>\n",
       "      <td>279828.402452</td>\n",
       "      <td>158379.832044</td>\n",
       "      <td>401276.972860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>68.0</td>\n",
       "      <td>272707.227686</td>\n",
       "      <td>151576.715020</td>\n",
       "      <td>393837.740352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>52.0</td>\n",
       "      <td>215737.829560</td>\n",
       "      <td>92950.942395</td>\n",
       "      <td>338524.716726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>58.0</td>\n",
       "      <td>237101.353858</td>\n",
       "      <td>115806.869618</td>\n",
       "      <td>358395.838097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>74.5</td>\n",
       "      <td>295851.045675</td>\n",
       "      <td>173266.083342</td>\n",
       "      <td>418436.008008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>60.8</td>\n",
       "      <td>247070.998530</td>\n",
       "      <td>126117.560983</td>\n",
       "      <td>368024.436076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>55.0</td>\n",
       "      <td>226419.591709</td>\n",
       "      <td>104507.444388</td>\n",
       "      <td>348331.739030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>78.0</td>\n",
       "      <td>308313.101515</td>\n",
       "      <td>184450.060488</td>\n",
       "      <td>432176.142542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>63.0</td>\n",
       "      <td>254904.290772</td>\n",
       "      <td>134057.999258</td>\n",
       "      <td>375750.582286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>74.4</td>\n",
       "      <td>295494.986937</td>\n",
       "      <td>172941.528691</td>\n",
       "      <td>418048.445182</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    grade_10_perc         pred_y    pred_y_left   pred_y_right\n",
       "6            70.0  279828.402452  158379.832044  401276.972860\n",
       "36           68.0  272707.227686  151576.715020  393837.740352\n",
       "37           52.0  215737.829560   92950.942395  338524.716726\n",
       "28           58.0  237101.353858  115806.869618  358395.838097\n",
       "43           74.5  295851.045675  173266.083342  418436.008008\n",
       "49           60.8  247070.998530  126117.560983  368024.436076\n",
       "5            55.0  226419.591709  104507.444388  348331.739030\n",
       "33           78.0  308313.101515  184450.060488  432176.142542\n",
       "20           63.0  254904.290772  134057.999258  375750.582286\n",
       "42           74.4  295494.986937  172941.528691  418048.445182"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_y_df[0:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}