CatBoostInsuranceRegression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f3d04198",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a846b3c3",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>...</th>\n",
       "      <th>cont6</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "      <th>loss</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.718367</td>\n",
       "      <td>0.335060</td>\n",
       "      <td>0.30260</td>\n",
       "      <td>0.67135</td>\n",
       "      <td>0.83510</td>\n",
       "      <td>0.569745</td>\n",
       "      <td>0.594646</td>\n",
       "      <td>0.822493</td>\n",
       "      <td>0.714843</td>\n",
       "      <td>2213.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.438917</td>\n",
       "      <td>0.436585</td>\n",
       "      <td>0.60087</td>\n",
       "      <td>0.35127</td>\n",
       "      <td>0.43919</td>\n",
       "      <td>0.338312</td>\n",
       "      <td>0.366307</td>\n",
       "      <td>0.611431</td>\n",
       "      <td>0.304496</td>\n",
       "      <td>1283.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.289648</td>\n",
       "      <td>0.315545</td>\n",
       "      <td>0.27320</td>\n",
       "      <td>0.26076</td>\n",
       "      <td>0.32446</td>\n",
       "      <td>0.381398</td>\n",
       "      <td>0.373424</td>\n",
       "      <td>0.195709</td>\n",
       "      <td>0.774425</td>\n",
       "      <td>3005.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.440945</td>\n",
       "      <td>0.391128</td>\n",
       "      <td>0.31796</td>\n",
       "      <td>0.32128</td>\n",
       "      <td>0.44467</td>\n",
       "      <td>0.327915</td>\n",
       "      <td>0.321570</td>\n",
       "      <td>0.605077</td>\n",
       "      <td>0.602642</td>\n",
       "      <td>939.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.178193</td>\n",
       "      <td>0.247408</td>\n",
       "      <td>0.24564</td>\n",
       "      <td>0.22089</td>\n",
       "      <td>0.21230</td>\n",
       "      <td>0.204687</td>\n",
       "      <td>0.202213</td>\n",
       "      <td>0.246011</td>\n",
       "      <td>0.432606</td>\n",
       "      <td>2763.85</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 132 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9  ...     cont6     cont7  \\\n",
       "0   1    A    B    A    B    A    A    A    A    B  ...  0.718367  0.335060   \n",
       "1   2    A    B    A    A    A    A    A    A    B  ...  0.438917  0.436585   \n",
       "2   5    A    B    A    A    B    A    A    A    B  ...  0.289648  0.315545   \n",
       "3  10    B    B    A    B    A    A    A    A    B  ...  0.440945  0.391128   \n",
       "4  11    A    B    A    B    A    A    A    A    B  ...  0.178193  0.247408   \n",
       "\n",
       "     cont8    cont9   cont10    cont11    cont12    cont13    cont14     loss  \n",
       "0  0.30260  0.67135  0.83510  0.569745  0.594646  0.822493  0.714843  2213.18  \n",
       "1  0.60087  0.35127  0.43919  0.338312  0.366307  0.611431  0.304496  1283.60  \n",
       "2  0.27320  0.26076  0.32446  0.381398  0.373424  0.195709  0.774425  3005.09  \n",
       "3  0.31796  0.32128  0.44467  0.327915  0.321570  0.605077  0.602642   939.85  \n",
       "4  0.24564  0.22089  0.21230  0.204687  0.202213  0.246011  0.432606  2763.85  \n",
       "\n",
       "[5 rows x 132 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('train.csv')\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "78f315cb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(188318, 132)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "602ed72f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull().sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "42f99097",
   "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>cont1</th>\n",
       "      <th>cont2</th>\n",
       "      <th>cont3</th>\n",
       "      <th>cont4</th>\n",
       "      <th>cont5</th>\n",
       "      <th>cont6</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "      <th>loss</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "      <td>188318.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>294135.982561</td>\n",
       "      <td>0.493861</td>\n",
       "      <td>0.507188</td>\n",
       "      <td>0.498918</td>\n",
       "      <td>0.491812</td>\n",
       "      <td>0.487428</td>\n",
       "      <td>0.490945</td>\n",
       "      <td>0.484970</td>\n",
       "      <td>0.486437</td>\n",
       "      <td>0.485506</td>\n",
       "      <td>0.498066</td>\n",
       "      <td>0.493511</td>\n",
       "      <td>0.493150</td>\n",
       "      <td>0.493138</td>\n",
       "      <td>0.495717</td>\n",
       "      <td>3037.337686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>169336.084867</td>\n",
       "      <td>0.187640</td>\n",
       "      <td>0.207202</td>\n",
       "      <td>0.202105</td>\n",
       "      <td>0.211292</td>\n",
       "      <td>0.209027</td>\n",
       "      <td>0.205273</td>\n",
       "      <td>0.178450</td>\n",
       "      <td>0.199370</td>\n",
       "      <td>0.181660</td>\n",
       "      <td>0.185877</td>\n",
       "      <td>0.209737</td>\n",
       "      <td>0.209427</td>\n",
       "      <td>0.212777</td>\n",
       "      <td>0.222488</td>\n",
       "      <td>2904.086186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>0.001149</td>\n",
       "      <td>0.002634</td>\n",
       "      <td>0.176921</td>\n",
       "      <td>0.281143</td>\n",
       "      <td>0.012683</td>\n",
       "      <td>0.069503</td>\n",
       "      <td>0.236880</td>\n",
       "      <td>0.000080</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.035321</td>\n",
       "      <td>0.036232</td>\n",
       "      <td>0.000228</td>\n",
       "      <td>0.179722</td>\n",
       "      <td>0.670000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>147748.250000</td>\n",
       "      <td>0.346090</td>\n",
       "      <td>0.358319</td>\n",
       "      <td>0.336963</td>\n",
       "      <td>0.327354</td>\n",
       "      <td>0.281143</td>\n",
       "      <td>0.336105</td>\n",
       "      <td>0.350175</td>\n",
       "      <td>0.312800</td>\n",
       "      <td>0.358970</td>\n",
       "      <td>0.364580</td>\n",
       "      <td>0.310961</td>\n",
       "      <td>0.311661</td>\n",
       "      <td>0.315758</td>\n",
       "      <td>0.294610</td>\n",
       "      <td>1204.460000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>294539.500000</td>\n",
       "      <td>0.475784</td>\n",
       "      <td>0.555782</td>\n",
       "      <td>0.527991</td>\n",
       "      <td>0.452887</td>\n",
       "      <td>0.422268</td>\n",
       "      <td>0.440945</td>\n",
       "      <td>0.438285</td>\n",
       "      <td>0.441060</td>\n",
       "      <td>0.441450</td>\n",
       "      <td>0.461190</td>\n",
       "      <td>0.457203</td>\n",
       "      <td>0.462286</td>\n",
       "      <td>0.363547</td>\n",
       "      <td>0.407403</td>\n",
       "      <td>2115.570000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>440680.500000</td>\n",
       "      <td>0.623912</td>\n",
       "      <td>0.681761</td>\n",
       "      <td>0.634224</td>\n",
       "      <td>0.652072</td>\n",
       "      <td>0.643315</td>\n",
       "      <td>0.655021</td>\n",
       "      <td>0.591045</td>\n",
       "      <td>0.623580</td>\n",
       "      <td>0.566820</td>\n",
       "      <td>0.614590</td>\n",
       "      <td>0.678924</td>\n",
       "      <td>0.675759</td>\n",
       "      <td>0.689974</td>\n",
       "      <td>0.724623</td>\n",
       "      <td>3864.045000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>587633.000000</td>\n",
       "      <td>0.984975</td>\n",
       "      <td>0.862654</td>\n",
       "      <td>0.944251</td>\n",
       "      <td>0.954297</td>\n",
       "      <td>0.983674</td>\n",
       "      <td>0.997162</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.980200</td>\n",
       "      <td>0.995400</td>\n",
       "      <td>0.994980</td>\n",
       "      <td>0.998742</td>\n",
       "      <td>0.998484</td>\n",
       "      <td>0.988494</td>\n",
       "      <td>0.844848</td>\n",
       "      <td>121012.250000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  id          cont1          cont2          cont3  \\\n",
       "count  188318.000000  188318.000000  188318.000000  188318.000000   \n",
       "mean   294135.982561       0.493861       0.507188       0.498918   \n",
       "std    169336.084867       0.187640       0.207202       0.202105   \n",
       "min         1.000000       0.000016       0.001149       0.002634   \n",
       "25%    147748.250000       0.346090       0.358319       0.336963   \n",
       "50%    294539.500000       0.475784       0.555782       0.527991   \n",
       "75%    440680.500000       0.623912       0.681761       0.634224   \n",
       "max    587633.000000       0.984975       0.862654       0.944251   \n",
       "\n",
       "               cont4          cont5          cont6          cont7  \\\n",
       "count  188318.000000  188318.000000  188318.000000  188318.000000   \n",
       "mean        0.491812       0.487428       0.490945       0.484970   \n",
       "std         0.211292       0.209027       0.205273       0.178450   \n",
       "min         0.176921       0.281143       0.012683       0.069503   \n",
       "25%         0.327354       0.281143       0.336105       0.350175   \n",
       "50%         0.452887       0.422268       0.440945       0.438285   \n",
       "75%         0.652072       0.643315       0.655021       0.591045   \n",
       "max         0.954297       0.983674       0.997162       1.000000   \n",
       "\n",
       "               cont8          cont9         cont10         cont11  \\\n",
       "count  188318.000000  188318.000000  188318.000000  188318.000000   \n",
       "mean        0.486437       0.485506       0.498066       0.493511   \n",
       "std         0.199370       0.181660       0.185877       0.209737   \n",
       "min         0.236880       0.000080       0.000000       0.035321   \n",
       "25%         0.312800       0.358970       0.364580       0.310961   \n",
       "50%         0.441060       0.441450       0.461190       0.457203   \n",
       "75%         0.623580       0.566820       0.614590       0.678924   \n",
       "max         0.980200       0.995400       0.994980       0.998742   \n",
       "\n",
       "              cont12         cont13         cont14           loss  \n",
       "count  188318.000000  188318.000000  188318.000000  188318.000000  \n",
       "mean        0.493150       0.493138       0.495717    3037.337686  \n",
       "std         0.209427       0.212777       0.222488    2904.086186  \n",
       "min         0.036232       0.000228       0.179722       0.670000  \n",
       "25%         0.311661       0.315758       0.294610    1204.460000  \n",
       "50%         0.462286       0.363547       0.407403    2115.570000  \n",
       "75%         0.675759       0.689974       0.724623    3864.045000  \n",
       "max         0.998484       0.988494       0.844848  121012.250000  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "eb01a4a2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 188318 entries, 0 to 188317\n",
      "Columns: 132 entries, id to loss\n",
      "dtypes: float64(15), int64(1), object(116)\n",
      "memory usage: 189.7+ MB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1e56ef3b",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_train = pd.read_csv('train.csv')\n",
    "data_test = pd.read_csv('test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "061debe3",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>...</th>\n",
       "      <th>cont5</th>\n",
       "      <th>cont6</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.281143</td>\n",
       "      <td>0.466591</td>\n",
       "      <td>0.317681</td>\n",
       "      <td>0.61229</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.38016</td>\n",
       "      <td>0.377724</td>\n",
       "      <td>0.369858</td>\n",
       "      <td>0.704052</td>\n",
       "      <td>0.392562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.836443</td>\n",
       "      <td>0.482425</td>\n",
       "      <td>0.443760</td>\n",
       "      <td>0.71330</td>\n",
       "      <td>0.51890</td>\n",
       "      <td>0.60401</td>\n",
       "      <td>0.689039</td>\n",
       "      <td>0.675759</td>\n",
       "      <td>0.453468</td>\n",
       "      <td>0.208045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.718531</td>\n",
       "      <td>0.212308</td>\n",
       "      <td>0.325779</td>\n",
       "      <td>0.29758</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.30529</td>\n",
       "      <td>0.245410</td>\n",
       "      <td>0.241676</td>\n",
       "      <td>0.258586</td>\n",
       "      <td>0.297232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.397069</td>\n",
       "      <td>0.369930</td>\n",
       "      <td>0.342355</td>\n",
       "      <td>0.40028</td>\n",
       "      <td>0.33237</td>\n",
       "      <td>0.31480</td>\n",
       "      <td>0.348867</td>\n",
       "      <td>0.341872</td>\n",
       "      <td>0.592264</td>\n",
       "      <td>0.555955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.302678</td>\n",
       "      <td>0.398862</td>\n",
       "      <td>0.391833</td>\n",
       "      <td>0.23688</td>\n",
       "      <td>0.43731</td>\n",
       "      <td>0.50556</td>\n",
       "      <td>0.359572</td>\n",
       "      <td>0.352251</td>\n",
       "      <td>0.301535</td>\n",
       "      <td>0.825823</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 131 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9  ...     cont5     cont6  \\\n",
       "0   4    A    B    A    A    A    A    A    A    B  ...  0.281143  0.466591   \n",
       "1   6    A    B    A    B    A    A    A    A    B  ...  0.836443  0.482425   \n",
       "2   9    A    B    A    B    B    A    B    A    B  ...  0.718531  0.212308   \n",
       "3  12    A    A    A    A    B    A    A    A    A  ...  0.397069  0.369930   \n",
       "4  15    B    A    A    A    A    B    A    A    A  ...  0.302678  0.398862   \n",
       "\n",
       "      cont7    cont8    cont9   cont10    cont11    cont12    cont13    cont14  \n",
       "0  0.317681  0.61229  0.34365  0.38016  0.377724  0.369858  0.704052  0.392562  \n",
       "1  0.443760  0.71330  0.51890  0.60401  0.689039  0.675759  0.453468  0.208045  \n",
       "2  0.325779  0.29758  0.34365  0.30529  0.245410  0.241676  0.258586  0.297232  \n",
       "3  0.342355  0.40028  0.33237  0.31480  0.348867  0.341872  0.592264  0.555955  \n",
       "4  0.391833  0.23688  0.43731  0.50556  0.359572  0.352251  0.301535  0.825823  \n",
       "\n",
       "[5 rows x 131 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a7821d0a",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_train['isTrain'] = True\n",
    "data_test['isTrain'] = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "60211794",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>...</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "      <th>loss</th>\n",
       "      <th>isTrain</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.335060</td>\n",
       "      <td>0.30260</td>\n",
       "      <td>0.67135</td>\n",
       "      <td>0.83510</td>\n",
       "      <td>0.569745</td>\n",
       "      <td>0.594646</td>\n",
       "      <td>0.822493</td>\n",
       "      <td>0.714843</td>\n",
       "      <td>2213.18</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.436585</td>\n",
       "      <td>0.60087</td>\n",
       "      <td>0.35127</td>\n",
       "      <td>0.43919</td>\n",
       "      <td>0.338312</td>\n",
       "      <td>0.366307</td>\n",
       "      <td>0.611431</td>\n",
       "      <td>0.304496</td>\n",
       "      <td>1283.60</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.315545</td>\n",
       "      <td>0.27320</td>\n",
       "      <td>0.26076</td>\n",
       "      <td>0.32446</td>\n",
       "      <td>0.381398</td>\n",
       "      <td>0.373424</td>\n",
       "      <td>0.195709</td>\n",
       "      <td>0.774425</td>\n",
       "      <td>3005.09</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.391128</td>\n",
       "      <td>0.31796</td>\n",
       "      <td>0.32128</td>\n",
       "      <td>0.44467</td>\n",
       "      <td>0.327915</td>\n",
       "      <td>0.321570</td>\n",
       "      <td>0.605077</td>\n",
       "      <td>0.602642</td>\n",
       "      <td>939.85</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.247408</td>\n",
       "      <td>0.24564</td>\n",
       "      <td>0.22089</td>\n",
       "      <td>0.21230</td>\n",
       "      <td>0.204687</td>\n",
       "      <td>0.202213</td>\n",
       "      <td>0.246011</td>\n",
       "      <td>0.432606</td>\n",
       "      <td>2763.85</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 133 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9  ...     cont7    cont8  \\\n",
       "0   1    A    B    A    B    A    A    A    A    B  ...  0.335060  0.30260   \n",
       "1   2    A    B    A    A    A    A    A    A    B  ...  0.436585  0.60087   \n",
       "2   5    A    B    A    A    B    A    A    A    B  ...  0.315545  0.27320   \n",
       "3  10    B    B    A    B    A    A    A    A    B  ...  0.391128  0.31796   \n",
       "4  11    A    B    A    B    A    A    A    A    B  ...  0.247408  0.24564   \n",
       "\n",
       "     cont9   cont10    cont11    cont12    cont13    cont14     loss isTrain  \n",
       "0  0.67135  0.83510  0.569745  0.594646  0.822493  0.714843  2213.18    True  \n",
       "1  0.35127  0.43919  0.338312  0.366307  0.611431  0.304496  1283.60    True  \n",
       "2  0.26076  0.32446  0.381398  0.373424  0.195709  0.774425  3005.09    True  \n",
       "3  0.32128  0.44467  0.327915  0.321570  0.605077  0.602642   939.85    True  \n",
       "4  0.22089  0.21230  0.204687  0.202213  0.246011  0.432606  2763.85    True  \n",
       "\n",
       "[5 rows x 133 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traintest = pd.concat([data_train, data_test], axis=0)\n",
    "traintest.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f17367bb",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>...</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "      <th>loss</th>\n",
       "      <th>isTrain</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>125541</th>\n",
       "      <td>587617</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.815941</td>\n",
       "      <td>0.39455</td>\n",
       "      <td>0.48740</td>\n",
       "      <td>0.40666</td>\n",
       "      <td>0.550529</td>\n",
       "      <td>0.538473</td>\n",
       "      <td>0.298734</td>\n",
       "      <td>0.345946</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125542</th>\n",
       "      <td>587621</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.424968</td>\n",
       "      <td>0.47669</td>\n",
       "      <td>0.25753</td>\n",
       "      <td>0.26894</td>\n",
       "      <td>0.324486</td>\n",
       "      <td>0.352251</td>\n",
       "      <td>0.490001</td>\n",
       "      <td>0.290576</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125543</th>\n",
       "      <td>587627</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.511502</td>\n",
       "      <td>0.72299</td>\n",
       "      <td>0.94438</td>\n",
       "      <td>0.83510</td>\n",
       "      <td>0.933174</td>\n",
       "      <td>0.926619</td>\n",
       "      <td>0.848129</td>\n",
       "      <td>0.808125</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125544</th>\n",
       "      <td>587629</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.388545</td>\n",
       "      <td>0.31796</td>\n",
       "      <td>0.32128</td>\n",
       "      <td>0.36974</td>\n",
       "      <td>0.307628</td>\n",
       "      <td>0.301921</td>\n",
       "      <td>0.608259</td>\n",
       "      <td>0.361542</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125545</th>\n",
       "      <td>587634</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.242044</td>\n",
       "      <td>0.25461</td>\n",
       "      <td>0.31399</td>\n",
       "      <td>0.25183</td>\n",
       "      <td>0.245410</td>\n",
       "      <td>0.241676</td>\n",
       "      <td>0.287682</td>\n",
       "      <td>0.220323</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 133 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            id cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9  ...     cont7  \\\n",
       "125541  587617    A    A    A    B    A    A    A    A    A  ...  0.815941   \n",
       "125542  587621    A    A    A    A    B    B    A    B    A  ...  0.424968   \n",
       "125543  587627    B    B    A    A    B    A    A    A    B  ...  0.511502   \n",
       "125544  587629    A    A    A    A    A    B    A    B    A  ...  0.388545   \n",
       "125545  587634    A    B    A    A    A    A    A    A    B  ...  0.242044   \n",
       "\n",
       "          cont8    cont9   cont10    cont11    cont12    cont13    cont14  \\\n",
       "125541  0.39455  0.48740  0.40666  0.550529  0.538473  0.298734  0.345946   \n",
       "125542  0.47669  0.25753  0.26894  0.324486  0.352251  0.490001  0.290576   \n",
       "125543  0.72299  0.94438  0.83510  0.933174  0.926619  0.848129  0.808125   \n",
       "125544  0.31796  0.32128  0.36974  0.307628  0.301921  0.608259  0.361542   \n",
       "125545  0.25461  0.31399  0.25183  0.245410  0.241676  0.287682  0.220323   \n",
       "\n",
       "       loss isTrain  \n",
       "125541  NaN   False  \n",
       "125542  NaN   False  \n",
       "125543  NaN   False  \n",
       "125544  NaN   False  \n",
       "125545  NaN   False  \n",
       "\n",
       "[5 rows x 133 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traintest.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c586de0f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1,\n",
       " 2,\n",
       " 3,\n",
       " 4,\n",
       " 5,\n",
       " 6,\n",
       " 7,\n",
       " 8,\n",
       " 9,\n",
       " 10,\n",
       " 11,\n",
       " 12,\n",
       " 13,\n",
       " 14,\n",
       " 15,\n",
       " 16,\n",
       " 17,\n",
       " 18,\n",
       " 19,\n",
       " 20,\n",
       " 21,\n",
       " 22,\n",
       " 23,\n",
       " 24,\n",
       " 25,\n",
       " 26,\n",
       " 27,\n",
       " 28,\n",
       " 29,\n",
       " 30,\n",
       " 31,\n",
       " 32,\n",
       " 33,\n",
       " 34,\n",
       " 35,\n",
       " 36,\n",
       " 37,\n",
       " 38,\n",
       " 39,\n",
       " 40,\n",
       " 41,\n",
       " 42,\n",
       " 43,\n",
       " 44,\n",
       " 45,\n",
       " 46,\n",
       " 47,\n",
       " 48,\n",
       " 49,\n",
       " 50,\n",
       " 51,\n",
       " 52,\n",
       " 53,\n",
       " 54,\n",
       " 55,\n",
       " 56,\n",
       " 57,\n",
       " 58,\n",
       " 59,\n",
       " 60,\n",
       " 61,\n",
       " 62,\n",
       " 63,\n",
       " 64,\n",
       " 65,\n",
       " 66,\n",
       " 67,\n",
       " 68,\n",
       " 69,\n",
       " 70,\n",
       " 71,\n",
       " 72,\n",
       " 73,\n",
       " 74,\n",
       " 75,\n",
       " 76,\n",
       " 77,\n",
       " 78,\n",
       " 79,\n",
       " 80,\n",
       " 81,\n",
       " 82,\n",
       " 83,\n",
       " 84,\n",
       " 85,\n",
       " 86,\n",
       " 87,\n",
       " 88,\n",
       " 89,\n",
       " 90,\n",
       " 91,\n",
       " 92,\n",
       " 93,\n",
       " 94,\n",
       " 95,\n",
       " 96,\n",
       " 97,\n",
       " 98,\n",
       " 99,\n",
       " 100,\n",
       " 101,\n",
       " 102,\n",
       " 103,\n",
       " 104,\n",
       " 105,\n",
       " 106,\n",
       " 107,\n",
       " 108,\n",
       " 109,\n",
       " 110,\n",
       " 111,\n",
       " 112,\n",
       " 113,\n",
       " 114,\n",
       " 115,\n",
       " 116]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import re\n",
    "cat_pattern = re.compile(\"^cat([1-9]|[1-9][0-9]|[1-9][0-9][0-9])$\")\n",
    "cat_index = [i for i in range(0, len(data_train.columns)) if cat_pattern.match(data_train.columns[i])]\n",
    "cat_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fa6f744a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cont_pattern = re.compile(\"^cont([1-9]|[1-9][0-9]|[1-9][0-9][0-9])$\")\n",
    "cont_index = [i for i in range(0, len(data_train.columns)) if cont_pattern.match(data_train.columns[i])]\n",
    "cont_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "af382a6c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['cont1',\n",
       " 'cont2',\n",
       " 'cont3',\n",
       " 'cont4',\n",
       " 'cont5',\n",
       " 'cont6',\n",
       " 'cont7',\n",
       " 'cont8',\n",
       " 'cont9',\n",
       " 'cont10',\n",
       " 'cont11',\n",
       " 'cont12',\n",
       " 'cont13',\n",
       " 'cont14']"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cont_col = sorted([cont for cont in data_train.columns if 'cont' in cont], key=lambda s: int(s[4:]))\n",
    "cont_col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b24182e5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['cat1',\n",
       " 'cat2',\n",
       " 'cat3',\n",
       " 'cat4',\n",
       " 'cat5',\n",
       " 'cat6',\n",
       " 'cat7',\n",
       " 'cat8',\n",
       " 'cat9',\n",
       " 'cat10',\n",
       " 'cat11',\n",
       " 'cat12',\n",
       " 'cat13',\n",
       " 'cat14',\n",
       " 'cat15',\n",
       " 'cat16',\n",
       " 'cat17',\n",
       " 'cat18',\n",
       " 'cat19',\n",
       " 'cat20',\n",
       " 'cat21',\n",
       " 'cat22',\n",
       " 'cat23',\n",
       " 'cat24',\n",
       " 'cat25',\n",
       " 'cat26',\n",
       " 'cat27',\n",
       " 'cat28',\n",
       " 'cat29',\n",
       " 'cat30',\n",
       " 'cat31',\n",
       " 'cat32',\n",
       " 'cat33',\n",
       " 'cat34',\n",
       " 'cat35',\n",
       " 'cat36',\n",
       " 'cat37',\n",
       " 'cat38',\n",
       " 'cat39',\n",
       " 'cat40',\n",
       " 'cat41',\n",
       " 'cat42',\n",
       " 'cat43',\n",
       " 'cat44',\n",
       " 'cat45',\n",
       " 'cat46',\n",
       " 'cat47',\n",
       " 'cat48',\n",
       " 'cat49',\n",
       " 'cat50',\n",
       " 'cat51',\n",
       " 'cat52',\n",
       " 'cat53',\n",
       " 'cat54',\n",
       " 'cat55',\n",
       " 'cat56',\n",
       " 'cat57',\n",
       " 'cat58',\n",
       " 'cat59',\n",
       " 'cat60',\n",
       " 'cat61',\n",
       " 'cat62',\n",
       " 'cat63',\n",
       " 'cat64',\n",
       " 'cat65',\n",
       " 'cat66',\n",
       " 'cat67',\n",
       " 'cat68',\n",
       " 'cat69',\n",
       " 'cat70',\n",
       " 'cat71',\n",
       " 'cat72',\n",
       " 'cat73',\n",
       " 'cat74',\n",
       " 'cat75',\n",
       " 'cat76',\n",
       " 'cat77',\n",
       " 'cat78',\n",
       " 'cat79',\n",
       " 'cat80',\n",
       " 'cat81',\n",
       " 'cat82',\n",
       " 'cat83',\n",
       " 'cat84',\n",
       " 'cat85',\n",
       " 'cat86',\n",
       " 'cat87',\n",
       " 'cat88',\n",
       " 'cat89',\n",
       " 'cat90',\n",
       " 'cat91',\n",
       " 'cat92',\n",
       " 'cat93',\n",
       " 'cat94',\n",
       " 'cat95',\n",
       " 'cat96',\n",
       " 'cat97',\n",
       " 'cat98',\n",
       " 'cat99',\n",
       " 'cat100',\n",
       " 'cat101',\n",
       " 'cat102',\n",
       " 'cat103',\n",
       " 'cat104',\n",
       " 'cat105',\n",
       " 'cat106',\n",
       " 'cat107',\n",
       " 'cat108',\n",
       " 'cat109',\n",
       " 'cat110',\n",
       " 'cat111',\n",
       " 'cat112',\n",
       " 'cat113',\n",
       " 'cat114',\n",
       " 'cat115',\n",
       " 'cat116']"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_col = sorted([cat for cat in data_train.columns if 'cat' in cat], key=lambda s: int(s[3:]))\n",
    "cat_col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "79458f54",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "8f7766ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mohamad/anaconda3/lib/python3.9/site-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='loss', ylabel='Density'>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 12))\n",
    "sns.distplot(data_train['loss'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "7ddd0e85",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mohamad/anaconda3/lib/python3.9/site-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='loss', ylabel='Density'>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 12))\n",
    "sns.distplot(np.log(data_train['loss']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "9c1b0c0c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from catboost import CatBoostRegressor\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "03295e2e",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = CatBoostRegressor(iterations=200, learning_rate=0.05, depth=6, eval_metric='MAE', verbose=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "833d860a",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data_train.drop(['id', 'loss', 'isTrain'], axis=1)\n",
    "y = np.log(data_train['loss'])\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "43a5a0e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(188318, 130)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.shape(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "6f700c87",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>cat10</th>\n",
       "      <th>...</th>\n",
       "      <th>cont5</th>\n",
       "      <th>cont6</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.310061</td>\n",
       "      <td>0.718367</td>\n",
       "      <td>0.335060</td>\n",
       "      <td>0.30260</td>\n",
       "      <td>0.67135</td>\n",
       "      <td>0.83510</td>\n",
       "      <td>0.569745</td>\n",
       "      <td>0.594646</td>\n",
       "      <td>0.822493</td>\n",
       "      <td>0.714843</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.885834</td>\n",
       "      <td>0.438917</td>\n",
       "      <td>0.436585</td>\n",
       "      <td>0.60087</td>\n",
       "      <td>0.35127</td>\n",
       "      <td>0.43919</td>\n",
       "      <td>0.338312</td>\n",
       "      <td>0.366307</td>\n",
       "      <td>0.611431</td>\n",
       "      <td>0.304496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.397069</td>\n",
       "      <td>0.289648</td>\n",
       "      <td>0.315545</td>\n",
       "      <td>0.27320</td>\n",
       "      <td>0.26076</td>\n",
       "      <td>0.32446</td>\n",
       "      <td>0.381398</td>\n",
       "      <td>0.373424</td>\n",
       "      <td>0.195709</td>\n",
       "      <td>0.774425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.422268</td>\n",
       "      <td>0.440945</td>\n",
       "      <td>0.391128</td>\n",
       "      <td>0.31796</td>\n",
       "      <td>0.32128</td>\n",
       "      <td>0.44467</td>\n",
       "      <td>0.327915</td>\n",
       "      <td>0.321570</td>\n",
       "      <td>0.605077</td>\n",
       "      <td>0.602642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.704268</td>\n",
       "      <td>0.178193</td>\n",
       "      <td>0.247408</td>\n",
       "      <td>0.24564</td>\n",
       "      <td>0.22089</td>\n",
       "      <td>0.21230</td>\n",
       "      <td>0.204687</td>\n",
       "      <td>0.202213</td>\n",
       "      <td>0.246011</td>\n",
       "      <td>0.432606</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 130 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9 cat10  ...     cont5     cont6  \\\n",
       "0    A    B    A    B    A    A    A    A    B     A  ...  0.310061  0.718367   \n",
       "1    A    B    A    A    A    A    A    A    B     B  ...  0.885834  0.438917   \n",
       "2    A    B    A    A    B    A    A    A    B     B  ...  0.397069  0.289648   \n",
       "3    B    B    A    B    A    A    A    A    B     A  ...  0.422268  0.440945   \n",
       "4    A    B    A    B    A    A    A    A    B     B  ...  0.704268  0.178193   \n",
       "\n",
       "      cont7    cont8    cont9   cont10    cont11    cont12    cont13    cont14  \n",
       "0  0.335060  0.30260  0.67135  0.83510  0.569745  0.594646  0.822493  0.714843  \n",
       "1  0.436585  0.60087  0.35127  0.43919  0.338312  0.366307  0.611431  0.304496  \n",
       "2  0.315545  0.27320  0.26076  0.32446  0.381398  0.373424  0.195709  0.774425  \n",
       "3  0.391128  0.31796  0.32128  0.44467  0.327915  0.321570  0.605077  0.602642  \n",
       "4  0.247408  0.24564  0.22089  0.21230  0.204687  0.202213  0.246011  0.432606  \n",
       "\n",
       "[5 rows x 130 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "1783e59a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  0,   1,   2,   3,   4,   5,   6,   7,   8,   9,  10,  11,  12,\n",
       "        13,  14,  15,  16,  17,  18,  19,  20,  21,  22,  23,  24,  25,\n",
       "        26,  27,  28,  29,  30,  31,  32,  33,  34,  35,  36,  37,  38,\n",
       "        39,  40,  41,  42,  43,  44,  45,  46,  47,  48,  49,  50,  51,\n",
       "        52,  53,  54,  55,  56,  57,  58,  59,  60,  61,  62,  63,  64,\n",
       "        65,  66,  67,  68,  69,  70,  71,  72,  73,  74,  75,  76,  77,\n",
       "        78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,  89,  90,\n",
       "        91,  92,  93,  94,  95,  96,  97,  98,  99, 100, 101, 102, 103,\n",
       "       104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.asarray(cat_index) - 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "4536a70d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 0.6480384\ttest: 0.6450584\tbest: 0.6450584 (0)\ttotal: 451ms\tremaining: 1m 29s\n",
      "10:\tlearn: 0.5600235\ttest: 0.5572567\tbest: 0.5572567 (10)\ttotal: 3.85s\tremaining: 1m 6s\n",
      "20:\tlearn: 0.5157233\ttest: 0.5135432\tbest: 0.5135432 (20)\ttotal: 7.12s\tremaining: 1m\n",
      "30:\tlearn: 0.4893528\ttest: 0.4874572\tbest: 0.4874572 (30)\ttotal: 10.2s\tremaining: 55.4s\n",
      "40:\tlearn: 0.4737406\ttest: 0.4720454\tbest: 0.4720454 (40)\ttotal: 13.4s\tremaining: 51.9s\n",
      "50:\tlearn: 0.4629948\ttest: 0.4615391\tbest: 0.4615391 (50)\ttotal: 16.5s\tremaining: 48.1s\n",
      "60:\tlearn: 0.4550304\ttest: 0.4537317\tbest: 0.4537317 (60)\ttotal: 19.7s\tremaining: 44.9s\n",
      "70:\tlearn: 0.4493657\ttest: 0.4482263\tbest: 0.4482263 (70)\ttotal: 22.9s\tremaining: 41.6s\n",
      "80:\tlearn: 0.4448871\ttest: 0.4437611\tbest: 0.4437611 (80)\ttotal: 26.1s\tremaining: 38.4s\n",
      "90:\tlearn: 0.4415576\ttest: 0.4404169\tbest: 0.4404169 (90)\ttotal: 29.3s\tremaining: 35.1s\n",
      "100:\tlearn: 0.4389563\ttest: 0.4378814\tbest: 0.4378814 (100)\ttotal: 32.5s\tremaining: 31.9s\n",
      "110:\tlearn: 0.4368814\ttest: 0.4359520\tbest: 0.4359520 (110)\ttotal: 35.8s\tremaining: 28.7s\n",
      "120:\tlearn: 0.4351447\ttest: 0.4344046\tbest: 0.4344046 (120)\ttotal: 39.1s\tremaining: 25.5s\n",
      "130:\tlearn: 0.4336076\ttest: 0.4329783\tbest: 0.4329783 (130)\ttotal: 42.3s\tremaining: 22.3s\n",
      "140:\tlearn: 0.4325275\ttest: 0.4320265\tbest: 0.4320265 (140)\ttotal: 45.5s\tremaining: 19s\n",
      "150:\tlearn: 0.4314289\ttest: 0.4310912\tbest: 0.4310912 (150)\ttotal: 48.8s\tremaining: 15.8s\n",
      "160:\tlearn: 0.4304728\ttest: 0.4302612\tbest: 0.4302612 (160)\ttotal: 52s\tremaining: 12.6s\n",
      "170:\tlearn: 0.4295278\ttest: 0.4294755\tbest: 0.4294755 (170)\ttotal: 55.3s\tremaining: 9.38s\n",
      "180:\tlearn: 0.4286654\ttest: 0.4287129\tbest: 0.4287129 (180)\ttotal: 58.5s\tremaining: 6.14s\n",
      "190:\tlearn: 0.4278236\ttest: 0.4279770\tbest: 0.4279770 (190)\ttotal: 1m 1s\tremaining: 2.91s\n",
      "199:\tlearn: 0.4271664\ttest: 0.4274354\tbest: 0.4274354 (199)\ttotal: 1m 4s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.4274353708\n",
      "bestIteration = 199\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<catboost.core.CatBoostRegressor at 0x7f33645da190>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train, np.asarray(cat_index) - 1, eval_set=(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "ebc20eea",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test = pd.read_csv('test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "80afab09",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>...</th>\n",
       "      <th>cont5</th>\n",
       "      <th>cont6</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.281143</td>\n",
       "      <td>0.466591</td>\n",
       "      <td>0.317681</td>\n",
       "      <td>0.61229</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.38016</td>\n",
       "      <td>0.377724</td>\n",
       "      <td>0.369858</td>\n",
       "      <td>0.704052</td>\n",
       "      <td>0.392562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.836443</td>\n",
       "      <td>0.482425</td>\n",
       "      <td>0.443760</td>\n",
       "      <td>0.71330</td>\n",
       "      <td>0.51890</td>\n",
       "      <td>0.60401</td>\n",
       "      <td>0.689039</td>\n",
       "      <td>0.675759</td>\n",
       "      <td>0.453468</td>\n",
       "      <td>0.208045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.718531</td>\n",
       "      <td>0.212308</td>\n",
       "      <td>0.325779</td>\n",
       "      <td>0.29758</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.30529</td>\n",
       "      <td>0.245410</td>\n",
       "      <td>0.241676</td>\n",
       "      <td>0.258586</td>\n",
       "      <td>0.297232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.397069</td>\n",
       "      <td>0.369930</td>\n",
       "      <td>0.342355</td>\n",
       "      <td>0.40028</td>\n",
       "      <td>0.33237</td>\n",
       "      <td>0.31480</td>\n",
       "      <td>0.348867</td>\n",
       "      <td>0.341872</td>\n",
       "      <td>0.592264</td>\n",
       "      <td>0.555955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.302678</td>\n",
       "      <td>0.398862</td>\n",
       "      <td>0.391833</td>\n",
       "      <td>0.23688</td>\n",
       "      <td>0.43731</td>\n",
       "      <td>0.50556</td>\n",
       "      <td>0.359572</td>\n",
       "      <td>0.352251</td>\n",
       "      <td>0.301535</td>\n",
       "      <td>0.825823</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 131 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   id cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9  ...     cont5     cont6  \\\n",
       "0   4    A    B    A    A    A    A    A    A    B  ...  0.281143  0.466591   \n",
       "1   6    A    B    A    B    A    A    A    A    B  ...  0.836443  0.482425   \n",
       "2   9    A    B    A    B    B    A    B    A    B  ...  0.718531  0.212308   \n",
       "3  12    A    A    A    A    B    A    A    A    A  ...  0.397069  0.369930   \n",
       "4  15    B    A    A    A    A    B    A    A    A  ...  0.302678  0.398862   \n",
       "\n",
       "      cont7    cont8    cont9   cont10    cont11    cont12    cont13    cont14  \n",
       "0  0.317681  0.61229  0.34365  0.38016  0.377724  0.369858  0.704052  0.392562  \n",
       "1  0.443760  0.71330  0.51890  0.60401  0.689039  0.675759  0.453468  0.208045  \n",
       "2  0.325779  0.29758  0.34365  0.30529  0.245410  0.241676  0.258586  0.297232  \n",
       "3  0.342355  0.40028  0.33237  0.31480  0.348867  0.341872  0.592264  0.555955  \n",
       "4  0.391833  0.23688  0.43731  0.50556  0.359572  0.352251  0.301535  0.825823  \n",
       "\n",
       "[5 rows x 131 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "c24afd50",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test.drop(['id'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "f9a2bb62",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>cat10</th>\n",
       "      <th>...</th>\n",
       "      <th>cont5</th>\n",
       "      <th>cont6</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.281143</td>\n",
       "      <td>0.466591</td>\n",
       "      <td>0.317681</td>\n",
       "      <td>0.61229</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.38016</td>\n",
       "      <td>0.377724</td>\n",
       "      <td>0.369858</td>\n",
       "      <td>0.704052</td>\n",
       "      <td>0.392562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.836443</td>\n",
       "      <td>0.482425</td>\n",
       "      <td>0.443760</td>\n",
       "      <td>0.71330</td>\n",
       "      <td>0.51890</td>\n",
       "      <td>0.60401</td>\n",
       "      <td>0.689039</td>\n",
       "      <td>0.675759</td>\n",
       "      <td>0.453468</td>\n",
       "      <td>0.208045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.718531</td>\n",
       "      <td>0.212308</td>\n",
       "      <td>0.325779</td>\n",
       "      <td>0.29758</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.30529</td>\n",
       "      <td>0.245410</td>\n",
       "      <td>0.241676</td>\n",
       "      <td>0.258586</td>\n",
       "      <td>0.297232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.397069</td>\n",
       "      <td>0.369930</td>\n",
       "      <td>0.342355</td>\n",
       "      <td>0.40028</td>\n",
       "      <td>0.33237</td>\n",
       "      <td>0.31480</td>\n",
       "      <td>0.348867</td>\n",
       "      <td>0.341872</td>\n",
       "      <td>0.592264</td>\n",
       "      <td>0.555955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.302678</td>\n",
       "      <td>0.398862</td>\n",
       "      <td>0.391833</td>\n",
       "      <td>0.23688</td>\n",
       "      <td>0.43731</td>\n",
       "      <td>0.50556</td>\n",
       "      <td>0.359572</td>\n",
       "      <td>0.352251</td>\n",
       "      <td>0.301535</td>\n",
       "      <td>0.825823</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 130 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9 cat10  ...     cont5     cont6  \\\n",
       "0    A    B    A    A    A    A    A    A    B     A  ...  0.281143  0.466591   \n",
       "1    A    B    A    B    A    A    A    A    B     A  ...  0.836443  0.482425   \n",
       "2    A    B    A    B    B    A    B    A    B     B  ...  0.718531  0.212308   \n",
       "3    A    A    A    A    B    A    A    A    A     A  ...  0.397069  0.369930   \n",
       "4    B    A    A    A    A    B    A    A    A     A  ...  0.302678  0.398862   \n",
       "\n",
       "      cont7    cont8    cont9   cont10    cont11    cont12    cont13    cont14  \n",
       "0  0.317681  0.61229  0.34365  0.38016  0.377724  0.369858  0.704052  0.392562  \n",
       "1  0.443760  0.71330  0.51890  0.60401  0.689039  0.675759  0.453468  0.208045  \n",
       "2  0.325779  0.29758  0.34365  0.30529  0.245410  0.241676  0.258586  0.297232  \n",
       "3  0.342355  0.40028  0.33237  0.31480  0.348867  0.341872  0.592264  0.555955  \n",
       "4  0.391833  0.23688  0.43731  0.50556  0.359572  0.352251  0.301535  0.825823  \n",
       "\n",
       "[5 rows x 130 columns]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "1b7619d2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7.40337149, 7.58380639, 9.1215944 , ..., 7.82274527, 6.80423938,\n",
       "       8.26050485])"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(data_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "29435d49",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1641.50944461, 1966.0984825 , 9150.77998536, ..., 2496.75026551,\n",
       "        901.66169111, 3868.04638745])"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.exp(model.predict(data_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "0a3e8554",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "125546"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(np.exp(model.predict(data_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "7b629cc7",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test['loss'] = np.exp(model.predict(data_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "e33e2157",
   "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>cat1</th>\n",
       "      <th>cat2</th>\n",
       "      <th>cat3</th>\n",
       "      <th>cat4</th>\n",
       "      <th>cat5</th>\n",
       "      <th>cat6</th>\n",
       "      <th>cat7</th>\n",
       "      <th>cat8</th>\n",
       "      <th>cat9</th>\n",
       "      <th>cat10</th>\n",
       "      <th>...</th>\n",
       "      <th>cont6</th>\n",
       "      <th>cont7</th>\n",
       "      <th>cont8</th>\n",
       "      <th>cont9</th>\n",
       "      <th>cont10</th>\n",
       "      <th>cont11</th>\n",
       "      <th>cont12</th>\n",
       "      <th>cont13</th>\n",
       "      <th>cont14</th>\n",
       "      <th>loss</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.466591</td>\n",
       "      <td>0.317681</td>\n",
       "      <td>0.61229</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.38016</td>\n",
       "      <td>0.377724</td>\n",
       "      <td>0.369858</td>\n",
       "      <td>0.704052</td>\n",
       "      <td>0.392562</td>\n",
       "      <td>1641.509445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.482425</td>\n",
       "      <td>0.443760</td>\n",
       "      <td>0.71330</td>\n",
       "      <td>0.51890</td>\n",
       "      <td>0.60401</td>\n",
       "      <td>0.689039</td>\n",
       "      <td>0.675759</td>\n",
       "      <td>0.453468</td>\n",
       "      <td>0.208045</td>\n",
       "      <td>1966.098483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>B</td>\n",
       "      <td>...</td>\n",
       "      <td>0.212308</td>\n",
       "      <td>0.325779</td>\n",
       "      <td>0.29758</td>\n",
       "      <td>0.34365</td>\n",
       "      <td>0.30529</td>\n",
       "      <td>0.245410</td>\n",
       "      <td>0.241676</td>\n",
       "      <td>0.258586</td>\n",
       "      <td>0.297232</td>\n",
       "      <td>9150.779985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.369930</td>\n",
       "      <td>0.342355</td>\n",
       "      <td>0.40028</td>\n",
       "      <td>0.33237</td>\n",
       "      <td>0.31480</td>\n",
       "      <td>0.348867</td>\n",
       "      <td>0.341872</td>\n",
       "      <td>0.592264</td>\n",
       "      <td>0.555955</td>\n",
       "      <td>5170.467404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>A</td>\n",
       "      <td>...</td>\n",
       "      <td>0.398862</td>\n",
       "      <td>0.391833</td>\n",
       "      <td>0.23688</td>\n",
       "      <td>0.43731</td>\n",
       "      <td>0.50556</td>\n",
       "      <td>0.359572</td>\n",
       "      <td>0.352251</td>\n",
       "      <td>0.301535</td>\n",
       "      <td>0.825823</td>\n",
       "      <td>887.316997</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 131 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  cat1 cat2 cat3 cat4 cat5 cat6 cat7 cat8 cat9 cat10  ...     cont6     cont7  \\\n",
       "0    A    B    A    A    A    A    A    A    B     A  ...  0.466591  0.317681   \n",
       "1    A    B    A    B    A    A    A    A    B     A  ...  0.482425  0.443760   \n",
       "2    A    B    A    B    B    A    B    A    B     B  ...  0.212308  0.325779   \n",
       "3    A    A    A    A    B    A    A    A    A     A  ...  0.369930  0.342355   \n",
       "4    B    A    A    A    A    B    A    A    A     A  ...  0.398862  0.391833   \n",
       "\n",
       "     cont8    cont9   cont10    cont11    cont12    cont13    cont14  \\\n",
       "0  0.61229  0.34365  0.38016  0.377724  0.369858  0.704052  0.392562   \n",
       "1  0.71330  0.51890  0.60401  0.689039  0.675759  0.453468  0.208045   \n",
       "2  0.29758  0.34365  0.30529  0.245410  0.241676  0.258586  0.297232   \n",
       "3  0.40028  0.33237  0.31480  0.348867  0.341872  0.592264  0.555955   \n",
       "4  0.23688  0.43731  0.50556  0.359572  0.352251  0.301535  0.825823   \n",
       "\n",
       "          loss  \n",
       "0  1641.509445  \n",
       "1  1966.098483  \n",
       "2  9150.779985  \n",
       "3  5170.467404  \n",
       "4   887.316997  \n",
       "\n",
       "[5 rows x 131 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "28ed260a",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test.to_excel('Prediction.xlsx')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}