source_code.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import Normalizer,StandardScaler\n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Data Directory Path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dir='6a62a5b4967411ea/Dataset/'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Load Training and Test Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data=pd.read_csv(os.path.join(data_dir,'Train.csv'))\n",
    "test_data=pd.read_csv(os.path.join(data_dir,'Test.csv'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Print first 5 rows of training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>INCIDENT_ID</th>\n",
       "      <th>DATE</th>\n",
       "      <th>X_1</th>\n",
       "      <th>X_2</th>\n",
       "      <th>X_3</th>\n",
       "      <th>X_4</th>\n",
       "      <th>X_5</th>\n",
       "      <th>X_6</th>\n",
       "      <th>X_7</th>\n",
       "      <th>X_8</th>\n",
       "      <th>X_9</th>\n",
       "      <th>X_10</th>\n",
       "      <th>X_11</th>\n",
       "      <th>X_12</th>\n",
       "      <th>X_13</th>\n",
       "      <th>X_14</th>\n",
       "      <th>X_15</th>\n",
       "      <th>MULTIPLE_OFFENSE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>CR_102659</td>\n",
       "      <td>04-JUL-04</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>34</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>92</td>\n",
       "      <td>29</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>CR_189752</td>\n",
       "      <td>18-JUL-17</td>\n",
       "      <td>1</td>\n",
       "      <td>37</td>\n",
       "      <td>37</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>236</td>\n",
       "      <td>1.0</td>\n",
       "      <td>103</td>\n",
       "      <td>142</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>CR_184637</td>\n",
       "      <td>15-MAR-17</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>110</td>\n",
       "      <td>93</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>CR_139071</td>\n",
       "      <td>13-FEB-09</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>249</td>\n",
       "      <td>1.0</td>\n",
       "      <td>72</td>\n",
       "      <td>29</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>CR_109335</td>\n",
       "      <td>13-APR-05</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>0.0</td>\n",
       "      <td>112</td>\n",
       "      <td>29</td>\n",
       "      <td>43</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  INCIDENT_ID       DATE  X_1  X_2  X_3  X_4  X_5  X_6  X_7  X_8  X_9  X_10  \\\n",
       "0   CR_102659  04-JUL-04    0   36   34    2    1    5    6    1    6     1   \n",
       "1   CR_189752  18-JUL-17    1   37   37    0    0   11   17    1    6     1   \n",
       "2   CR_184637  15-MAR-17    0    3    2    3    5    1    0    2    3     1   \n",
       "3   CR_139071  13-FEB-09    0   33   32    2    1    7    1    1    6     1   \n",
       "4   CR_109335  13-APR-05    0   33   32    2    1    8    3    0    5     1   \n",
       "\n",
       "   X_11  X_12  X_13  X_14  X_15  MULTIPLE_OFFENSE  \n",
       "0   174   1.0    92    29    36                 0  \n",
       "1   236   1.0   103   142    34                 1  \n",
       "2   174   1.0   110    93    34                 1  \n",
       "3   249   1.0    72    29    34                 1  \n",
       "4   174   0.0   112    29    43                 1  "
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Print Shape of Training and Test Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(23856, 18)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(15903, 17)"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(train_data.shape)\n",
    "test_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Test Data have one less column than train data,which is obvious as test data does not have the target column."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Check if there are any null values in the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "INCIDENT_ID           0\n",
       "DATE                  0\n",
       "X_1                   0\n",
       "X_2                   0\n",
       "X_3                   0\n",
       "X_4                   0\n",
       "X_5                   0\n",
       "X_6                   0\n",
       "X_7                   0\n",
       "X_8                   0\n",
       "X_9                   0\n",
       "X_10                  0\n",
       "X_11                  0\n",
       "X_12                182\n",
       "X_13                  0\n",
       "X_14                  0\n",
       "X_15                  0\n",
       "MULTIPLE_OFFENSE      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "INCIDENT_ID      0\n",
       "DATE             0\n",
       "X_1              0\n",
       "X_2              0\n",
       "X_3              0\n",
       "X_4              0\n",
       "X_5              0\n",
       "X_6              0\n",
       "X_7              0\n",
       "X_8              0\n",
       "X_9              0\n",
       "X_10             0\n",
       "X_11             0\n",
       "X_12           127\n",
       "X_13             0\n",
       "X_14             0\n",
       "X_15             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### See the description of Training Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>X_1</th>\n",
       "      <th>X_2</th>\n",
       "      <th>X_3</th>\n",
       "      <th>X_4</th>\n",
       "      <th>X_5</th>\n",
       "      <th>X_6</th>\n",
       "      <th>X_7</th>\n",
       "      <th>X_8</th>\n",
       "      <th>X_9</th>\n",
       "      <th>X_10</th>\n",
       "      <th>X_11</th>\n",
       "      <th>X_12</th>\n",
       "      <th>X_13</th>\n",
       "      <th>X_14</th>\n",
       "      <th>X_15</th>\n",
       "      <th>MULTIPLE_OFFENSE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>count</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23674.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "      <td>23856.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>mean</td>\n",
       "      <td>0.483778</td>\n",
       "      <td>24.791206</td>\n",
       "      <td>24.637450</td>\n",
       "      <td>4.276744</td>\n",
       "      <td>2.455609</td>\n",
       "      <td>6.154175</td>\n",
       "      <td>4.876509</td>\n",
       "      <td>0.972460</td>\n",
       "      <td>4.924128</td>\n",
       "      <td>1.244802</td>\n",
       "      <td>206.954519</td>\n",
       "      <td>0.974064</td>\n",
       "      <td>85.237383</td>\n",
       "      <td>72.674296</td>\n",
       "      <td>33.464747</td>\n",
       "      <td>0.955231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>std</td>\n",
       "      <td>1.439738</td>\n",
       "      <td>15.240231</td>\n",
       "      <td>15.135093</td>\n",
       "      <td>2.944672</td>\n",
       "      <td>1.963095</td>\n",
       "      <td>4.471756</td>\n",
       "      <td>3.881931</td>\n",
       "      <td>1.453144</td>\n",
       "      <td>1.362625</td>\n",
       "      <td>1.119301</td>\n",
       "      <td>93.033348</td>\n",
       "      <td>1.167725</td>\n",
       "      <td>27.597226</td>\n",
       "      <td>43.297320</td>\n",
       "      <td>8.386834</td>\n",
       "      <td>0.206800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>min</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>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</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",
       "      <td>25%</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>174.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>72.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50%</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>249.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>62.000000</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>75%</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>249.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>103.000000</td>\n",
       "      <td>107.000000</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>max</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>52.000000</td>\n",
       "      <td>52.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>99.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>332.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>116.000000</td>\n",
       "      <td>142.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                X_1           X_2           X_3           X_4           X_5  \\\n",
       "count  23856.000000  23856.000000  23856.000000  23856.000000  23856.000000   \n",
       "mean       0.483778     24.791206     24.637450      4.276744      2.455609   \n",
       "std        1.439738     15.240231     15.135093      2.944672      1.963095   \n",
       "min        0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%        0.000000      7.000000      8.000000      2.000000      1.000000   \n",
       "50%        0.000000     24.000000     24.000000      4.000000      3.000000   \n",
       "75%        0.000000     36.000000     35.000000      6.000000      5.000000   \n",
       "max        7.000000     52.000000     52.000000     10.000000      5.000000   \n",
       "\n",
       "                X_6           X_7           X_8           X_9          X_10  \\\n",
       "count  23856.000000  23856.000000  23856.000000  23856.000000  23856.000000   \n",
       "mean       6.154175      4.876509      0.972460      4.924128      1.244802   \n",
       "std        4.471756      3.881931      1.453144      1.362625      1.119301   \n",
       "min        1.000000      0.000000      0.000000      0.000000      1.000000   \n",
       "25%        3.000000      2.000000      0.000000      5.000000      1.000000   \n",
       "50%        5.000000      4.000000      1.000000      5.000000      1.000000   \n",
       "75%        8.000000      7.000000      1.000000      6.000000      1.000000   \n",
       "max       19.000000     18.000000     99.000000      6.000000     90.000000   \n",
       "\n",
       "               X_11          X_12          X_13          X_14          X_15  \\\n",
       "count  23856.000000  23674.000000  23856.000000  23856.000000  23856.000000   \n",
       "mean     206.954519      0.974064     85.237383     72.674296     33.464747   \n",
       "std       93.033348      1.167725     27.597226     43.297320      8.386834   \n",
       "min        0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%      174.000000      1.000000     72.000000     29.000000     34.000000   \n",
       "50%      249.000000      1.000000     98.000000     62.000000     34.000000   \n",
       "75%      249.000000      1.000000    103.000000    107.000000     34.000000   \n",
       "max      332.000000     90.000000    116.000000    142.000000     50.000000   \n",
       "\n",
       "       MULTIPLE_OFFENSE  \n",
       "count      23856.000000  \n",
       "mean           0.955231  \n",
       "std            0.206800  \n",
       "min            0.000000  \n",
       "25%            1.000000  \n",
       "50%            1.000000  \n",
       "75%            1.000000  \n",
       "max            1.000000  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Fetch day, month and year from the date column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fetch_day(sample):\n",
    "    return sample[:2]\n",
    "def fetch_month(sample):\n",
    "    return sample[3:6]\n",
    "def fetch_year(sample):\n",
    "    return sample[7:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data['day']=train_data['DATE'].apply(fetch_day)\n",
    "test_data['day']=test_data['DATE'].apply(fetch_day)\n",
    "\n",
    "train_data['month']=train_data['DATE'].apply(fetch_month)\n",
    "test_data['month']=test_data['DATE'].apply(fetch_month)\n",
    "\n",
    "train_data['year']=train_data['DATE'].apply(fetch_year)\n",
    "test_data['year']=test_data['DATE'].apply(fetch_year)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data.drop(['DATE'],axis=1,inplace=True)\n",
    "test_data.drop(['DATE'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "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>INCIDENT_ID</th>\n",
       "      <th>X_1</th>\n",
       "      <th>X_2</th>\n",
       "      <th>X_3</th>\n",
       "      <th>X_4</th>\n",
       "      <th>X_5</th>\n",
       "      <th>X_6</th>\n",
       "      <th>X_7</th>\n",
       "      <th>X_8</th>\n",
       "      <th>X_9</th>\n",
       "      <th>X_10</th>\n",
       "      <th>X_11</th>\n",
       "      <th>X_12</th>\n",
       "      <th>X_13</th>\n",
       "      <th>X_14</th>\n",
       "      <th>X_15</th>\n",
       "      <th>MULTIPLE_OFFENSE</th>\n",
       "      <th>day</th>\n",
       "      <th>month</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>CR_102659</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>34</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>92</td>\n",
       "      <td>29</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "      <td>04</td>\n",
       "      <td>JUL</td>\n",
       "      <td>04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>CR_189752</td>\n",
       "      <td>1</td>\n",
       "      <td>37</td>\n",
       "      <td>37</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>236</td>\n",
       "      <td>1.0</td>\n",
       "      <td>103</td>\n",
       "      <td>142</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "      <td>18</td>\n",
       "      <td>JUL</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>CR_184637</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>110</td>\n",
       "      <td>93</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>MAR</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>CR_139071</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>249</td>\n",
       "      <td>1.0</td>\n",
       "      <td>72</td>\n",
       "      <td>29</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "      <td>13</td>\n",
       "      <td>FEB</td>\n",
       "      <td>09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>CR_109335</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>0.0</td>\n",
       "      <td>112</td>\n",
       "      <td>29</td>\n",
       "      <td>43</td>\n",
       "      <td>1</td>\n",
       "      <td>13</td>\n",
       "      <td>APR</td>\n",
       "      <td>05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  INCIDENT_ID  X_1  X_2  X_3  X_4  X_5  X_6  X_7  X_8  X_9  X_10  X_11  X_12  \\\n",
       "0   CR_102659    0   36   34    2    1    5    6    1    6     1   174   1.0   \n",
       "1   CR_189752    1   37   37    0    0   11   17    1    6     1   236   1.0   \n",
       "2   CR_184637    0    3    2    3    5    1    0    2    3     1   174   1.0   \n",
       "3   CR_139071    0   33   32    2    1    7    1    1    6     1   249   1.0   \n",
       "4   CR_109335    0   33   32    2    1    8    3    0    5     1   174   0.0   \n",
       "\n",
       "   X_13  X_14  X_15  MULTIPLE_OFFENSE day month year  \n",
       "0    92    29    36                 0  04   JUL   04  \n",
       "1   103   142    34                 1  18   JUL   17  \n",
       "2   110    93    34                 1  15   MAR   17  \n",
       "3    72    29    34                 1  13   FEB   09  \n",
       "4   112    29    43                 1  13   APR   05  "
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### change month into numeric form"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data['month']=pd.to_datetime(train_data['month'], format='%b').dt.month\n",
    "test_data['month']=pd.to_datetime(test_data['month'], format='%b').dt.month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "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>INCIDENT_ID</th>\n",
       "      <th>X_1</th>\n",
       "      <th>X_2</th>\n",
       "      <th>X_3</th>\n",
       "      <th>X_4</th>\n",
       "      <th>X_5</th>\n",
       "      <th>X_6</th>\n",
       "      <th>X_7</th>\n",
       "      <th>X_8</th>\n",
       "      <th>X_9</th>\n",
       "      <th>X_10</th>\n",
       "      <th>X_11</th>\n",
       "      <th>X_12</th>\n",
       "      <th>X_13</th>\n",
       "      <th>X_14</th>\n",
       "      <th>X_15</th>\n",
       "      <th>MULTIPLE_OFFENSE</th>\n",
       "      <th>day</th>\n",
       "      <th>month</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>CR_102659</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>34</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>92</td>\n",
       "      <td>29</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "      <td>04</td>\n",
       "      <td>7</td>\n",
       "      <td>04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>CR_189752</td>\n",
       "      <td>1</td>\n",
       "      <td>37</td>\n",
       "      <td>37</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>236</td>\n",
       "      <td>1.0</td>\n",
       "      <td>103</td>\n",
       "      <td>142</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "      <td>18</td>\n",
       "      <td>7</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>CR_184637</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>110</td>\n",
       "      <td>93</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "      <td>15</td>\n",
       "      <td>3</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>CR_139071</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>249</td>\n",
       "      <td>1.0</td>\n",
       "      <td>72</td>\n",
       "      <td>29</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "      <td>13</td>\n",
       "      <td>2</td>\n",
       "      <td>09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>CR_109335</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>0.0</td>\n",
       "      <td>112</td>\n",
       "      <td>29</td>\n",
       "      <td>43</td>\n",
       "      <td>1</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  INCIDENT_ID  X_1  X_2  X_3  X_4  X_5  X_6  X_7  X_8  X_9  X_10  X_11  X_12  \\\n",
       "0   CR_102659    0   36   34    2    1    5    6    1    6     1   174   1.0   \n",
       "1   CR_189752    1   37   37    0    0   11   17    1    6     1   236   1.0   \n",
       "2   CR_184637    0    3    2    3    5    1    0    2    3     1   174   1.0   \n",
       "3   CR_139071    0   33   32    2    1    7    1    1    6     1   249   1.0   \n",
       "4   CR_109335    0   33   32    2    1    8    3    0    5     1   174   0.0   \n",
       "\n",
       "   X_13  X_14  X_15  MULTIPLE_OFFENSE day  month year  \n",
       "0    92    29    36                 0  04      7   04  \n",
       "1   103   142    34                 1  18      7   17  \n",
       "2   110    93    34                 1  15      3   17  \n",
       "3    72    29    34                 1  13      2   09  \n",
       "4   112    29    43                 1  13      4   05  "
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Plot Histogram for all numerical columns of training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_data=train_data.select_dtypes(include=['int','float'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_data=num_data.drop('MULTIPLE_OFFENSE',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib64/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:307: MatplotlibDeprecationWarning: \n",
      "The rowNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().rowspan.start instead.\n",
      "  layout[ax.rowNum, ax.colNum] = ax.get_visible()\n",
      "/usr/local/lib64/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:307: MatplotlibDeprecationWarning: \n",
      "The colNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().colspan.start instead.\n",
      "  layout[ax.rowNum, ax.colNum] = ax.get_visible()\n",
      "/usr/local/lib64/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:313: MatplotlibDeprecationWarning: \n",
      "The rowNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().rowspan.start instead.\n",
      "  if not layout[ax.rowNum + 1, ax.colNum]:\n",
      "/usr/local/lib64/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:313: MatplotlibDeprecationWarning: \n",
      "The colNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().colspan.start instead.\n",
      "  if not layout[ax.rowNum + 1, ax.colNum]:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "num_data.hist(bins=50,figsize=(20,15))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### As you can see from description result and histograms that data colums have different range and does not follow normal distribution.\n",
    "#### We need to standardize the data but before that we need to handle missing values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### As you can see from the graph of X_12 , data is skewed so we will impute missing values with median."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "median_X12=train_data['X_12'].median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data['X_12']=train_data['X_12'].fillna(median_X12)\n",
    "test_data['X_12']=test_data['X_12'].fillna(median_X12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "INCIDENT_ID         0\n",
       "X_1                 0\n",
       "X_2                 0\n",
       "X_3                 0\n",
       "X_4                 0\n",
       "X_5                 0\n",
       "X_6                 0\n",
       "X_7                 0\n",
       "X_8                 0\n",
       "X_9                 0\n",
       "X_10                0\n",
       "X_11                0\n",
       "X_12                0\n",
       "X_13                0\n",
       "X_14                0\n",
       "X_15                0\n",
       "MULTIPLE_OFFENSE    0\n",
       "day                 0\n",
       "month               0\n",
       "year                0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "INCIDENT_ID    0\n",
       "X_1            0\n",
       "X_2            0\n",
       "X_3            0\n",
       "X_4            0\n",
       "X_5            0\n",
       "X_6            0\n",
       "X_7            0\n",
       "X_8            0\n",
       "X_9            0\n",
       "X_10           0\n",
       "X_11           0\n",
       "X_12           0\n",
       "X_13           0\n",
       "X_14           0\n",
       "X_15           0\n",
       "day            0\n",
       "month          0\n",
       "year           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Now there are no missing values in train and test data, we will apply standardization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "X=train_data.drop(['MULTIPLE_OFFENSE','INCIDENT_ID'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "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>X_1</th>\n",
       "      <th>X_2</th>\n",
       "      <th>X_3</th>\n",
       "      <th>X_4</th>\n",
       "      <th>X_5</th>\n",
       "      <th>X_6</th>\n",
       "      <th>X_7</th>\n",
       "      <th>X_8</th>\n",
       "      <th>X_9</th>\n",
       "      <th>X_10</th>\n",
       "      <th>X_11</th>\n",
       "      <th>X_12</th>\n",
       "      <th>X_13</th>\n",
       "      <th>X_14</th>\n",
       "      <th>X_15</th>\n",
       "      <th>day</th>\n",
       "      <th>month</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>34</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>92</td>\n",
       "      <td>29</td>\n",
       "      <td>36</td>\n",
       "      <td>04</td>\n",
       "      <td>7</td>\n",
       "      <td>04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>37</td>\n",
       "      <td>37</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>236</td>\n",
       "      <td>1.0</td>\n",
       "      <td>103</td>\n",
       "      <td>142</td>\n",
       "      <td>34</td>\n",
       "      <td>18</td>\n",
       "      <td>7</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>1.0</td>\n",
       "      <td>110</td>\n",
       "      <td>93</td>\n",
       "      <td>34</td>\n",
       "      <td>15</td>\n",
       "      <td>3</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>249</td>\n",
       "      <td>1.0</td>\n",
       "      <td>72</td>\n",
       "      <td>29</td>\n",
       "      <td>34</td>\n",
       "      <td>13</td>\n",
       "      <td>2</td>\n",
       "      <td>09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>32</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>174</td>\n",
       "      <td>0.0</td>\n",
       "      <td>112</td>\n",
       "      <td>29</td>\n",
       "      <td>43</td>\n",
       "      <td>13</td>\n",
       "      <td>4</td>\n",
       "      <td>05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   X_1  X_2  X_3  X_4  X_5  X_6  X_7  X_8  X_9  X_10  X_11  X_12  X_13  X_14  \\\n",
       "0    0   36   34    2    1    5    6    1    6     1   174   1.0    92    29   \n",
       "1    1   37   37    0    0   11   17    1    6     1   236   1.0   103   142   \n",
       "2    0    3    2    3    5    1    0    2    3     1   174   1.0   110    93   \n",
       "3    0   33   32    2    1    7    1    1    6     1   249   1.0    72    29   \n",
       "4    0   33   32    2    1    8    3    0    5     1   174   0.0   112    29   \n",
       "\n",
       "   X_15 day  month year  \n",
       "0    36  04      7   04  \n",
       "1    34  18      7   17  \n",
       "2    34  15      3   17  \n",
       "3    34  13      2   09  \n",
       "4    43  13      4   05  "
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler=StandardScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaled_data=scaler.fit_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaled_test_data=test_data.drop(['INCIDENT_ID'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaled_test_data=scaler.transform(scaled_test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df=pd.DataFrame(scaled_data,columns=X.columns)\n",
    "test_final=pd.DataFrame(scaled_test_data,columns=X.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Split the train_df ino train and test data so that we can calculate metrics on test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df,test_df,train_labels,test_labels=train_test_split(train_df,train_data['MULTIPLE_OFFENSE'],test_size=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Apply Logitsic Regression Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr=LogisticRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib64/python3.7/site-packages/sklearn/linear_model/logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n",
      "  FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
       "                   multi_class='warn', n_jobs=None, penalty='l2',\n",
       "                   random_state=None, solver='warn', tol=0.0001, verbose=0,\n",
       "                   warm_start=False)"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr.fit(train_df,train_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_test=lr.predict(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions=lr.predict(test_final)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculate Precision , Recall and Accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score,precision_score,recall_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9545264040234702"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(test_labels,predictions_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9572451558550968"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "precision_score(test_labels,predictions_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9969291511296337"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recall_score(test_labels,predictions_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Print Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  10,  203],\n",
       "       [  14, 4545]])"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(test_labels,predictions_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Save test predictions to submission file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission=pd.DataFrame({'INCIDENT_ID':test_data.INCIDENT_ID,\"MULTIPLE_OFFENSE\":predictions})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission.to_csv(\"submission_one.csv\",index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Apply gradient boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "gb=GradientBoostingClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "                           learning_rate=0.1, loss='deviance', max_depth=3,\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, n_estimators=100,\n",
       "                           n_iter_no_change=None, presort='auto',\n",
       "                           random_state=None, subsample=1.0, tol=0.0001,\n",
       "                           validation_fraction=0.1, verbose=0,\n",
       "                           warm_start=False)"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gb.fit(train_df,train_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_test=lr.predict(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9545264040234702"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(test_labels,predictions_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 439,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9592716211077781"
      ]
     },
     "execution_count": 439,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "precision_score(train_data['MULTIPLE_OFFENSE'],predictions_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 440,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9963577321397227"
      ]
     },
     "execution_count": 440,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recall_score(train_data['MULTIPLE_OFFENSE'],predictions_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 441,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  104,   964],\n",
       "       [   83, 22705]])"
      ]
     },
     "execution_count": 441,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(train_data['MULTIPLE_OFFENSE'],predictions_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 442,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds=gb.predict(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 443,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission=pd.DataFrame({'INCIDENT_ID':test_data.INCIDENT_ID,\"MULTIPLE_OFFENSE\":preds})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 444,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission.to_csv(\"submission_second.csv\",index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}