data_preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Preprocessing\n",
    "\n",
    "In this part we will prepare our data to model building. Here we explore what components we should change, but working code, which we could use in our work I will post with another file, where you can find pipeline of work from getting data to model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "#importing main data analysis libraries  \n",
    "import pandas as pd \n",
    "import numpy as np\n",
    "#importing main visualization libraries\n",
    "%matplotlib inline \n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.pylab as pylab\n",
    "import seaborn as sns\n",
    "\n",
    "%matplotlib inline \n",
    "\n",
    "#set visualization settings\n",
    "large = 30; med = 14; small = 12\n",
    "params = {'axes.titlesize': large,\n",
    "          'legend.fontsize': med,\n",
    "          'figure.figsize': (13, 10),\n",
    "          'axes.labelsize': med,\n",
    "          'axes.titlesize': med,\n",
    "          'xtick.labelsize': med,\n",
    "          'ytick.labelsize': med,\n",
    "          'figure.titlesize': large}\n",
    "plt.rcParams.update(params)\n",
    "plt.style.use('seaborn-whitegrid')\n",
    "sns.set_style(\"white\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_csv('C:/Users/user/Projects/Health_Insurance_Sell_Analysis/data/train.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Encoding features\n",
    "Firstly, we should encode our data, because we will calculate by Random Forest, which our features are important, and this algorithm is working only with encode data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [],
   "source": [
    "#But we should rename our columns for easier working with it\n",
    "for column in train.columns:\n",
    "    column_lower = column.lower()\n",
    "    train = train.rename(columns = {column:column_lower})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "#this func encode our features, we could use OneHotEncoder, but we need strict values, which we need\n",
    "def encoder(x):\n",
    "    if x== 'Male' or x == 'No' or x == '< 1 Year':\n",
    "        return int(0)\n",
    "    elif x== 'Male' or x == 'Yes' or x == '1-2 Year':\n",
    "        return int(1)\n",
    "    else:\n",
    "        return int(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "names = ['gender', 'vehicle_age', 'vehicle_damage']\n",
    "features = [train.gender, train.vehicle_age, train.vehicle_damage]\n",
    "for feature in features:\n",
    "    for name in names:\n",
    "        train[name] = feature.apply(encoder)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.policy_sales_channel = train.policy_sales_channel.apply(lambda x: int(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "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>id</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>driving_license</th>\n",
       "      <th>region_code</th>\n",
       "      <th>previously_insured</th>\n",
       "      <th>vehicle_age</th>\n",
       "      <th>vehicle_damage</th>\n",
       "      <th>annual_premium</th>\n",
       "      <th>policy_sales_channel</th>\n",
       "      <th>vintage</th>\n",
       "      <th>response</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>44</td>\n",
       "      <td>1</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>40454.0</td>\n",
       "      <td>26</td>\n",
       "      <td>217</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>76</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>33536.0</td>\n",
       "      <td>26</td>\n",
       "      <td>183</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>47</td>\n",
       "      <td>1</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38294.0</td>\n",
       "      <td>26</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>1</td>\n",
       "      <td>11.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28619.0</td>\n",
       "      <td>152</td>\n",
       "      <td>203</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>1</td>\n",
       "      <td>41.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27496.0</td>\n",
       "      <td>152</td>\n",
       "      <td>39</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  gender  age  driving_license  region_code  previously_insured  \\\n",
       "0   1       1   44                1         28.0                   0   \n",
       "1   2       0   76                1          3.0                   0   \n",
       "2   3       1   47                1         28.0                   0   \n",
       "3   4       0   21                1         11.0                   1   \n",
       "4   5       0   29                1         41.0                   1   \n",
       "\n",
       "   vehicle_age  vehicle_damage  annual_premium  policy_sales_channel  vintage  \\\n",
       "0            1               1         40454.0                    26      217   \n",
       "1            0               0         33536.0                    26      183   \n",
       "2            1               1         38294.0                    26       27   \n",
       "3            0               0         28619.0                   152      203   \n",
       "4            0               0         27496.0                   152       39   \n",
       "\n",
       "   response  \n",
       "0         1  \n",
       "1         0  \n",
       "2         1  \n",
       "3         0  \n",
       "4         0  "
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dropping features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = train.drop(columns= ['id', 'region_code'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['gender', 'age', 'driving_license', 'previously_insured', 'vehicle_age',\n",
       "       'vehicle_damage', 'annual_premium', 'policy_sales_channel', 'vintage',\n",
       "       'response'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Find not important features using Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    train.drop(columns= 'response', axis = 1), \n",
    "    train['response'], \n",
    "    test_size=0.2,\n",
    "    random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
       "                       criterion='gini', max_depth=None, max_features='auto',\n",
       "                       max_leaf_nodes=None, max_samples=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_jobs=None, oob_score=False, random_state=None,\n",
       "                       verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "rfc_ftrs = RandomForestClassifier()\n",
    "\n",
    "rfc_ftrs.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "scrolled": true
   },
   "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>features</th>\n",
       "      <th>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>id</td>\n",
       "      <td>0.211183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>gender</td>\n",
       "      <td>0.012660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>age</td>\n",
       "      <td>0.121872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>driving_license</td>\n",
       "      <td>0.000502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>region_code</td>\n",
       "      <td>0.086872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>previously_insured</td>\n",
       "      <td>0.057700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>vehicle_age</td>\n",
       "      <td>0.018765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>vehicle_damage</td>\n",
       "      <td>0.069356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>annual_premium</td>\n",
       "      <td>0.172413</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>policy_sales_channel</td>\n",
       "      <td>0.056777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>vintage</td>\n",
       "      <td>0.191899</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                features  importance\n",
       "0                     id    0.211183\n",
       "1                 gender    0.012660\n",
       "2                    age    0.121872\n",
       "3        driving_license    0.000502\n",
       "4            region_code    0.086872\n",
       "5     previously_insured    0.057700\n",
       "6            vehicle_age    0.018765\n",
       "7         vehicle_damage    0.069356\n",
       "8         annual_premium    0.172413\n",
       "9   policy_sales_channel    0.056777\n",
       "10               vintage    0.191899"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_ftrs = pd.DataFrame({\n",
    "    'features':X_train.columns,\n",
    "    'importance':rfc_ftrs.feature_importances_\n",
    "})\n",
    "\n",
    "df_ftrs\n",
    "\n",
    "#as we can see gender and driving license doesn't make a big sense in our dataset, so we can drop them\n",
    "#we could drop previously insured, vehicle_age and etc. as well, but it can make our data too small and our models\n",
    "#would be underfitted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = train.drop(columns = ['gender', 'driving_license'], axis = 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Remove outliers\n",
    "Here we will remove outliers to make our models work more efficient. In EDA part we saw, that annual_premium had a lot of outliers, which we should remove."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "##if we look on first boxplot, we can see a lot of outliers, they started after 60000\n",
    "ap_boxplot_bf = plt.subplot()\n",
    "ap_boxplot_bf = sns.boxplot(\n",
    "    x = 'annual_premium',\n",
    "    data = train\n",
    ")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x28482484e88>"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#if we limit it we got normal boxplot, without big outliers, lets check how much do we lost\n",
    "sns.boxplot(\n",
    "    x = train[train.annual_premium <= 60000].annual_premium,\n",
    "    data = train\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(381109, 8) \n",
      " (368950, 8)\n"
     ]
    }
   ],
   "source": [
    "#we didn't lost too much, only 13 000 rows, in my opinion it would improve work of our model, so we can continue with this\n",
    "print(train.shape,'\\n',\n",
    "      train[train.annual_premium <= 60000].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = train[train.annual_premium <= 60000]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Scaling\n",
    "Here we will scale our data with MinMaxScaler, because we need this type of data to improve work with SVC and other algorithms. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "\n",
    "\n",
    "train[['annual_premium', 'vintage', 'age']] = scaler.fit_transform(train[['annual_premium', 'vintage', 'age']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {
    "scrolled": true
   },
   "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>age</th>\n",
       "      <th>previously_insured</th>\n",
       "      <th>vehicle_age</th>\n",
       "      <th>vehicle_damage</th>\n",
       "      <th>annual_premium</th>\n",
       "      <th>policy_sales_channel</th>\n",
       "      <th>vintage</th>\n",
       "      <th>response</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.369231</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.659299</td>\n",
       "      <td>26</td>\n",
       "      <td>0.716263</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.861538</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.538714</td>\n",
       "      <td>26</td>\n",
       "      <td>0.598616</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.415385</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.621649</td>\n",
       "      <td>26</td>\n",
       "      <td>0.058824</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.015385</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.453007</td>\n",
       "      <td>152</td>\n",
       "      <td>0.667820</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.138462</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.433432</td>\n",
       "      <td>152</td>\n",
       "      <td>0.100346</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        age  previously_insured  vehicle_age  vehicle_damage  annual_premium  \\\n",
       "0  0.369231                   0            1               1        0.659299   \n",
       "1  0.861538                   0            0               0        0.538714   \n",
       "2  0.415385                   0            1               1        0.621649   \n",
       "3  0.015385                   1            0               0        0.453007   \n",
       "4  0.138462                   1            0               0        0.433432   \n",
       "\n",
       "   policy_sales_channel   vintage  response  \n",
       "0                    26  0.716263         1  \n",
       "1                    26  0.598616         0  \n",
       "2                    26  0.058824         1  \n",
       "3                   152  0.667820         0  \n",
       "4                   152  0.100346         0  "
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepared data output\n",
    "Here we will output our data to another file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.to_csv('prepared_data', index= False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}