part1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Standard toolset\n",
    "%matplotlib inline\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plot\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt \n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2310, 11)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df = pd.read_csv('flats.csv')\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>balcony</th>\n",
       "      <th>total_area</th>\n",
       "      <th>year</th>\n",
       "      <th>cost</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1859.000000</td>\n",
       "      <td>2310.000000</td>\n",
       "      <td>2081.000000</td>\n",
       "      <td>2.310000e+03</td>\n",
       "      <td>2310.000000</td>\n",
       "      <td>2310.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.010758</td>\n",
       "      <td>36.169697</td>\n",
       "      <td>1997.595387</td>\n",
       "      <td>2.848634e+06</td>\n",
       "      <td>60.595955</td>\n",
       "      <td>56.823525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.117804</td>\n",
       "      <td>7.792332</td>\n",
       "      <td>20.895129</td>\n",
       "      <td>9.269107e+05</td>\n",
       "      <td>0.060531</td>\n",
       "      <td>0.043765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>1925.000000</td>\n",
       "      <td>7.800000e+05</td>\n",
       "      <td>60.407310</td>\n",
       "      <td>56.739841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1977.000000</td>\n",
       "      <td>2.290000e+06</td>\n",
       "      <td>60.550942</td>\n",
       "      <td>56.790414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>2009.000000</td>\n",
       "      <td>2.680000e+06</td>\n",
       "      <td>60.601184</td>\n",
       "      <td>56.815863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>2015.000000</td>\n",
       "      <td>3.150000e+06</td>\n",
       "      <td>60.633243</td>\n",
       "      <td>56.856063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>106.000000</td>\n",
       "      <td>2019.000000</td>\n",
       "      <td>1.390000e+07</td>\n",
       "      <td>60.782653</td>\n",
       "      <td>56.958709</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           balcony   total_area         year          cost          lon  \\\n",
       "count  1859.000000  2310.000000  2081.000000  2.310000e+03  2310.000000   \n",
       "mean      1.010758    36.169697  1997.595387  2.848634e+06    60.595955   \n",
       "std       0.117804     7.792332    20.895129  9.269107e+05     0.060531   \n",
       "min       1.000000    20.000000  1925.000000  7.800000e+05    60.407310   \n",
       "25%       1.000000    31.000000  1977.000000  2.290000e+06    60.550942   \n",
       "50%       1.000000    35.000000  2009.000000  2.680000e+06    60.601184   \n",
       "75%       1.000000    40.000000  2015.000000  3.150000e+06    60.633243   \n",
       "max       3.000000   106.000000  2019.000000  1.390000e+07    60.782653   \n",
       "\n",
       "               lat  \n",
       "count  2310.000000  \n",
       "mean     56.823525  \n",
       "std       0.043765  \n",
       "min      56.739841  \n",
       "25%      56.790414  \n",
       "50%      56.815863  \n",
       "75%      56.856063  \n",
       "max      56.958709  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df.total_area >= 20]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "numerical_fields = ['total_area','cost']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for col in numerical_fields:\n",
    "    mask = ~np.isnan(df[col])\n",
    "    sns.distplot(df[col][mask],  color=\"r\",label=col)\n",
    "    plot.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x7f0d5a5b63c8>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.pairplot(df[numerical_fields])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[abs(df.total_area - df.total_area.mean()) <= (3 * df.total_area.std())]\n",
    "df = df[abs(df.cost - df.cost.mean()) <= (3 * df.cost.std())]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x7f0d5a5b6be0>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.pairplot(df[numerical_fields])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>balcony</th>\n",
       "      <th>walls</th>\n",
       "      <th>total_area</th>\n",
       "      <th>district</th>\n",
       "      <th>year</th>\n",
       "      <th>cost</th>\n",
       "      <th>description</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>type_house</th>\n",
       "      <th>repair</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>Панельный</td>\n",
       "      <td>30</td>\n",
       "      <td>Уралмаш</td>\n",
       "      <td>1963.0</td>\n",
       "      <td>2250000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>60.600987</td>\n",
       "      <td>56.890172</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Хорошее</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Кирпичный</td>\n",
       "      <td>32</td>\n",
       "      <td>Эльмаш</td>\n",
       "      <td>1959.0</td>\n",
       "      <td>2090000</td>\n",
       "      <td>Продается отличная,теплая, уютная однокомнатна...</td>\n",
       "      <td>60.615153</td>\n",
       "      <td>56.884861</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Хорошее</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>42</td>\n",
       "      <td>Ботанический</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>2800000</td>\n",
       "      <td>Квартира с хорошим ремонтом.</td>\n",
       "      <td>60.615800</td>\n",
       "      <td>56.790526</td>\n",
       "      <td>NaN</td>\n",
       "      <td>В</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29</td>\n",
       "      <td>Ю-З</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2200000</td>\n",
       "      <td>Продается светлая,теплая квартира в районе Бур...</td>\n",
       "      <td>60.572762</td>\n",
       "      <td>56.823636</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>Панельный</td>\n",
       "      <td>34</td>\n",
       "      <td>Компрессорный</td>\n",
       "      <td>2001.0</td>\n",
       "      <td>2000000</td>\n",
       "      <td>Уютная, теплая, светлая однокомнатная квартира...</td>\n",
       "      <td>60.763187</td>\n",
       "      <td>56.796718</td>\n",
       "      <td>Улучшенной</td>\n",
       "      <td>Хорошее</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   balcony      walls  total_area       district    year     cost  \\\n",
       "0      1.0  Панельный          30        Уралмаш  1963.0  2250000   \n",
       "1      NaN  Кирпичный          32         Эльмаш  1959.0  2090000   \n",
       "2      NaN        NaN          42   Ботанический  2017.0  2800000   \n",
       "3      1.0        NaN          29            Ю-З     NaN  2200000   \n",
       "4      1.0  Панельный          34  Компрессорный  2001.0  2000000   \n",
       "\n",
       "                                         description        lon        lat  \\\n",
       "0                                                NaN  60.600987  56.890172   \n",
       "1  Продается отличная,теплая, уютная однокомнатна...  60.615153  56.884861   \n",
       "2                       Квартира с хорошим ремонтом.  60.615800  56.790526   \n",
       "3  Продается светлая,теплая квартира в районе Бур...  60.572762  56.823636   \n",
       "4  Уютная, теплая, светлая однокомнатная квартира...  60.763187  56.796718   \n",
       "\n",
       "   type_house   repair  \n",
       "0         NaN  Хорошее  \n",
       "1         NaN  Хорошее  \n",
       "2         NaN        В  \n",
       "3         NaN      NaN  \n",
       "4  Улучшенной  Хорошее  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['age'] = 2019 -df['year']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "flats_on_map = df[[\"lon\",\"lat\",\"cost\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "flats_on_map.to_csv(\"on_map.csv\",index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(columns=[\"lon\",\"lat\",\"description\"],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>balcony</th>\n",
       "      <th>walls</th>\n",
       "      <th>total_area</th>\n",
       "      <th>district</th>\n",
       "      <th>year</th>\n",
       "      <th>cost</th>\n",
       "      <th>type_house</th>\n",
       "      <th>repair</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>Панельный</td>\n",
       "      <td>30</td>\n",
       "      <td>Уралмаш</td>\n",
       "      <td>1963.0</td>\n",
       "      <td>2250000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Хорошее</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Кирпичный</td>\n",
       "      <td>32</td>\n",
       "      <td>Эльмаш</td>\n",
       "      <td>1959.0</td>\n",
       "      <td>2090000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Хорошее</td>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>42</td>\n",
       "      <td>Ботанический</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>2800000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>В</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29</td>\n",
       "      <td>Ю-З</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2200000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>Панельный</td>\n",
       "      <td>34</td>\n",
       "      <td>Компрессорный</td>\n",
       "      <td>2001.0</td>\n",
       "      <td>2000000</td>\n",
       "      <td>Улучшенной</td>\n",
       "      <td>Хорошее</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   balcony      walls  total_area       district    year     cost  type_house  \\\n",
       "0      1.0  Панельный          30        Уралмаш  1963.0  2250000         NaN   \n",
       "1      NaN  Кирпичный          32         Эльмаш  1959.0  2090000         NaN   \n",
       "2      NaN        NaN          42   Ботанический  2017.0  2800000         NaN   \n",
       "3      1.0        NaN          29            Ю-З     NaN  2200000         NaN   \n",
       "4      1.0  Панельный          34  Компрессорный  2001.0  2000000  Улучшенной   \n",
       "\n",
       "    repair   age  \n",
       "0  Хорошее  56.0  \n",
       "1  Хорошее  60.0  \n",
       "2        В   2.0  \n",
       "3      NaN   NaN  \n",
       "4  Хорошее  18.0  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>balcony</th>\n",
       "      <th>total_area</th>\n",
       "      <th>year</th>\n",
       "      <th>cost</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1824.000000</td>\n",
       "      <td>2261.000000</td>\n",
       "      <td>2038.00000</td>\n",
       "      <td>2.261000e+03</td>\n",
       "      <td>2038.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.010417</td>\n",
       "      <td>35.675807</td>\n",
       "      <td>1997.29735</td>\n",
       "      <td>2.779700e+06</td>\n",
       "      <td>21.70265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.116641</td>\n",
       "      <td>6.918724</td>\n",
       "      <td>21.00202</td>\n",
       "      <td>7.488420e+05</td>\n",
       "      <td>21.00202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>1925.00000</td>\n",
       "      <td>7.800000e+05</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1977.00000</td>\n",
       "      <td>2.280000e+06</td>\n",
       "      <td>4.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>2008.00000</td>\n",
       "      <td>2.650000e+06</td>\n",
       "      <td>11.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>2015.00000</td>\n",
       "      <td>3.100000e+06</td>\n",
       "      <td>42.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>2019.00000</td>\n",
       "      <td>5.300000e+06</td>\n",
       "      <td>94.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           balcony   total_area        year          cost         age\n",
       "count  1824.000000  2261.000000  2038.00000  2.261000e+03  2038.00000\n",
       "mean      1.010417    35.675807  1997.29735  2.779700e+06    21.70265\n",
       "std       0.116641     6.918724    21.00202  7.488420e+05    21.00202\n",
       "min       1.000000    20.000000  1925.00000  7.800000e+05     0.00000\n",
       "25%       1.000000    31.000000  1977.00000  2.280000e+06     4.00000\n",
       "50%       1.000000    35.000000  2008.00000  2.650000e+06    11.00000\n",
       "75%       1.000000    40.000000  2015.00000  3.100000e+06    42.00000\n",
       "max       3.000000    59.000000  2019.00000  5.300000e+06    94.00000"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Fill empty values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>balcony</th>\n",
       "      <th>walls</th>\n",
       "      <th>total_area</th>\n",
       "      <th>district</th>\n",
       "      <th>year</th>\n",
       "      <th>cost</th>\n",
       "      <th>type_house</th>\n",
       "      <th>repair</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>Панельный</td>\n",
       "      <td>30</td>\n",
       "      <td>Уралмаш</td>\n",
       "      <td>1963.0</td>\n",
       "      <td>2250000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Хорошее</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>Кирпичный</td>\n",
       "      <td>32</td>\n",
       "      <td>Эльмаш</td>\n",
       "      <td>1959.0</td>\n",
       "      <td>2090000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Хорошее</td>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>42</td>\n",
       "      <td>Ботанический</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>2800000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>В</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29</td>\n",
       "      <td>Ю-З</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2200000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>Панельный</td>\n",
       "      <td>34</td>\n",
       "      <td>Компрессорный</td>\n",
       "      <td>2001.0</td>\n",
       "      <td>2000000</td>\n",
       "      <td>Улучшенной</td>\n",
       "      <td>Хорошее</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   balcony      walls  total_area       district    year     cost  type_house  \\\n",
       "0      1.0  Панельный          30        Уралмаш  1963.0  2250000         NaN   \n",
       "1      NaN  Кирпичный          32         Эльмаш  1959.0  2090000         NaN   \n",
       "2      NaN        NaN          42   Ботанический  2017.0  2800000         NaN   \n",
       "3      1.0        NaN          29            Ю-З     NaN  2200000         NaN   \n",
       "4      1.0  Панельный          34  Компрессорный  2001.0  2000000  Улучшенной   \n",
       "\n",
       "    repair   age  \n",
       "0  Хорошее  56.0  \n",
       "1  Хорошее  60.0  \n",
       "2        В   2.0  \n",
       "3      NaN   NaN  \n",
       "4  Хорошее  18.0  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2261, 9)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Transformation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "district_map = {'автовокзал': 1, 'академический': 2, 'березит': 4,\n",
    "                'ботанический': 2, 'верхнемакарово': 4, 'виз': 1,\n",
    "                'вокзальный': 1, 'втузгородок': 1, 'горный щит': 4,\n",
    "                'елизавет': 3, 'жби': 2, 'завокзальный': 3, 'заречный': 1,\n",
    "                'изоплит': 4, 'исток': 4, 'калиновский': 4, 'карасьеозерск': 2,\n",
    "                'кольцово': 4, 'компрессорный': 3, 'лечебный': 3,\n",
    "                'малый исток': 4, 'медный': 4, 'н.сортировка': 2,\n",
    "                'нижнеисетский': 4, 'палкин.торфяник': 4, 'палкино': 4,\n",
    "                'палникс': 2, 'парковый': 1, 'пионерский': 1, 'полеводство': 4,\n",
    "                'птицефабрика': 4, 'рудный': 4, 'с.сортировка': 3, 'садовый': 4,\n",
    "                'северка': 4, 'семь ключей': 4, 'сибирский': 2,\n",
    "                'синие камни': 2, 'совхоз': 3, 'солнечный': 3,\n",
    "                'сулим.торфяник': 4, 'технопарк': 4, 'уктус': 2, 'унц': 3,\n",
    "                'уралмаш': 2, 'химмаш': 3, 'центр': 0, 'чермет': 3,\n",
    "                'чусовское озеро': 4, 'шабровский': 4, 'шарташ': 4,\n",
    "                'шарташский р-к': 1, 'широкая речка': 2, 'шувакиш': 4,\n",
    "                'эльмаш': 2, 'ю-з': 1}\n",
    "\n",
    "df.district = df.district.str.lower()\n",
    "df.replace({\"district\": district_map}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "repair = {'В': 1,\n",
    "          'Требует': 0.6,\n",
    "          'Косметический': 0.7,\n",
    "          'Хорошее': 0.8}\n",
    "df.repair.fillna('Косметический', inplace=True)\n",
    "df.replace({\"repair\": repair}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "walls_map = {'Кирпичный': 1.0,\n",
    "             'Кирпич': 1.0,\n",
    "             'Блочный': 0.8,\n",
    "             'Монолитный': 0.9,\n",
    "             'Деревянный': 0.4,\n",
    "             'Панельный': 0.8,\n",
    "             'Шлакоблочный': 0.6,\n",
    "             'Газозолоблоки': 0.9,\n",
    "             'Твин-блоки': 0.9,\n",
    "             'Твинблоки': 0.9,\n",
    "             'Пенозолобетон': 0.9,\n",
    "             'Бетонные': 0.8,\n",
    "             'Другой': 0.8,\n",
    "             'Железобетон': 0.8}\n",
    "mask = df[df['walls'].isna()][df.year >= 2010].index\n",
    "df.loc[mask, 'walls'] = 'Газозолоблоки'\n",
    "mask = df[df['walls'].isna()][df.year >= 2000].index\n",
    "df.loc[mask, 'walls'] = 'Бетонные'\n",
    "mask = df[df['walls'].isna()][df.year >= 1990].index\n",
    "df.loc[mask, 'walls'] = 'Панельный'\n",
    "mask = df[df['walls'].isna()].index\n",
    "df.loc[mask, 'walls'] = 'Шлакоблочный'\n",
    "\n",
    "df.replace({\"walls\": walls_map}, inplace=True)\n",
    "df.drop(columns=['year'],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.balcony.fillna(0,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>balcony</th>\n",
       "      <th>total_area</th>\n",
       "      <th>district</th>\n",
       "      <th>cost</th>\n",
       "      <th>repair</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2261.000000</td>\n",
       "      <td>2261.000000</td>\n",
       "      <td>2261.000000</td>\n",
       "      <td>2.261000e+03</td>\n",
       "      <td>2261.000000</td>\n",
       "      <td>2038.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.815126</td>\n",
       "      <td>35.675807</td>\n",
       "      <td>1.804069</td>\n",
       "      <td>2.779700e+06</td>\n",
       "      <td>0.806325</td>\n",
       "      <td>21.70265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.412591</td>\n",
       "      <td>6.918724</td>\n",
       "      <td>0.855022</td>\n",
       "      <td>7.488420e+05</td>\n",
       "      <td>0.139277</td>\n",
       "      <td>21.00202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.800000e+05</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.280000e+06</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>4.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.650000e+06</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>11.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.100000e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>42.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.300000e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>94.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           balcony   total_area     district          cost       repair  \\\n",
       "count  2261.000000  2261.000000  2261.000000  2.261000e+03  2261.000000   \n",
       "mean      0.815126    35.675807     1.804069  2.779700e+06     0.806325   \n",
       "std       0.412591     6.918724     0.855022  7.488420e+05     0.139277   \n",
       "min       0.000000    20.000000     0.000000  7.800000e+05     0.600000   \n",
       "25%       1.000000    31.000000     1.000000  2.280000e+06     0.700000   \n",
       "50%       1.000000    35.000000     2.000000  2.650000e+06     0.800000   \n",
       "75%       1.000000    40.000000     2.000000  3.100000e+06     1.000000   \n",
       "max       3.000000    59.000000     4.000000  5.300000e+06     1.000000   \n",
       "\n",
       "              age  \n",
       "count  2038.00000  \n",
       "mean     21.70265  \n",
       "std      21.00202  \n",
       "min       0.00000  \n",
       "25%       4.00000  \n",
       "50%      11.00000  \n",
       "75%      42.00000  \n",
       "max      94.00000  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df.drop(columns=['type_house'],inplace=True)\n",
    "df = df.astype(np.float64)\n",
    "df.dropna(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>balcony</th>\n",
       "      <th>walls</th>\n",
       "      <th>total_area</th>\n",
       "      <th>district</th>\n",
       "      <th>cost</th>\n",
       "      <th>repair</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2.038000e+03</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.823847</td>\n",
       "      <td>0.905839</td>\n",
       "      <td>35.800294</td>\n",
       "      <td>1.795388</td>\n",
       "      <td>2.799436e+06</td>\n",
       "      <td>0.808832</td>\n",
       "      <td>21.70265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.408405</td>\n",
       "      <td>0.087176</td>\n",
       "      <td>6.954855</td>\n",
       "      <td>0.851715</td>\n",
       "      <td>7.524326e+05</td>\n",
       "      <td>0.140238</td>\n",
       "      <td>21.00202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.800000e+05</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.300000e+06</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>4.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.680000e+06</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>11.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.116000e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>42.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.300000e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>94.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           balcony        walls   total_area     district          cost  \\\n",
       "count  2038.000000  2038.000000  2038.000000  2038.000000  2.038000e+03   \n",
       "mean      0.823847     0.905839    35.800294     1.795388  2.799436e+06   \n",
       "std       0.408405     0.087176     6.954855     0.851715  7.524326e+05   \n",
       "min       0.000000     0.400000    20.000000     0.000000  7.800000e+05   \n",
       "25%       1.000000     0.800000    31.000000     1.000000  2.300000e+06   \n",
       "50%       1.000000     0.900000    35.000000     2.000000  2.680000e+06   \n",
       "75%       1.000000     1.000000    40.000000     2.000000  3.116000e+06   \n",
       "max       3.000000     1.000000    59.000000     4.000000  5.300000e+06   \n",
       "\n",
       "            repair         age  \n",
       "count  2038.000000  2038.00000  \n",
       "mean      0.808832    21.70265  \n",
       "std       0.140238    21.00202  \n",
       "min       0.600000     0.00000  \n",
       "25%       0.700000     4.00000  \n",
       "50%       0.800000    11.00000  \n",
       "75%       1.000000    42.00000  \n",
       "max       1.000000    94.00000  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x792 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def show_correlation(df):\n",
    "    sns.set(style=\"whitegrid\")\n",
    "    corr = df.corr() * 100\n",
    "\n",
    "    # Select upper triangle of correlation matrix\n",
    "    mask = np.zeros_like(corr, dtype=np.bool)\n",
    "    mask[np.triu_indices_from(mask)] = True\n",
    "\n",
    "    # Set up the matplotlib figure\n",
    "    f, ax = plt.subplots(figsize=(15, 11))\n",
    "\n",
    "    # Generate a custom diverging colormap\n",
    "    cmap = sns.diverging_palette(220, 10)\n",
    "\n",
    "    # Draw the heatmap with the mask and correct aspect ratio\n",
    "    sns.heatmap(corr, mask=mask, cmap=cmap, center=0,\n",
    "                linewidths=1, cbar_kws={\"shrink\": .7}, annot=True,\n",
    "                fmt=\".2f\")\n",
    "    plot.show()\n",
    "    # df[columns] = scale(df[columns])\n",
    "    return df\n",
    "df1 = show_correlation(df.drop(columns=['cost']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([30., 32., 42., ..., 44., 45., 40.])"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['total_area'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0, 1.0, 0, array([30., 32., 42., ..., 44., 45., 40.]))"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_selection import chi2\n",
    "from scipy.stats import chi2_contingency\n",
    "chi2_contingency(df['total_area'], df['district'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f0d53fed748>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6,6))\n",
    "corr = df.corr()*100.0\n",
    "sns.heatmap(corr[['cost']],\n",
    "            cmap= sns.diverging_palette(220, 10),\n",
    "            center=0,\n",
    "            linewidths=1, cbar_kws={\"shrink\": .7}, annot=True,\n",
    "            fmt=\".2f\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cost</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>balcony</th>\n",
       "      <td>20.679068</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>walls</th>\n",
       "      <td>13.607172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_area</th>\n",
       "      <td>73.391879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>district</th>\n",
       "      <td>-41.234944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cost</th>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>repair</th>\n",
       "      <td>16.053086</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>-44.282268</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  cost\n",
       "balcony      20.679068\n",
       "walls        13.607172\n",
       "total_area   73.391879\n",
       "district    -41.234944\n",
       "cost        100.000000\n",
       "repair       16.053086\n",
       "age         -44.282268"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr[['cost']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split \n",
    "\n",
    "\n",
    "y = df.cost\n",
    "X = df.drop(columns=['cost'])\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naive predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>balcony</th>\n",
       "      <th>walls</th>\n",
       "      <th>total_area</th>\n",
       "      <th>district</th>\n",
       "      <th>repair</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.823847</td>\n",
       "      <td>0.905839</td>\n",
       "      <td>35.800294</td>\n",
       "      <td>1.795388</td>\n",
       "      <td>0.808832</td>\n",
       "      <td>21.70265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.408405</td>\n",
       "      <td>0.087176</td>\n",
       "      <td>6.954855</td>\n",
       "      <td>0.851715</td>\n",
       "      <td>0.140238</td>\n",
       "      <td>21.00202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>4.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>11.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>42.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>94.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           balcony        walls   total_area     district       repair  \\\n",
       "count  2038.000000  2038.000000  2038.000000  2038.000000  2038.000000   \n",
       "mean      0.823847     0.905839    35.800294     1.795388     0.808832   \n",
       "std       0.408405     0.087176     6.954855     0.851715     0.140238   \n",
       "min       0.000000     0.400000    20.000000     0.000000     0.600000   \n",
       "25%       1.000000     0.800000    31.000000     1.000000     0.700000   \n",
       "50%       1.000000     0.900000    35.000000     2.000000     0.800000   \n",
       "75%       1.000000     1.000000    40.000000     2.000000     1.000000   \n",
       "max       3.000000     1.000000    59.000000     4.000000     1.000000   \n",
       "\n",
       "              age  \n",
       "count  2038.00000  \n",
       "mean     21.70265  \n",
       "std      21.00202  \n",
       "min       0.00000  \n",
       "25%       4.00000  \n",
       "50%      11.00000  \n",
       "75%      42.00000  \n",
       "max      94.00000  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.describe()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score on LinearRegression is 76.67\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "76.67"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "def predict(X, y_test, model):\n",
    "    y = model.predict(X)\n",
    "    score = round((r2_score(y_test, y) * 100), 2)\n",
    "    print(f'Score on {model.__class__.__name__} is {score}')\n",
    "    return score\n",
    "\n",
    "def train_model(X, y, regressor):\n",
    "    model = regressor.fit(X, y)\n",
    "    return model\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "regressor = LinearRegression()\n",
    "model = train_model(X_train, y_train, regressor)\n",
    "predict(X_test, y_test, model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Crossvaladaion accuracy on LinearRegression is 73.32300000000001\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "73.32300000000001"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def do_cross_validation(X, y, model):\n",
    "    from sklearn.model_selection import KFold, cross_val_score\n",
    "    regressor_name = model.__class__.__name__\n",
    "    fold = KFold(n_splits=10, shuffle=True, random_state=0)\n",
    "    scores_on_this_split = cross_val_score(estimator=model, X=X,\n",
    "                                           y=y, cv=fold, scoring='r2')\n",
    "    scores_on_this_split = np.round(scores_on_this_split * 100, 2)\n",
    "\n",
    "    mean_accuracy = scores_on_this_split.mean()\n",
    "    print(f'Crossvaladaion accuracy on {model.__class__.__name__} is {mean_accuracy}')\n",
    "    return mean_accuracy\n",
    "do_cross_validation(X, y, model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>balcony</th>\n",
       "      <th>walls</th>\n",
       "      <th>total_area</th>\n",
       "      <th>district</th>\n",
       "      <th>cost</th>\n",
       "      <th>repair</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2.038000e+03</td>\n",
       "      <td>2038.000000</td>\n",
       "      <td>2038.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.823847</td>\n",
       "      <td>0.905839</td>\n",
       "      <td>35.800294</td>\n",
       "      <td>1.795388</td>\n",
       "      <td>2.799436e+06</td>\n",
       "      <td>0.808832</td>\n",
       "      <td>21.70265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.408405</td>\n",
       "      <td>0.087176</td>\n",
       "      <td>6.954855</td>\n",
       "      <td>0.851715</td>\n",
       "      <td>7.524326e+05</td>\n",
       "      <td>0.140238</td>\n",
       "      <td>21.00202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.800000e+05</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.300000e+06</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>4.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.680000e+06</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>11.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.116000e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>42.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.300000e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>94.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           balcony        walls   total_area     district          cost  \\\n",
       "count  2038.000000  2038.000000  2038.000000  2038.000000  2.038000e+03   \n",
       "mean      0.823847     0.905839    35.800294     1.795388  2.799436e+06   \n",
       "std       0.408405     0.087176     6.954855     0.851715  7.524326e+05   \n",
       "min       0.000000     0.400000    20.000000     0.000000  7.800000e+05   \n",
       "25%       1.000000     0.800000    31.000000     1.000000  2.300000e+06   \n",
       "50%       1.000000     0.900000    35.000000     2.000000  2.680000e+06   \n",
       "75%       1.000000     1.000000    40.000000     2.000000  3.116000e+06   \n",
       "max       3.000000     1.000000    59.000000     4.000000  5.300000e+06   \n",
       "\n",
       "            repair         age  \n",
       "count  2038.000000  2038.00000  \n",
       "mean      0.808832    21.70265  \n",
       "std       0.140238    21.00202  \n",
       "min       0.600000     0.00000  \n",
       "25%       0.700000     4.00000  \n",
       "50%       0.800000    11.00000  \n",
       "75%       1.000000    42.00000  \n",
       "max       1.000000    94.00000  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def estimate_model(model):\n",
    "    \n",
    "    sns.set(style=\"white\", context=\"talk\")\n",
    "    f, ax = plot.subplots(1, 1, figsize=(10, 10), sharex=True)\n",
    "    sns.barplot(x=model.coef_, y=X.columns, palette=\"vlag\", ax=ax)\n",
    "    for i, v in enumerate(model.coef_.astype(int)):\n",
    "        ax.text(v + 3, i + .25, str(v), color='black')\n",
    "\n",
    "    ax.set_title(f\"Coefficients\")\n",
    "estimate_model(regressor)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}