The Oscars go to....ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# And the Oscar goes to ....\n",
    "\n",
    "This is a great supervised learning problem: we’ve got a nice set of data with information (training data) that we suspect will help us predict who will win this year. This is called classification, since there are only two options for each datapoint: win or not win.\n",
    "Now the problem is, we don’t know the important trends and relationships within our training data. Our goal is to come up with a program that models these relationships, so we can just input The Green Book and get back win.\n",
    "\n",
    "</br>\n",
    "<center>\n",
    "    <img src =oscars.1.jpg width = 400 />\n",
    "    *Tadaaaa...*\n",
    "</center>\n",
    "\n",
    "Let’s start simple:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "hideCode": false,
    "hidePrompt": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import tree\n",
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Reading the dataset and Identify the anomalies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>category</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "      <th>MovieLensrating</th>\n",
       "      <th>runtimeMinutes</th>\n",
       "      <th>IMDBRating</th>\n",
       "      <th>imdb_numVotes</th>\n",
       "      <th>RottenCriticRating</th>\n",
       "      <th>rottencritic_numVotes</th>\n",
       "      <th>RottenRating</th>\n",
       "      <th>rotten_numVotes</th>\n",
       "      <th>won_bafta</th>\n",
       "      <th>won_sa</th>\n",
       "      <th>won_sag</th>\n",
       "      <th>won_cc</th>\n",
       "      <th>won_dga</th>\n",
       "      <th>won_nbra</th>\n",
       "      <th>won_oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1944</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Double Indemnity</td>\n",
       "      <td>Crime|Drama|Film-Noir</td>\n",
       "      <td>4.20</td>\n",
       "      <td>107</td>\n",
       "      <td>8.3</td>\n",
       "      <td>125662</td>\n",
       "      <td>96</td>\n",
       "      <td>57</td>\n",
       "      <td>95</td>\n",
       "      <td>35629</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1944</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Gaslight</td>\n",
       "      <td>Drama|Thriller</td>\n",
       "      <td>4.03</td>\n",
       "      <td>114</td>\n",
       "      <td>7.8</td>\n",
       "      <td>20733</td>\n",
       "      <td>86</td>\n",
       "      <td>28</td>\n",
       "      <td>90</td>\n",
       "      <td>8460</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1944</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Going My Way</td>\n",
       "      <td>Comedy|Drama|Musical</td>\n",
       "      <td>3.56</td>\n",
       "      <td>126</td>\n",
       "      <td>7.2</td>\n",
       "      <td>9228</td>\n",
       "      <td>79</td>\n",
       "      <td>24</td>\n",
       "      <td>76</td>\n",
       "      <td>5748</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1944</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Since You Went Away</td>\n",
       "      <td>Drama|War</td>\n",
       "      <td>3.65</td>\n",
       "      <td>177</td>\n",
       "      <td>7.6</td>\n",
       "      <td>3879</td>\n",
       "      <td>80</td>\n",
       "      <td>10</td>\n",
       "      <td>82</td>\n",
       "      <td>1305</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1944</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Wilson</td>\n",
       "      <td>Drama</td>\n",
       "      <td>3.21</td>\n",
       "      <td>154</td>\n",
       "      <td>6.8</td>\n",
       "      <td>1314</td>\n",
       "      <td>88</td>\n",
       "      <td>8</td>\n",
       "      <td>41</td>\n",
       "      <td>363</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   year      category                title                 genres  \\\n",
       "0  1944  BEST PICTURE     Double Indemnity  Crime|Drama|Film-Noir   \n",
       "1  1944  BEST PICTURE             Gaslight         Drama|Thriller   \n",
       "2  1944  BEST PICTURE         Going My Way   Comedy|Drama|Musical   \n",
       "3  1944  BEST PICTURE  Since You Went Away              Drama|War   \n",
       "4  1944  BEST PICTURE               Wilson                  Drama   \n",
       "\n",
       "   MovieLensrating  runtimeMinutes  IMDBRating  imdb_numVotes  \\\n",
       "0             4.20             107         8.3         125662   \n",
       "1             4.03             114         7.8          20733   \n",
       "2             3.56             126         7.2           9228   \n",
       "3             3.65             177         7.6           3879   \n",
       "4             3.21             154         6.8           1314   \n",
       "\n",
       "   RottenCriticRating  rottencritic_numVotes  RottenRating  rotten_numVotes  \\\n",
       "0                  96                     57            95            35629   \n",
       "1                  86                     28            90             8460   \n",
       "2                  79                     24            76             5748   \n",
       "3                  80                     10            82             1305   \n",
       "4                  88                      8            41              363   \n",
       "\n",
       "   won_bafta  won_sa  won_sag  won_cc  won_dga  won_nbra  won_oscar  \n",
       "0          0       0        0       0        0         0          0  \n",
       "1          0       0        0       0        0         0          0  \n",
       "2          0       0        0       0        0         0          1  \n",
       "3          0       0        0       0        0         0          0  \n",
       "4          0       0        0       0        0         0          0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in data as pandas dataframe and display the oscar dataset\n",
    "dataset = pd.read_csv('OscarData.csv')\n",
    "dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>MovieLensrating</th>\n",
       "      <th>runtimeMinutes</th>\n",
       "      <th>IMDBRating</th>\n",
       "      <th>imdb_numVotes</th>\n",
       "      <th>RottenCriticRating</th>\n",
       "      <th>rottencritic_numVotes</th>\n",
       "      <th>RottenRating</th>\n",
       "      <th>rotten_numVotes</th>\n",
       "      <th>won_bafta</th>\n",
       "      <th>won_sa</th>\n",
       "      <th>won_sag</th>\n",
       "      <th>won_cc</th>\n",
       "      <th>won_dga</th>\n",
       "      <th>won_nbra</th>\n",
       "      <th>won_oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>4.110000e+02</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>4.110000e+02</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "      <td>411.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1984.000000</td>\n",
       "      <td>3.744891</td>\n",
       "      <td>130.605839</td>\n",
       "      <td>7.703163</td>\n",
       "      <td>2.289712e+05</td>\n",
       "      <td>87.581509</td>\n",
       "      <td>117.270073</td>\n",
       "      <td>83.464720</td>\n",
       "      <td>6.384332e+05</td>\n",
       "      <td>0.133820</td>\n",
       "      <td>0.060827</td>\n",
       "      <td>0.053528</td>\n",
       "      <td>0.063260</td>\n",
       "      <td>0.170316</td>\n",
       "      <td>0.148418</td>\n",
       "      <td>0.182482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>22.762185</td>\n",
       "      <td>0.295022</td>\n",
       "      <td>25.593356</td>\n",
       "      <td>0.481071</td>\n",
       "      <td>3.218728e+05</td>\n",
       "      <td>11.907778</td>\n",
       "      <td>114.005338</td>\n",
       "      <td>9.163586</td>\n",
       "      <td>4.055703e+06</td>\n",
       "      <td>0.340874</td>\n",
       "      <td>0.239304</td>\n",
       "      <td>0.225358</td>\n",
       "      <td>0.243727</td>\n",
       "      <td>0.376369</td>\n",
       "      <td>0.355948</td>\n",
       "      <td>0.386712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1944.000000</td>\n",
       "      <td>2.820000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>6.200000</td>\n",
       "      <td>1.147000e+03</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>3.340000e+02</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1964.000000</td>\n",
       "      <td>3.565000</td>\n",
       "      <td>113.000000</td>\n",
       "      <td>7.400000</td>\n",
       "      <td>2.047700e+04</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>79.000000</td>\n",
       "      <td>9.507500e+03</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1985.000000</td>\n",
       "      <td>3.760000</td>\n",
       "      <td>126.000000</td>\n",
       "      <td>7.700000</td>\n",
       "      <td>9.304000e+04</td>\n",
       "      <td>91.000000</td>\n",
       "      <td>60.000000</td>\n",
       "      <td>85.000000</td>\n",
       "      <td>4.973300e+04</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2006.000000</td>\n",
       "      <td>3.935000</td>\n",
       "      <td>143.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>3.050575e+05</td>\n",
       "      <td>95.000000</td>\n",
       "      <td>212.500000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>1.798465e+05</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2018.000000</td>\n",
       "      <td>4.720000</td>\n",
       "      <td>220.000000</td>\n",
       "      <td>9.300000</td>\n",
       "      <td>2.070726e+06</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>472.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>3.579764e+07</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              year  MovieLensrating  runtimeMinutes  IMDBRating  \\\n",
       "count   411.000000       411.000000      411.000000  411.000000   \n",
       "mean   1984.000000         3.744891      130.605839    7.703163   \n",
       "std      22.762185         0.295022       25.593356    0.481071   \n",
       "min    1944.000000         2.820000       84.000000    6.200000   \n",
       "25%    1964.000000         3.565000      113.000000    7.400000   \n",
       "50%    1985.000000         3.760000      126.000000    7.700000   \n",
       "75%    2006.000000         3.935000      143.000000    8.000000   \n",
       "max    2018.000000         4.720000      220.000000    9.300000   \n",
       "\n",
       "       imdb_numVotes  RottenCriticRating  rottencritic_numVotes  RottenRating  \\\n",
       "count   4.110000e+02          411.000000             411.000000    411.000000   \n",
       "mean    2.289712e+05           87.581509             117.270073     83.464720   \n",
       "std     3.218728e+05           11.907778             114.005338      9.163586   \n",
       "min     1.147000e+03            0.000000               0.000000     41.000000   \n",
       "25%     2.047700e+04           84.000000              30.000000     79.000000   \n",
       "50%     9.304000e+04           91.000000              60.000000     85.000000   \n",
       "75%     3.050575e+05           95.000000             212.500000     90.000000   \n",
       "max     2.070726e+06          100.000000             472.000000     98.000000   \n",
       "\n",
       "       rotten_numVotes   won_bafta      won_sa     won_sag      won_cc  \\\n",
       "count     4.110000e+02  411.000000  411.000000  411.000000  411.000000   \n",
       "mean      6.384332e+05    0.133820    0.060827    0.053528    0.063260   \n",
       "std       4.055703e+06    0.340874    0.239304    0.225358    0.243727   \n",
       "min       3.340000e+02    0.000000    0.000000    0.000000    0.000000   \n",
       "25%       9.507500e+03    0.000000    0.000000    0.000000    0.000000   \n",
       "50%       4.973300e+04    0.000000    0.000000    0.000000    0.000000   \n",
       "75%       1.798465e+05    0.000000    0.000000    0.000000    0.000000   \n",
       "max       3.579764e+07    1.000000    1.000000    1.000000    1.000000   \n",
       "\n",
       "          won_dga    won_nbra   won_oscar  \n",
       "count  411.000000  411.000000  411.000000  \n",
       "mean     0.170316    0.148418    0.182482  \n",
       "std      0.376369    0.355948    0.386712  \n",
       "min      0.000000    0.000000    0.000000  \n",
       "25%      0.000000    0.000000    0.000000  \n",
       "50%      0.000000    0.000000    0.000000  \n",
       "75%      0.000000    0.000000    0.000000  \n",
       "max      1.000000    1.000000    1.000000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Descriptive statistics for each column\n",
    "dataset.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>category</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "      <th>MovieLensrating</th>\n",
       "      <th>runtimeMinutes</th>\n",
       "      <th>IMDBRating</th>\n",
       "      <th>imdb_numVotes</th>\n",
       "      <th>RottenCriticRating</th>\n",
       "      <th>rottencritic_numVotes</th>\n",
       "      <th>RottenRating</th>\n",
       "      <th>rotten_numVotes</th>\n",
       "      <th>won_bafta</th>\n",
       "      <th>won_sa</th>\n",
       "      <th>won_sag</th>\n",
       "      <th>won_cc</th>\n",
       "      <th>won_dga</th>\n",
       "      <th>won_nbra</th>\n",
       "      <th>won_oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>406</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>The Favourite</td>\n",
       "      <td>Drama|Comedy</td>\n",
       "      <td>3.82</td>\n",
       "      <td>119</td>\n",
       "      <td>7.7</td>\n",
       "      <td>103011</td>\n",
       "      <td>93</td>\n",
       "      <td>362</td>\n",
       "      <td>67</td>\n",
       "      <td>5754</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>407</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Green Book</td>\n",
       "      <td>Drama|Comedy</td>\n",
       "      <td>3.94</td>\n",
       "      <td>130</td>\n",
       "      <td>8.3</td>\n",
       "      <td>170827</td>\n",
       "      <td>78</td>\n",
       "      <td>316</td>\n",
       "      <td>92</td>\n",
       "      <td>7859</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>408</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Roma</td>\n",
       "      <td>Drama</td>\n",
       "      <td>3.76</td>\n",
       "      <td>135</td>\n",
       "      <td>7.8</td>\n",
       "      <td>100158</td>\n",
       "      <td>96</td>\n",
       "      <td>353</td>\n",
       "      <td>71</td>\n",
       "      <td>4667</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>409</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>A Star Is Born</td>\n",
       "      <td>Drama|Romance</td>\n",
       "      <td>3.67</td>\n",
       "      <td>135</td>\n",
       "      <td>7.8</td>\n",
       "      <td>229903</td>\n",
       "      <td>89</td>\n",
       "      <td>472</td>\n",
       "      <td>80</td>\n",
       "      <td>18019</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Vice</td>\n",
       "      <td>Drama|Comedy</td>\n",
       "      <td>3.49</td>\n",
       "      <td>132</td>\n",
       "      <td>7.2</td>\n",
       "      <td>58938</td>\n",
       "      <td>66</td>\n",
       "      <td>326</td>\n",
       "      <td>58</td>\n",
       "      <td>3902</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     year      category           title         genres  MovieLensrating  \\\n",
       "406  2018  BEST PICTURE   The Favourite   Drama|Comedy             3.82   \n",
       "407  2018  BEST PICTURE      Green Book   Drama|Comedy             3.94   \n",
       "408  2018  BEST PICTURE            Roma          Drama             3.76   \n",
       "409  2018  BEST PICTURE  A Star Is Born  Drama|Romance             3.67   \n",
       "410  2018  BEST PICTURE            Vice   Drama|Comedy             3.49   \n",
       "\n",
       "     runtimeMinutes  IMDBRating  imdb_numVotes  RottenCriticRating  \\\n",
       "406             119         7.7         103011                  93   \n",
       "407             130         8.3         170827                  78   \n",
       "408             135         7.8         100158                  96   \n",
       "409             135         7.8         229903                  89   \n",
       "410             132         7.2          58938                  66   \n",
       "\n",
       "     rottencritic_numVotes  RottenRating  rotten_numVotes  won_bafta  won_sa  \\\n",
       "406                    362            67             5754          0       1   \n",
       "407                    316            92             7859          0       0   \n",
       "408                    353            71             4667          1       1   \n",
       "409                    472            80            18019          0       1   \n",
       "410                    326            58             3902          0       0   \n",
       "\n",
       "     won_sag  won_cc  won_dga  won_nbra  won_oscar  \n",
       "406        0       0        0         0          0  \n",
       "407        0       0        0         1          1  \n",
       "408        0       1        1         0          0  \n",
       "409        0       0        0         0          0  \n",
       "410        0       0        0         0          0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = dataset.fillna(dataset.mean())\n",
    "dataset.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['MovieLensrating',\n",
       " 'runtimeMinutes',\n",
       " 'IMDBRating',\n",
       " 'imdb_numVotes',\n",
       " 'RottenCriticRating',\n",
       " 'rottencritic_numVotes',\n",
       " 'RottenRating',\n",
       " 'rotten_numVotes',\n",
       " 'won_bafta',\n",
       " 'won_sa',\n",
       " 'won_sag',\n",
       " 'won_cc',\n",
       " 'won_dga']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#feature list for variable importance\n",
    "features = dataset.iloc[:,4:17]\n",
    "feature_list = list(features.columns)\n",
    "feature_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Divide the dataset into features(X) and the label(y) which we have to predict\n",
    "X = dataset.iloc[:, 4 : 17].values\n",
    "y = dataset.iloc[:, 18].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The shape of our features is: (411, 13)\n"
     ]
    }
   ],
   "source": [
    "print('The shape of our features is:', X.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0,\n",
       "       0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1,\n",
       "       0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,\n",
       "       0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0,\n",
       "       0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0,\n",
       "       0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,\n",
       "       0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,\n",
       "       0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,\n",
       "       1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0,\n",
       "       0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0,\n",
       "       0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
       "       1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0,\n",
       "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,\n",
       "       0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Viewing the sliced label data\n",
    "y"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Splitting the dataset into the Training set and Test set\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data is spilt with random_state = 0 that is the data remains the same throughout the entire execution and for all the algorithms. This is done to maintain consistency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Using Skicit-learn to split data into training and testing sets\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.26, random_state = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Features Shape: (304, 13)\n",
      "Training Labels Shape: (304,)\n",
      "Testing Features Shape: (107, 13)\n",
      "Testing Labels Shape: (107,)\n"
     ]
    }
   ],
   "source": [
    "print('Training Features Shape:', X_train.shape)\n",
    "print('Training Labels Shape:', y_train.shape)\n",
    "print('Testing Features Shape:', X_test.shape)\n",
    "print('Testing Labels Shape:', y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Functions which will be used throughout the entire code"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. This function will be used plot the confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "import itertools\n",
    "\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Oranges):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    Source: http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "\n",
    "    plt.figure(figsize = (10, 10))\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title, size = 24)\n",
    "    plt.colorbar(aspect=4)\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45, size = 14)\n",
    "    plt.yticks(tick_marks, classes, size = 14)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    \n",
    "    # Labeling the plot\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt), fontsize = 20,\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "        \n",
    "    plt.grid(None)\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label', size = 18)\n",
    "    plt.xlabel('Predicted label', size = 18)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. This function will be used to plot the ROC curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_model(predictions_lr, probs_lr, train_predictions_lr, train_probs_lr):\n",
    "    \"\"\"Compare machine learning model to baseline performance.\n",
    "    Computes statistics and shows ROC curve.\"\"\"\n",
    "    \n",
    "    baseline = {}\n",
    "    \n",
    "    baseline['recall'] = recall_score(y_test, [1 for _ in range(len(y_test))])\n",
    "    baseline['precision'] = precision_score(y_test, [1 for _ in range(len(y_test))])\n",
    "    baseline['roc'] = 0.5\n",
    "    \n",
    "    results = {}\n",
    "    \n",
    "    results['recall'] = recall_score(y_test, predictions_lr)\n",
    "    results['precision'] = precision_score(y_test, predictions_lr)\n",
    "    results['roc'] = roc_auc_score(y_test, probs_lr)\n",
    "    \n",
    "    train_results = {}\n",
    "    train_results['recall'] = recall_score(y_train, train_predictions_lr)\n",
    "    train_results['precision'] = precision_score(y_train, train_predictions_lr)\n",
    "    train_results['roc'] = roc_auc_score(y_train, train_probs_lr)\n",
    "    \n",
    "    for metric in ['recall', 'precision', 'roc']:\n",
    "        print(f'{metric.capitalize()} Baseline: {round(baseline[metric], 2)} Test: {round(results[metric], 2)} Train: {round(train_results[metric], 2)}')\n",
    "    \n",
    "    # Calculate false positive rates and true positive rates\n",
    "    base_fpr, base_tpr, _ = roc_curve(y_test, [1 for _ in range(len(y_test))])\n",
    "    model_fpr, model_tpr, _ = roc_curve(y_test, probs_lr)\n",
    "\n",
    "    plt.figure(figsize = (8, 6))\n",
    "    plt.rcParams['font.size'] = 16\n",
    "    \n",
    "    # Plot both curves\n",
    "    plt.plot(base_fpr, base_tpr, 'g', label = 'baseline')\n",
    "    plt.plot(model_fpr, model_tpr, 'r', label = 'model')\n",
    "    plt.legend();\n",
    "    plt.xlabel('False Positive Rate'); plt.ylabel('True Positive Rate'); plt.title('ROC Curves');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naive Bayes \n",
    "Naive Bayes is a very good method for classification as it uses probabilites for prediction. Since predicting oscars is using probabilities, this will be a good option.\n",
    "\n",
    "The first 2 steps will be the same and now we move on to fit the Oscar data to the logisitic regression model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Fitting Naive Bayes to the training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GaussianNB(priors=None, var_smoothing=1e-09)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "classifier_nb = GaussianNB()\n",
    "classifier_nb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Classes and Class prior based on model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classes and their prior -\n",
      "0 : 0.83\n",
      "1 : 0.17\n"
     ]
    }
   ],
   "source": [
    "print(\"Classes and their prior -\")\n",
    "print(classifier_nb.classes_[0], \":\", round(classifier_nb.class_prior_[0],2))\n",
    "print(classifier_nb.classes_[1], \":\", round(classifier_nb.class_prior_[1],2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Assess Naive Bayes Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predicting the Test set results\n",
    "y_pred_nb = classifier_nb.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Making the Confusion Matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "cm = confusion_matrix(y_test, y_pred_nb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[82  3]\n",
      " [21  1]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(cm, classes = ['Lost Oscar', 'Won Oscar'],\n",
    "                      title = 'Oscar Confusion Matrix',cmap=plt.cm.Blues)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> This matrix shows that there are 82+1=83 correct predictions and only 21+3=24 incorrect predictions.\n",
    "The naive bayes model predicts 21 false negatives which means the movies which actually won the oscar but predicted that the movie lost and the model also predicted 3 false postives which means the movies which actually lost the oscar but predicted that the movie won the oscar. So in all, the model predicts more false negatives which is good. This could be improved using a Random Forest Classifier</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.7757009345794392\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "print(\"Accuracy\", metrics.accuracy_score(y_test, y_pred_nb))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make probability predictions\n",
    "train_probs_nb = classifier_nb.predict_proba(X_train)[:, 1]\n",
    "probs_nb = classifier_nb.predict_proba(X_test)[:, 1]\n",
    "\n",
    "train_predictions_nb = classifier_nb.predict(X_train)\n",
    "predictions_nb = classifier_nb.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train ROC AUC Score: 0.6678944598962641\n",
      "Test ROC AUC  Score: 0.6053475935828877\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score, recall_score, roc_auc_score, roc_curve\n",
    "\n",
    "print(f'Train ROC AUC Score: {roc_auc_score(y_train, train_probs_nb)}')\n",
    "print(f'Test ROC AUC  Score: {roc_auc_score(y_test, probs_nb)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baseline ROC AUC: 0.5\n"
     ]
    }
   ],
   "source": [
    "print(f'Baseline ROC AUC: {roc_auc_score(y_test, [1 for _ in range(len(y_test))])}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall Baseline: 1.0 Test: 0.05 Train: 0.13\n",
      "Precision Baseline: 0.21 Test: 0.25 Train: 0.5\n",
      "Roc Baseline: 0.5 Test: 0.61 Train: 0.67\n"
     ]
    },
    {
     "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": [
    "evaluate_model(predictions_nb, probs_nb, train_predictions_nb, train_probs_nb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Finally our Naive Bayes Model predicts the Oscars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_features = dataset[feature_list].values\n",
    "\n",
    "pred_naive = classifier_nb.predict_proba(test_features)[:,1]\n",
    "movie_name = np.array(dataset['title'])\n",
    "year = np.array(dataset['year'])\n",
    "won_oscar = np.array(dataset['won_oscar'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "naive_prediction = pd.DataFrame(pred_naive,movie_name,columns =[\"Probability\"] )\n",
    "naive_prediction[\"Year\"] = year \n",
    "naive_prediction[\"Won Oscar\"] = won_oscar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>Probability</th>\n",
       "      <th>Year</th>\n",
       "      <th>Won Oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Arrival</th>\n",
       "      <td>0.050988</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fences</th>\n",
       "      <td>0.026981</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hacksaw Ridge</th>\n",
       "      <td>0.037200</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hell or High Water</th>\n",
       "      <td>0.028102</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hidden Figures</th>\n",
       "      <td>0.028228</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>La La Land</th>\n",
       "      <td>0.041534</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lion</th>\n",
       "      <td>0.028417</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Manchester by the Sea</th>\n",
       "      <td>0.029393</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moonlight</th>\n",
       "      <td>0.029331</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Call Me by Your Name</th>\n",
       "      <td>0.027758</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Darkest Hour</th>\n",
       "      <td>0.027472</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dunkirk</th>\n",
       "      <td>0.044161</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Get Out</th>\n",
       "      <td>0.038238</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ladybird</th>\n",
       "      <td>0.027641</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Phantom Thread</th>\n",
       "      <td>0.026760</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Post</th>\n",
       "      <td>0.026586</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Shape of Water</th>\n",
       "      <td>0.032273</td>\n",
       "      <td>2017</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Three Billboards Outside Ebbing, Missouri</th>\n",
       "      <td>0.034474</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Black Panther</th>\n",
       "      <td>0.047965</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BlacKkKlansman</th>\n",
       "      <td>0.027050</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bohemian Rhapsody</th>\n",
       "      <td>0.033908</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Favourite</th>\n",
       "      <td>0.026600</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Green Book</th>\n",
       "      <td>0.028264</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Roma</th>\n",
       "      <td>0.026817</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Star Is Born</th>\n",
       "      <td>0.028545</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Vice</th>\n",
       "      <td>0.026500</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           Probability  Year  Won Oscar\n",
       "Arrival                                       0.050988  2016          0\n",
       "Fences                                        0.026981  2016          0\n",
       "Hacksaw Ridge                                 0.037200  2016          0\n",
       "Hell or High Water                            0.028102  2016          0\n",
       "Hidden Figures                                0.028228  2016          0\n",
       "La La Land                                    0.041534  2016          0\n",
       "Lion                                          0.028417  2016          0\n",
       "Manchester by the Sea                         0.029393  2016          0\n",
       "Moonlight                                     0.029331  2016          1\n",
       "Call Me by Your Name                          0.027758  2017          0\n",
       "Darkest Hour                                  0.027472  2017          0\n",
       "Dunkirk                                       0.044161  2017          0\n",
       "Get Out                                       0.038238  2017          0\n",
       "Ladybird                                      0.027641  2017          0\n",
       "Phantom Thread                                0.026760  2017          0\n",
       "The Post                                      0.026586  2017          0\n",
       "The Shape of Water                            0.032273  2017          1\n",
       "Three Billboards Outside Ebbing, Missouri     0.034474  2017          0\n",
       "Black Panther                                 0.047965  2018          0\n",
       "BlacKkKlansman                                0.027050  2018          0\n",
       "Bohemian Rhapsody                             0.033908  2018          0\n",
       "The Favourite                                 0.026600  2018          0\n",
       "Green Book                                    0.028264  2018          1\n",
       "Roma                                          0.026817  2018          0\n",
       "A Star Is Born                                0.028545  2018          0\n",
       "Vice                                          0.026500  2018          0"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "naive_prediction.tail(26)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "</br>\n",
    "<center>\n",
    "    <img src = predictions.jpg width = 400 />\n",
    "    *Please work...*\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>The navie bayes model is a start to our machine learning classification problem for predicting the Oscars. It did not give us good results with the test data and it also does not predict correct results. It predicts \"Bohemian Rhapsody\" as the Oscar winner rather than the actual winner \"Green Book\"</b> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Logistic Regression \n",
    "\n",
    "Logistic regression is the go-to method for binary classification problems (problems with two class values). We started off our Oscar prediction with trying out logistic regression."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Fitting Logistic Regression to the data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='warn',\n",
       "          n_jobs=None, penalty='l2', random_state=0, solver='lbfgs',\n",
       "          tol=0.0001, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "classifier_lr = LogisticRegression(random_state=0,solver=\"lbfgs\")\n",
    "classifier_lr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Assess Logistic Regression Performance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix for Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predicting the Test set results\n",
    "y_pred_lr = classifier_lr.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[85  0]\n",
      " [21  1]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "cm_lg = confusion_matrix(y_test, y_pred_lr)\n",
    "plot_confusion_matrix(cm_lg, classes = ['Lost Oscar', 'Won Oscar'],\n",
    "                      title = 'Oscar Confusion Matrix',cmap=plt.cm.Blues)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>This matrix shows that there are 85+1=86 correct predictions and only 21+1=22 incorrect predictions. The logistic regression model predicts 21 false negatives which means the movies which actually won the oscar but predicted that the movie lost and the model also predicted 0 false postives which means the movies which actually lost the oscar but predicted that the movie won the oscar. \n",
    "\n",
    "\n",
    "The logistic regression model does not give us great results. Too many false negatives and not correct predictions</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.8037383177570093\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "print(\"Accuracy\", metrics.accuracy_score(y_test, y_pred_lr))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Let's try the k-Fold Cross Validation is applied to improve the model performance</b>\n",
    "\n",
    "\"This approach involves randomly dividing the set of observations into k groups, or folds, of approximately equal size. The first fold is treated as a validation set, and the method is fit on the remaining k − 1 folds.\"\n",
    "\n",
    "— Page 181, An Introduction to Statistical Learning, 2013."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8327486559139785"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Applying k-Fold Cross Validation\n",
    "from sklearn.model_selection import cross_val_score\n",
    "accuracies_lr = cross_val_score(estimator = classifier_lr, X = X_train, y = y_train, cv = 10)\n",
    "accuracies_lr.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>The Logistic Regression performance shows 83% which is good in it's terms.</b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try and build a ROC curve to check the performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make probability predictions\n",
    "train_probs_lr = classifier_lr.predict_proba(X_train)[:, 1]\n",
    "probs_lr = classifier_lr.predict_proba(X_test)[:, 1]\n",
    "\n",
    "train_predictions_lr = classifier_lr.predict(X_train)\n",
    "predictions_lr = classifier_lr.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train ROC AUC Score: 0.5874614748552959\n",
      "Test ROC AUC  Score: 0.46310160427807484\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score, recall_score, roc_auc_score, roc_curve\n",
    "\n",
    "print(f'Train ROC AUC Score: {roc_auc_score(y_train, train_probs_lr)}')\n",
    "print(f'Test ROC AUC  Score: {roc_auc_score(y_test, probs_lr)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baseline ROC AUC: 0.5\n"
     ]
    }
   ],
   "source": [
    "print(f'Baseline ROC AUC: {roc_auc_score(y_test, [1 for _ in range(len(y_test))])}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall Baseline: 1.0 Test: 0.05 Train: 0.13\n",
      "Precision Baseline: 0.21 Test: 1.0 Train: 0.7\n",
      "Roc Baseline: 0.5 Test: 0.46 Train: 0.59\n"
     ]
    },
    {
     "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": [
    "evaluate_model(predictions_lr, probs_lr, train_predictions_lr, train_probs_lr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Our Logistic regression gives a score of 59 for training data but 46 for test which are bad results."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Finally the Logistic Regression Model predicts the Oscars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>category</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "      <th>MovieLensrating</th>\n",
       "      <th>runtimeMinutes</th>\n",
       "      <th>IMDBRating</th>\n",
       "      <th>imdb_numVotes</th>\n",
       "      <th>RottenCriticRating</th>\n",
       "      <th>rottencritic_numVotes</th>\n",
       "      <th>RottenRating</th>\n",
       "      <th>rotten_numVotes</th>\n",
       "      <th>won_bafta</th>\n",
       "      <th>won_sa</th>\n",
       "      <th>won_sag</th>\n",
       "      <th>won_cc</th>\n",
       "      <th>won_dga</th>\n",
       "      <th>won_nbra</th>\n",
       "      <th>won_oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>277</th>\n",
       "      <td>2000</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Chocolat</td>\n",
       "      <td>Drama|Romance</td>\n",
       "      <td>3.68</td>\n",
       "      <td>121</td>\n",
       "      <td>7.3</td>\n",
       "      <td>167453</td>\n",
       "      <td>62</td>\n",
       "      <td>117</td>\n",
       "      <td>83</td>\n",
       "      <td>328835</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>278</th>\n",
       "      <td>2000</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Crouching Tiger, Hidden Dragon</td>\n",
       "      <td>Action|Adventure|Fantasy</td>\n",
       "      <td>3.54</td>\n",
       "      <td>120</td>\n",
       "      <td>7.9</td>\n",
       "      <td>238676</td>\n",
       "      <td>97</td>\n",
       "      <td>153</td>\n",
       "      <td>86</td>\n",
       "      <td>427169</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>279</th>\n",
       "      <td>2000</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Erin Brockovich</td>\n",
       "      <td>Drama</td>\n",
       "      <td>3.58</td>\n",
       "      <td>131</td>\n",
       "      <td>7.3</td>\n",
       "      <td>157701</td>\n",
       "      <td>84</td>\n",
       "      <td>145</td>\n",
       "      <td>81</td>\n",
       "      <td>418491</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>280</th>\n",
       "      <td>2000</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Gladiator</td>\n",
       "      <td>Action|Adventure|Drama</td>\n",
       "      <td>3.96</td>\n",
       "      <td>155</td>\n",
       "      <td>8.5</td>\n",
       "      <td>1197072</td>\n",
       "      <td>76</td>\n",
       "      <td>194</td>\n",
       "      <td>87</td>\n",
       "      <td>34128074</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>281</th>\n",
       "      <td>2000</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Traffic</td>\n",
       "      <td>Crime|Drama|Thriller</td>\n",
       "      <td>3.80</td>\n",
       "      <td>147</td>\n",
       "      <td>7.6</td>\n",
       "      <td>185430</td>\n",
       "      <td>92</td>\n",
       "      <td>158</td>\n",
       "      <td>85</td>\n",
       "      <td>190305</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     year      category                           title  \\\n",
       "277  2000  BEST PICTURE                        Chocolat   \n",
       "278  2000  BEST PICTURE  Crouching Tiger, Hidden Dragon   \n",
       "279  2000  BEST PICTURE                 Erin Brockovich   \n",
       "280  2000  BEST PICTURE                       Gladiator   \n",
       "281  2000  BEST PICTURE                         Traffic   \n",
       "\n",
       "                       genres  MovieLensrating  runtimeMinutes  IMDBRating  \\\n",
       "277             Drama|Romance             3.68             121         7.3   \n",
       "278  Action|Adventure|Fantasy             3.54             120         7.9   \n",
       "279                     Drama             3.58             131         7.3   \n",
       "280    Action|Adventure|Drama             3.96             155         8.5   \n",
       "281      Crime|Drama|Thriller             3.80             147         7.6   \n",
       "\n",
       "     imdb_numVotes  RottenCriticRating  rottencritic_numVotes  RottenRating  \\\n",
       "277         167453                  62                    117            83   \n",
       "278         238676                  97                    153            86   \n",
       "279         157701                  84                    145            81   \n",
       "280        1197072                  76                    194            87   \n",
       "281         185430                  92                    158            85   \n",
       "\n",
       "     rotten_numVotes  won_bafta  won_sa  won_sag  won_cc  won_dga  won_nbra  \\\n",
       "277           328835          0       0        0       0        0         0   \n",
       "278           427169          0       1        0       0        1         1   \n",
       "279           418491          0       0        0       0        0         0   \n",
       "280         34128074          1       0        0       1        0         0   \n",
       "281           190305          0       1        1       0        0         0   \n",
       "\n",
       "     won_oscar  \n",
       "277          0  \n",
       "278          0  \n",
       "279          0  \n",
       "280          1  \n",
       "281          0  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Separate the dataset into train and test to predict the Oscar's\n",
    "train_lr = dataset[dataset['year']<=2000]\n",
    "train_lr.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year</th>\n",
       "      <th>category</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "      <th>MovieLensrating</th>\n",
       "      <th>runtimeMinutes</th>\n",
       "      <th>IMDBRating</th>\n",
       "      <th>imdb_numVotes</th>\n",
       "      <th>RottenCriticRating</th>\n",
       "      <th>rottencritic_numVotes</th>\n",
       "      <th>RottenRating</th>\n",
       "      <th>rotten_numVotes</th>\n",
       "      <th>won_bafta</th>\n",
       "      <th>won_sa</th>\n",
       "      <th>won_sag</th>\n",
       "      <th>won_cc</th>\n",
       "      <th>won_dga</th>\n",
       "      <th>won_nbra</th>\n",
       "      <th>won_oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>406</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>The Favourite</td>\n",
       "      <td>Drama|Comedy</td>\n",
       "      <td>3.82</td>\n",
       "      <td>119</td>\n",
       "      <td>7.7</td>\n",
       "      <td>103011</td>\n",
       "      <td>93</td>\n",
       "      <td>362</td>\n",
       "      <td>67</td>\n",
       "      <td>5754</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>407</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Green Book</td>\n",
       "      <td>Drama|Comedy</td>\n",
       "      <td>3.94</td>\n",
       "      <td>130</td>\n",
       "      <td>8.3</td>\n",
       "      <td>170827</td>\n",
       "      <td>78</td>\n",
       "      <td>316</td>\n",
       "      <td>92</td>\n",
       "      <td>7859</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>408</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Roma</td>\n",
       "      <td>Drama</td>\n",
       "      <td>3.76</td>\n",
       "      <td>135</td>\n",
       "      <td>7.8</td>\n",
       "      <td>100158</td>\n",
       "      <td>96</td>\n",
       "      <td>353</td>\n",
       "      <td>71</td>\n",
       "      <td>4667</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>409</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>A Star Is Born</td>\n",
       "      <td>Drama|Romance</td>\n",
       "      <td>3.67</td>\n",
       "      <td>135</td>\n",
       "      <td>7.8</td>\n",
       "      <td>229903</td>\n",
       "      <td>89</td>\n",
       "      <td>472</td>\n",
       "      <td>80</td>\n",
       "      <td>18019</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410</th>\n",
       "      <td>2018</td>\n",
       "      <td>BEST PICTURE</td>\n",
       "      <td>Vice</td>\n",
       "      <td>Drama|Comedy</td>\n",
       "      <td>3.49</td>\n",
       "      <td>132</td>\n",
       "      <td>7.2</td>\n",
       "      <td>58938</td>\n",
       "      <td>66</td>\n",
       "      <td>326</td>\n",
       "      <td>58</td>\n",
       "      <td>3902</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     year      category           title         genres  MovieLensrating  \\\n",
       "406  2018  BEST PICTURE   The Favourite   Drama|Comedy             3.82   \n",
       "407  2018  BEST PICTURE      Green Book   Drama|Comedy             3.94   \n",
       "408  2018  BEST PICTURE            Roma          Drama             3.76   \n",
       "409  2018  BEST PICTURE  A Star Is Born  Drama|Romance             3.67   \n",
       "410  2018  BEST PICTURE            Vice   Drama|Comedy             3.49   \n",
       "\n",
       "     runtimeMinutes  IMDBRating  imdb_numVotes  RottenCriticRating  \\\n",
       "406             119         7.7         103011                  93   \n",
       "407             130         8.3         170827                  78   \n",
       "408             135         7.8         100158                  96   \n",
       "409             135         7.8         229903                  89   \n",
       "410             132         7.2          58938                  66   \n",
       "\n",
       "     rottencritic_numVotes  RottenRating  rotten_numVotes  won_bafta  won_sa  \\\n",
       "406                    362            67             5754          0       1   \n",
       "407                    316            92             7859          0       0   \n",
       "408                    353            71             4667          1       1   \n",
       "409                    472            80            18019          0       1   \n",
       "410                    326            58             3902          0       0   \n",
       "\n",
       "     won_sag  won_cc  won_dga  won_nbra  won_oscar  \n",
       "406        0       0        0         0          0  \n",
       "407        0       0        0         1          1  \n",
       "408        0       1        1         0          0  \n",
       "409        0       0        0         0          0  \n",
       "410        0       0        0         0          0  "
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_lr = dataset[dataset['year']>2000]\n",
    "test_lr.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0,\n",
       "       0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1,\n",
       "       0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,\n",
       "       0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0,\n",
       "       0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0,\n",
       "       0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,\n",
       "       0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,\n",
       "       0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,\n",
       "       1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0,\n",
       "       0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#target varaiable which you want to predict\n",
    "target_lr = train_lr['won_oscar'].values\n",
    "target_lr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0. , 0. , 0. , 0. , 0. , 8.3],\n",
       "       [0. , 0. , 0. , 0. , 0. , 7.8],\n",
       "       [0. , 0. , 0. , 0. , 0. , 7.2],\n",
       "       ...,\n",
       "       [0. , 0. , 0. , 0. , 0. , 7.3],\n",
       "       [1. , 0. , 0. , 0. , 1. , 8.5],\n",
       "       [0. , 0. , 1. , 1. , 0. , 7.6]])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Define the feature names based on which you want the Oscar's to be predicted\n",
    "feature_names_lr = [\n",
    "    'won_bafta',\n",
    "    'won_dga',\n",
    "    'won_sa',\n",
    "    'won_sag',\n",
    "    'won_cc',\n",
    "    'IMDBRating'\n",
    "]\n",
    "features_lr = train_lr[feature_names_lr].values\n",
    "features_lr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='warn',\n",
       "          n_jobs=None, penalty='l2', random_state=0, solver='lbfgs',\n",
       "          tol=0.0001, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier_lr.fit(features_lr, target_lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_features_lr = test_lr[feature_names_lr].values\n",
    "pred_LR = classifier_lr.predict_proba(test_features_lr)[:,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "movie_name = np.array(test_lr['title'])\n",
    "year = np.array(test_lr['year'])\n",
    "oscar = np.array(test_lr['won_oscar'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "prediction = pd.DataFrame(pred_LR.round(2), movie_name,columns =[\"Probability\"] )\n",
    "prediction[\"year\"] = year\n",
    "prediction[\"Actually Oscar\"] = oscar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>Probability</th>\n",
       "      <th>year</th>\n",
       "      <th>Actually Oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Arrival</th>\n",
       "      <td>0.08</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fences</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hacksaw Ridge</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hell or High Water</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hidden Figures</th>\n",
       "      <td>0.08</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>La La Land</th>\n",
       "      <td>0.82</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lion</th>\n",
       "      <td>0.67</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Manchester by the Sea</th>\n",
       "      <td>0.05</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moonlight</th>\n",
       "      <td>0.06</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Call Me by Your Name</th>\n",
       "      <td>0.08</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Darkest Hour</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dunkirk</th>\n",
       "      <td>0.08</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Get Out</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ladybird</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Phantom Thread</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Post</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Shape of Water</th>\n",
       "      <td>0.62</td>\n",
       "      <td>2017</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Three Billboards Outside Ebbing, Missouri</th>\n",
       "      <td>0.16</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Black Panther</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BlacKkKlansman</th>\n",
       "      <td>0.05</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bohemian Rhapsody</th>\n",
       "      <td>0.08</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Favourite</th>\n",
       "      <td>0.05</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Green Book</th>\n",
       "      <td>0.08</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Roma</th>\n",
       "      <td>0.80</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Star Is Born</th>\n",
       "      <td>0.05</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Vice</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           Probability  year  Actually Oscar\n",
       "Arrival                                           0.08  2016               0\n",
       "Fences                                            0.07  2016               0\n",
       "Hacksaw Ridge                                     0.07  2016               0\n",
       "Hell or High Water                                0.07  2016               0\n",
       "Hidden Figures                                    0.08  2016               0\n",
       "La La Land                                        0.82  2016               0\n",
       "Lion                                              0.67  2016               0\n",
       "Manchester by the Sea                             0.05  2016               0\n",
       "Moonlight                                         0.06  2016               1\n",
       "Call Me by Your Name                              0.08  2017               0\n",
       "Darkest Hour                                      0.07  2017               0\n",
       "Dunkirk                                           0.08  2017               0\n",
       "Get Out                                           0.07  2017               0\n",
       "Ladybird                                          0.07  2017               0\n",
       "Phantom Thread                                    0.07  2017               0\n",
       "The Post                                          0.07  2017               0\n",
       "The Shape of Water                                0.62  2017               1\n",
       "Three Billboards Outside Ebbing, Missouri         0.16  2017               0\n",
       "Black Panther                                     0.07  2018               0\n",
       "BlacKkKlansman                                    0.05  2018               0\n",
       "Bohemian Rhapsody                                 0.08  2018               0\n",
       "The Favourite                                     0.05  2018               0\n",
       "Green Book                                        0.08  2018               1\n",
       "Roma                                              0.80  2018               0\n",
       "A Star Is Born                                    0.05  2018               0\n",
       "Vice                                              0.07  2018               0"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction.tail(26)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "</br>\n",
    "<center>\n",
    "    <img src = notplanned.jpg width = 400 />\n",
    "    Awkward....\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Well logistic regression model didn't go as planned either. It did not give us great results with the test data and it also does not predict correct results. It predicts \"Roma\" as the Oscar winner rather than the actual winner \"Green Book\"</b> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Decision Trees\n",
    "\n",
    "A decision tree is the building block of a random forest and is an intuitive model. We can think of a decision tree as a series of yes/no questions asked about our data eventually leading to a predicted class. \n",
    "\n",
    "Presumably, the odds of an Oscar win could depend on critic ratings,IMDb ratings,previous awards won(BAFTA,SAG,NBRA) or even the genre. According to the Atlantic (see references below), dramas are more likely to win than any other category. And maybe a win also depends on the time of year the movie was released, so if the movie debuted near Oscars nominations season, you might have a better shot\n",
    "\n",
    "https://www.theatlantic.com/entertainment/archive/2014/03/the-secrets-to-winning-an-academy-award/284158/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Fitting Decision Tree to the training set\n",
    "To create a decision tree and train (fit) it on the data, we use Scikit-Learn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(class_weight=None, criterion='entropy', max_depth=None,\n",
       "            max_features=None, max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, presort=False, random_state=0,\n",
       "            splitter='best')"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "classifier_dt = DecisionTreeClassifier(criterion = 'entropy',random_state = 0)\n",
    "classifier_dt.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision tree has 61 nodes with maximum depth 8.\n"
     ]
    }
   ],
   "source": [
    "print(f'Decision tree has {classifier_dt.tree_.node_count} nodes with maximum depth {classifier_dt.tree_.max_depth}.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Variable Importance based on model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variable: won_dga              Importance: 0.43\n",
      "Variable: rottencritic_numVotes Importance: 0.14\n",
      "Variable: runtimeMinutes       Importance: 0.09\n",
      "Variable: rotten_numVotes      Importance: 0.08\n",
      "Variable: RottenCriticRating   Importance: 0.07\n",
      "Variable: MovieLensrating      Importance: 0.05\n",
      "Variable: IMDBRating           Importance: 0.05\n",
      "Variable: imdb_numVotes        Importance: 0.05\n",
      "Variable: RottenRating         Importance: 0.05\n",
      "Variable: won_bafta            Importance: 0.0\n",
      "Variable: won_sa               Importance: 0.0\n",
      "Variable: won_sag              Importance: 0.0\n",
      "Variable: won_cc               Importance: 0.0\n"
     ]
    }
   ],
   "source": [
    "# Get numerical feature importances\n",
    "importances_tree = list(classifier_dt.feature_importances_)\n",
    "\n",
    "# List of tuples with variable and importance\n",
    "feature_importances = [(feature, round(importance, 2)) for feature, importance in zip(feature_list, importances_tree)]\n",
    "\n",
    "# Sort the feature importances by most important first\n",
    "feature_importances = sorted(feature_importances, key = lambda x: x[1], reverse = True)\n",
    "\n",
    "# Print out the feature and importances \n",
    "[print('Variable: {:20} Importance: {}'.format(*pair)) for pair in feature_importances];"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Visualization of the Decision Tree Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import tools needed for visualization\n",
    "from sklearn.tree import export_graphviz\n",
    "import pydot\n",
    "\n",
    "# Export the image to a dot file\n",
    "export_graphviz(classifier_dt, out_file = 'tree_oscars_dt.dot', feature_names = feature_list, class_names=['won','lost'],max_depth = 10,rounded = True, precision = 1)\n",
    "\n",
    "# Use dot file to create a graph\n",
    "(graph, ) = pydot.graph_from_dot_file('tree_oscars_dt.dot')\n",
    "\n",
    "# Write graph to a png file\n",
    "graph.write_png('tree_oscars_dt.png'); "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br />\n",
    "<center>\n",
    "<img src = tree_oscars_dt.png width = 1000 />\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Assess Decision Tree Performance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Confusion Matrix for Decision Tree </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predicting the Test set results\n",
    "y_pred_dt = classifier_dt.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Making the Confusion Matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "cm_dt = confusion_matrix(y_test, y_pred_dt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[75 10]\n",
      " [ 9 13]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(cm_dt, classes = ['Lost Oscar', 'Won Oscar'],\n",
    "                      title = 'Oscar Confusion Matrix',cmap=plt.cm.Blues)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> This matrix shows that there are 75+13=88 correct predictions and only 10+9=19 incorrect predictions.\n",
    "The decision tree model predicts 9 false negatives which means the movies which actually won the oscar but predicted that the movie lost and the model also predicted 10 false postives which means the movies which actually lost the oscar but predicted that the movie won the oscar. So in all, the model predicts more false positives which is bad. This could be improved using a Random Forest Classifier</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.822429906542056\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "print(\"Accuracy\", metrics.accuracy_score(y_test, y_pred_dt))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Let's try the k-Fold Cross Validation is applied to improve the model performance</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8714852150537634"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Applying k-Fold Cross Validation\n",
    "from sklearn.model_selection import cross_val_score\n",
    "accuracies_dt = cross_val_score(estimator = classifier_dt, X = X_train, y = y_train, cv = 10)\n",
    "accuracies_dt.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>The Decision tree performance is 87% which is decent.</b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "ROC curves typically feature true positive rate on the Y axis, and false positive rate on the X axis. This means that the top left corner of the plot is the “ideal” point - a false positive rate of zero, and a true positive rate of one. This is not very realistic, but it does mean that a larger area under the curve (AUC) is usually better.\n",
    "\n",
    "The “steepness” of ROC curves is also important, since it is ideal to maximize the true positive rate while minimizing the false positive rate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make probability predictions\n",
    "train_probs_dt = classifier_dt.predict_proba(X_train)[:, 1]\n",
    "probs_dt = classifier_dt.predict_proba(X_test)[:, 1]\n",
    "\n",
    "train_predictions_dt = classifier_dt.predict(X_train)\n",
    "predictions_dt = classifier_dt.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train ROC AUC Score: 1.0\n",
      "Test ROC AUC  Score: 0.7366310160427807\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score, recall_score, roc_auc_score, roc_curve\n",
    "\n",
    "print(f'Train ROC AUC Score: {roc_auc_score(y_train, train_probs_dt)}')\n",
    "print(f'Test ROC AUC  Score: {roc_auc_score(y_test, probs_dt)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baseline ROC AUC: 0.5\n"
     ]
    }
   ],
   "source": [
    "print(f'Baseline ROC AUC: {roc_auc_score(y_test, [1 for _ in range(len(y_test))])}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall Baseline: 1.0 Test: 0.59 Train: 1.0\n",
      "Precision Baseline: 0.21 Test: 0.57 Train: 1.0\n",
      "Roc Baseline: 0.5 Test: 0.74 Train: 1.0\n"
     ]
    },
    {
     "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": [
    "evaluate_model(predictions_dt, probs_dt, train_predictions_dt, train_probs_dt)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Our decision tree for Oscars achieves the perfect score for the training data. But the test data is not great, and can be seen that there is overfitting of the training data.</b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. Finally our Decision Tree Model predicts the Oscars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_features = dataset[feature_list].values\n",
    "\n",
    "pred_tree = classifier_dt.predict_proba(test_features)[:,1]\n",
    "movie_name = np.array(dataset['title'])\n",
    "year = np.array(dataset['year'])\n",
    "won_oscar = np.array(dataset['won_oscar'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_prediction = pd.DataFrame(pred_tree,movie_name,columns =[\"Probability\"] )\n",
    "tree_prediction[\"Year\"] = year \n",
    "tree_prediction[\"Won Oscar\"] = won_oscar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>Probability</th>\n",
       "      <th>Year</th>\n",
       "      <th>Won Oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Lion</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Manchester by the Sea</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moonlight</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Call Me by Your Name</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Darkest Hour</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dunkirk</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Get Out</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ladybird</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Phantom Thread</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Post</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Shape of Water</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Three Billboards Outside Ebbing, Missouri</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Black Panther</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BlacKkKlansman</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bohemian Rhapsody</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Favourite</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Green Book</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Roma</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Star Is Born</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Vice</th>\n",
       "      <td>0.0</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           Probability  Year  Won Oscar\n",
       "Lion                                               1.0  2016          0\n",
       "Manchester by the Sea                              0.0  2016          0\n",
       "Moonlight                                          1.0  2016          1\n",
       "Call Me by Your Name                               0.0  2017          0\n",
       "Darkest Hour                                       0.0  2017          0\n",
       "Dunkirk                                            0.0  2017          0\n",
       "Get Out                                            0.0  2017          0\n",
       "Ladybird                                           0.0  2017          0\n",
       "Phantom Thread                                     0.0  2017          0\n",
       "The Post                                           0.0  2017          0\n",
       "The Shape of Water                                 1.0  2017          1\n",
       "Three Billboards Outside Ebbing, Missouri          0.0  2017          0\n",
       "Black Panther                                      0.0  2018          0\n",
       "BlacKkKlansman                                     0.0  2018          0\n",
       "Bohemian Rhapsody                                  0.0  2018          0\n",
       "The Favourite                                      0.0  2018          0\n",
       "Green Book                                         1.0  2018          1\n",
       "Roma                                               0.0  2018          0\n",
       "A Star Is Born                                     0.0  2018          0\n",
       "Vice                                               0.0  2018          0"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree_prediction.tail(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "</br>\n",
    "<center>\n",
    "    <img src =working.gif width = 400 />\n",
    "    *Yassss...*\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>The decision tree predicts the correct Oscar winner and did way better than logisitc regression model.Andddd the Decision tree predicts \"Green Book\" as the winner.Yay!  </b>\n",
    "\n",
    "But not so good results with the model accuracy and performance. We could try to improve the results by trying another classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest Classifier \n",
    "\n",
    "The random forest combines hundreds or thousands of decision trees, trains each one on a slightly different set of the observations, splitting nodes in each tree considering a limited number of the features. The final predictions of the random forest are made by averaging the predictions of each individual tree.\n",
    "\n",
    "<b>Not great results using a decision tree as we can see above. Let's try using a Random Forest Classifier</b>\n",
    "\n",
    "Here we directly jump to the <b>3rd</b> step from above. As we have done the data pre-processing in the decision tree steps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Fitting Random Forest to the training set\n",
    "To create a decision tree and train (fit) it on the data, we use Scikit-Learn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='entropy',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=None,\n",
       "            oob_score=False, random_state=0, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fitting Random Forest Classification to the Training set\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "classifier_rf = RandomForestClassifier(n_estimators = 100,random_state = 0,criterion = 'entropy')\n",
    "classifier_rf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Variable Importance based on model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variable: won_dga              Importance: 0.31\n",
      "Variable: rottencritic_numVotes Importance: 0.12\n",
      "Variable: imdb_numVotes        Importance: 0.09\n",
      "Variable: rotten_numVotes      Importance: 0.09\n",
      "Variable: MovieLensrating      Importance: 0.07\n",
      "Variable: runtimeMinutes       Importance: 0.07\n",
      "Variable: RottenCriticRating   Importance: 0.07\n",
      "Variable: IMDBRating           Importance: 0.06\n",
      "Variable: RottenRating         Importance: 0.06\n",
      "Variable: won_bafta            Importance: 0.02\n",
      "Variable: won_cc               Importance: 0.02\n",
      "Variable: won_sa               Importance: 0.01\n",
      "Variable: won_sag              Importance: 0.01\n"
     ]
    }
   ],
   "source": [
    "# Get numerical feature importances\n",
    "importances_forest = list(classifier_rf.feature_importances_)\n",
    "\n",
    "# List of tuples with variable and importance\n",
    "feature_importances = [(feature, round(importance, 2)) for feature, importance in zip(feature_list, importances_forest)]\n",
    "\n",
    "# Sort the feature importances by most important first\n",
    "feature_importances = sorted(feature_importances, key = lambda x: x[1], reverse = True)\n",
    "\n",
    "# Print out the feature and importances \n",
    "[print('Variable: {:20} Importance: {}'.format(*pair)) for pair in feature_importances];"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "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": [
    "# list of x locations for plotting\n",
    "x_values = list(range(len(importances_forest)))\n",
    "\n",
    "# Make a horizontal bar chart\n",
    "plt.barh(x_values, importances_forest,color=\"blue\",align=\"center\")\n",
    "\n",
    "# Tick labels for y axis\n",
    "plt.yticks(x_values, feature_list)\n",
    "#plt.ylim(-0.5,11.5)\n",
    "\n",
    "# Axis labels and title\n",
    "plt.ylabel('Feature List'); plt.xlabel('Importance'); plt.title('Variable Importances'); "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Visualization of the Random Forest Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import tools needed for visualization\n",
    "from sklearn.tree import export_graphviz\n",
    "import pydot\n",
    "\n",
    "# Pull out one tree from the forest\n",
    "tree = classifier_rf.estimators_[13]\n",
    "\n",
    "# Export the image to a dot file\n",
    "export_graphviz(tree, out_file = 'tree_oscars_rf.dot', feature_names = feature_list, rounded = True, precision = 1)\n",
    "\n",
    "# Use dot file to create a graph\n",
    "(graph, ) = pydot.graph_from_dot_file('tree_oscars_rf.dot')\n",
    "\n",
    "# Write graph to a png file\n",
    "graph.write_png('tree_oscars_rf.png'); "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br />\n",
    "<center>\n",
    "<img src = tree_oscars_rf.png width = 1000 />\n",
    "</center>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The depth of this tree is: 14\n"
     ]
    }
   ],
   "source": [
    "print('The depth of this tree is:', tree.tree_.max_depth)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Limit depth of tree to 2 levels\n",
    "rf_small = RandomForestClassifier(n_estimators=10, max_depth = 4,criterion = 'entropy', random_state = 0)\n",
    "rf_small.fit(X_train, y_train)\n",
    "\n",
    "# Extract the small tree\n",
    "tree_small = rf_small.estimators_[5]\n",
    "\n",
    "# Save the tree as a png image\n",
    "export_graphviz(tree_small, out_file = 'small_tree_oscars.dot', feature_names = feature_list, class_names=['won','lost'],rounded = True, precision = 1)\n",
    "\n",
    "(graph, ) = pydot.graph_from_dot_file('small_tree_oscars.dot')\n",
    "\n",
    "graph.write_png('small_tree_oscars.png');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br/>\n",
    "<center>\n",
    "<img src = small_tree_oscars.png width = 1000 />\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Assess Random Forest Performance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Confusion Matrix for Random Forest </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predicting the Test set results\n",
    "y_pred_rf = classifier_rf.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Making the Confusion Matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "cm_rf = confusion_matrix(y_test, y_pred_rf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[82  3]\n",
      " [11 11]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(cm_rf, classes = ['Lost Oscar', 'Won Oscar'],\n",
    "                      title = 'Oscar Confusion Matrix',cmap=plt.cm.Blues)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> This matrix shows that there are 82+11=93 correct predictions and only 11+3=14 incorrect predictions. This seems to be a better model than the decision tree.\n",
    "\n",
    "The random forest model predicts 11 false negatives which means the movies which actually won the oscar but predicted that the movie lost and the model also predicted 3 false postives which means the movies which actually lost the oscar but predicted that the movie won the oscar.\n",
    "\n",
    "Using the random forest model the false positives are lesser,which means our model is headed in the right direction and is doing better than the decision tree model</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.8691588785046729\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "print(\"Accuracy\", metrics.accuracy_score(y_test, y_pred_rf))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Let's try the k-Fold Cross Validation is applied to improve the model performance</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9009543010752689"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Applying k-Fold Cross Validation\n",
    "from sklearn.model_selection import cross_val_score\n",
    "accuracies_rf = cross_val_score(estimator = classifier_rf, X = X_train, y = y_train, cv = 10)\n",
    "accuracies_rf.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>The Random Forest performance is 90% which is better than the Decision tree's model.</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make probability predictions\n",
    "train_probs_rf = classifier_rf.predict_proba(X_train)[:, 1]\n",
    "probs_rf = classifier_rf.predict_proba(X_test)[:, 1]\n",
    "\n",
    "train_predictions_rf = classifier_rf.predict(X_train)\n",
    "predictions_rf = classifier_rf.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train ROC AUC Score: 1.0\n",
      "Test ROC AUC  Score: 0.9379679144385027\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score, recall_score, roc_auc_score, roc_curve\n",
    "\n",
    "print(f'Train ROC AUC Score: {roc_auc_score(y_train, train_probs_rf)}')\n",
    "print(f'Test ROC AUC  Score: {roc_auc_score(y_test, probs_rf)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baseline ROC AUC: 0.5\n"
     ]
    }
   ],
   "source": [
    "print(f'Baseline ROC AUC: {roc_auc_score(y_test, [1 for _ in range(len(y_test))])}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall Baseline: 1.0 Test: 0.5 Train: 1.0\n",
      "Precision Baseline: 0.21 Test: 0.79 Train: 1.0\n",
      "Roc Baseline: 0.5 Test: 0.94 Train: 1.0\n"
     ]
    },
    {
     "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": [
    "evaluate_model(predictions_rf, probs_rf, train_predictions_rf, train_probs_rf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Sooo, our random forest model has done wayy better than the deicison tree model. The training data as usual is perfect, but the testing data has also done a better with a Roc of 0.94. Also the AUC is greater which means its a better model for our Oscar data.</b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. Finally our Decision Tree Model predicts the Oscars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_features = dataset[feature_list].values\n",
    "\n",
    "pred_forest = classifier_rf.predict_proba(test_features)[:,1]\n",
    "movie_name = np.array(dataset['title'])\n",
    "year = np.array(dataset['year'])\n",
    "won_oscar = np.array(dataset['won_oscar'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "forest_prediction = pd.DataFrame(pred_forest,movie_name,columns =[\"Probability\"] )\n",
    "forest_prediction[\"Year\"] = year \n",
    "forest_prediction[\"Won Oscar\"] = won_oscar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "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>Probability</th>\n",
       "      <th>Year</th>\n",
       "      <th>Won Oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Lion</th>\n",
       "      <td>0.65</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Manchester by the Sea</th>\n",
       "      <td>0.02</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moonlight</th>\n",
       "      <td>0.69</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Call Me by Your Name</th>\n",
       "      <td>0.01</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Darkest Hour</th>\n",
       "      <td>0.03</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dunkirk</th>\n",
       "      <td>0.01</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Get Out</th>\n",
       "      <td>0.02</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ladybird</th>\n",
       "      <td>0.06</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Phantom Thread</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Post</th>\n",
       "      <td>0.09</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Shape of Water</th>\n",
       "      <td>0.70</td>\n",
       "      <td>2017</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Three Billboards Outside Ebbing, Missouri</th>\n",
       "      <td>0.11</td>\n",
       "      <td>2017</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Black Panther</th>\n",
       "      <td>0.02</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BlacKkKlansman</th>\n",
       "      <td>0.02</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bohemian Rhapsody</th>\n",
       "      <td>0.08</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Favourite</th>\n",
       "      <td>0.11</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Green Book</th>\n",
       "      <td>0.55</td>\n",
       "      <td>2018</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Roma</th>\n",
       "      <td>0.16</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Star Is Born</th>\n",
       "      <td>0.04</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Vice</th>\n",
       "      <td>0.20</td>\n",
       "      <td>2018</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           Probability  Year  Won Oscar\n",
       "Lion                                              0.65  2016          0\n",
       "Manchester by the Sea                             0.02  2016          0\n",
       "Moonlight                                         0.69  2016          1\n",
       "Call Me by Your Name                              0.01  2017          0\n",
       "Darkest Hour                                      0.03  2017          0\n",
       "Dunkirk                                           0.01  2017          0\n",
       "Get Out                                           0.02  2017          0\n",
       "Ladybird                                          0.06  2017          0\n",
       "Phantom Thread                                    0.07  2017          0\n",
       "The Post                                          0.09  2017          0\n",
       "The Shape of Water                                0.70  2017          1\n",
       "Three Billboards Outside Ebbing, Missouri         0.11  2017          0\n",
       "Black Panther                                     0.02  2018          0\n",
       "BlacKkKlansman                                    0.02  2018          0\n",
       "Bohemian Rhapsody                                 0.08  2018          0\n",
       "The Favourite                                     0.11  2018          0\n",
       "Green Book                                        0.55  2018          1\n",
       "Roma                                              0.16  2018          0\n",
       "A Star Is Born                                    0.04  2018          0\n",
       "Vice                                              0.20  2018          0"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest_prediction.tail(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Model Accuracy is better and the Oscar Prediction shows the correct result!.\n",
    "    \n",
    "    And we predict Green Book as the winner with a higher probability.\n",
    "    \n",
    " *Let us check for the ever so famous \"La La Land\" and \"Moon Light\" for 2016*</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "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>Probability</th>\n",
       "      <th>Year</th>\n",
       "      <th>Won Oscar</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Arrival</th>\n",
       "      <td>0.04</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fences</th>\n",
       "      <td>0.01</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hacksaw Ridge</th>\n",
       "      <td>0.07</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hell or High Water</th>\n",
       "      <td>0.01</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hidden Figures</th>\n",
       "      <td>0.02</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>La La Land</th>\n",
       "      <td>0.19</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lion</th>\n",
       "      <td>0.65</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Manchester by the Sea</th>\n",
       "      <td>0.02</td>\n",
       "      <td>2016</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Moonlight</th>\n",
       "      <td>0.69</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Probability  Year  Won Oscar\n",
       "Arrival                       0.04  2016          0\n",
       "Fences                        0.01  2016          0\n",
       "Hacksaw Ridge                 0.07  2016          0\n",
       "Hell or High Water            0.01  2016          0\n",
       "Hidden Figures                0.02  2016          0\n",
       "La La Land                    0.19  2016          0\n",
       "Lion                          0.65  2016          0\n",
       "Manchester by the Sea         0.02  2016          0\n",
       "Moonlight                     0.69  2016          1"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest_prediction.loc[forest_prediction['Year'] == 2016]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusion\n",
    "\n",
    "<b>The random forest still predicts Moonlight as the Oscar Winner.* Well the data cannot go wrong and we users really were rooting for La La Land to win. According to the data our model is correct.\n",
    "\n",
    "\n",
    "So the Random Forest predicts \"Green Book\" as the Best Picture winner. \n",
    "\n",
    "We have predicted the correct Oscar winners for the 2019 which is \"Green Book\". Not to brag our model also predicted that 2018 winners \"The Shape of Water\".\n",
    "\n",
    "    All in all we predicted pretty good results with training the machine from data all the way from 1944.\n",
    "\n",
    "</b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "<center>\n",
    "    <img src =giphy.gif width = 600 />\n",
    "    *It was all worth it...*\n",
    "</center>\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "hide_code_all_hidden": false,
  "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}