Modeling Cian.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CIAN.ru"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import os, gc\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import cross_val_score,train_test_split\n",
    "\n",
    "from sklearn.metrics import make_scorer, mean_squared_error,  mean_absolute_error\n",
    "from bayes_opt import BayesianOptimization\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data = pd.read_csv('Data_cian_full.csv',index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>District</th>\n",
       "      <th>Price</th>\n",
       "      <th>Tel</th>\n",
       "      <th>combined_bs</th>\n",
       "      <th>cooking_space</th>\n",
       "      <th>dist</th>\n",
       "      <th>floor</th>\n",
       "      <th>house_type</th>\n",
       "      <th>lat</th>\n",
       "      <th>link</th>\n",
       "      <th>living_space</th>\n",
       "      <th>lon</th>\n",
       "      <th>metro_station</th>\n",
       "      <th>metro_time</th>\n",
       "      <th>new</th>\n",
       "      <th>rooms</th>\n",
       "      <th>separate_bs</th>\n",
       "      <th>total_floors</th>\n",
       "      <th>total_space</th>\n",
       "      <th>walk</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>528391080</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.720885</td>\n",
       "      <td>8</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.749603</td>\n",
       "      <td>150090567</td>\n",
       "      <td>NaN</td>\n",
       "      <td>37.597872</td>\n",
       "      <td>Арбатская</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>399.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>311298500</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.512113</td>\n",
       "      <td>5</td>\n",
       "      <td>монолитный</td>\n",
       "      <td>55.757526</td>\n",
       "      <td>150453732</td>\n",
       "      <td>165.0</td>\n",
       "      <td>37.590632</td>\n",
       "      <td>Баррикадная</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>14000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>8.9</td>\n",
       "      <td>6.482659</td>\n",
       "      <td>5</td>\n",
       "      <td>кирпичный</td>\n",
       "      <td>55.743831</td>\n",
       "      <td>150613273</td>\n",
       "      <td>50.9</td>\n",
       "      <td>37.679835</td>\n",
       "      <td>Римская</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.0</td>\n",
       "      <td>75.5</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>4403057</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.864250</td>\n",
       "      <td>3</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.775753</td>\n",
       "      <td>149031353</td>\n",
       "      <td>22.7</td>\n",
       "      <td>37.671938</td>\n",
       "      <td>Красносельская</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22.7</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>4814581</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.903571</td>\n",
       "      <td>2</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.776294</td>\n",
       "      <td>148567438</td>\n",
       "      <td>NaN</td>\n",
       "      <td>37.672163</td>\n",
       "      <td>Бауманская</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>17.3</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   District      Price  Tel  combined_bs  cooking_space      dist  floor  \\\n",
       "0         0  528391080    0            0            NaN  2.720885      8   \n",
       "1         0  311298500    1            3            NaN  3.512113      5   \n",
       "2         0   14000000    1            0            8.9  6.482659      5   \n",
       "3         0    4403057    0            1            NaN  5.864250      3   \n",
       "4         0    4814581    0            1            NaN  5.903571      2   \n",
       "\n",
       "   house_type        lat       link  living_space        lon   metro_station  \\\n",
       "0        Else  55.749603  150090567           NaN  37.597872       Арбатская   \n",
       "1  монолитный  55.757526  150453732         165.0  37.590632     Баррикадная   \n",
       "2   кирпичный  55.743831  150613273          50.9  37.679835         Римская   \n",
       "3        Else  55.775753  149031353          22.7  37.671938  Красносельская   \n",
       "4        Else  55.776294  148567438           NaN  37.672163      Бауманская   \n",
       "\n",
       "   metro_time  new  rooms  separate_bs  total_floors  total_space  walk  \n",
       "0         7.0    0    NaN            0           9.0        399.0   1.0  \n",
       "1        10.0    0    5.0            0           7.0        200.0   1.0  \n",
       "2         5.0    0    3.0            1           5.0         75.5   NaN  \n",
       "3         6.0    1    1.0            0           NaN         22.7   1.0  \n",
       "4         5.0    0    1.0            0           5.0         17.3   1.0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sum(data.duplicated())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4397, 20)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "District            0\n",
       "Price               0\n",
       "Tel                 0\n",
       "combined_bs         0\n",
       "cooking_space     781\n",
       "dist                0\n",
       "floor               0\n",
       "house_type          0\n",
       "lat                 0\n",
       "link                0\n",
       "living_space      444\n",
       "lon                 0\n",
       "metro_station     599\n",
       "metro_time        625\n",
       "new                 0\n",
       "rooms               4\n",
       "separate_bs         0\n",
       "total_floors       68\n",
       "total_space         0\n",
       "walk             2040\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data['walk'] = data['walk'].fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data['metro_station'] = data['metro_station'].fillna('Нет') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xcbc9710>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAqgAAAHcCAYAAAAa41gWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAPYQAAD2EBqD+naQAAIABJREFUeJzt3X1wVPWh//EPSbqbiGyNeYKmxEp0JAmYxDx0qEZ/gw+d\nWsCZLtzpg21DrZlKQO/12jbQ1gesBcWiNjwUItUWx6lpttZSO/QOrVclRIsJmJSIulExoSTZUCAp\nbjgm2d8fHfZ2DUHO7gn5JrxfM4yT8z373e/55Cx+2D05mRQKhUICAAAADBE31gsAAAAA/h0FFQAA\nAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgAAAAwiu2C+o9//EO3\n3367SkpK9PnPf17PPvtseKyjo0OLFy9WYWGh5s2bp/r6+ojH7tq1S/Pnz1dBQYHKy8vV3t4e+xEA\nAABgQrFdUJcsWaLu7m5t3bpVK1as0OrVq7Vjx47wWHp6unw+nxYsWKClS5eqs7NTknTo0CFVVlbK\n6/XK5/MpOTlZlZWVzh4NAAAAxr1JoVAodKY7/+1vf9OiRYu0Y8cOZWZmSpJqamr05z//Wf/5n/+p\nJUuWqKGhQW63W5K0ePFiFRUVaenSpXrsscfU2NioX/3qV5Kk/v5+XXnllfr5z3+ukpKSUTg0AAAA\njEe23kFtb2/XhRdeGC6nknTZZZfpb3/7m1577TXl5eWFy6kkFRUVae/evZKk5ubmiCKamJio3Nxc\n7dmzJ9ZjAAAAwARiq6Cmpqaqt7dXJ06cCG87dOiQBgYGdPjwYaWnp0fsn5KSoq6uLklSd3f3sPHU\n1NTwOAAAACDZLKj5+flKS0vTypUrFQwGdeDAAT355JOaNGmSTpw4IZfLFbG/y+WSZVmS/vWR/unG\nAQAAAMlmQXW5XPrZz36mV199VUVFRbr55pv15S9/WaFQSHFxccPKpmVZSkxMlCS53e7Tjp8JG5fL\nAgAAYJxKsPuAWbNmaceOHTp8+LCSk5P18ssv68ILL1RWVpZ27twZsW9PT4/S0tIkSRkZGQoEAsPG\nc3Jyzvi5J02apN7eoAYHh+wu+5wUHx8njyeJzGwiN/vILDrkZh+ZRYfc7COz6JzMLVa2CuqxY8d0\n2223aePGjUpJSZEk/e///q9KS0t1+eWXa9OmTbIsK/xRfmNjo4qLiyX96/KApqam8FzBYFCtra1a\ntmyZrQUPDg5pYIATxQ4yiw652Udm0SE3+8gsOuRmH5mNDVsf8X/yk59UMBjUmjVr1N7ert/85jd6\n9tlndeutt6q0tFSf+tSnVFVVJb/fr82bN6ulpUULFy6UJHm9XjU1NammpkZ+v1/Lly9XVlaWSktL\nR+XAAAAAMD7ZvlH/I488ogMHDmjBggX61a9+pccee0x5eXmKi4vThg0bFAgE5PV6tW3bNq1fv15T\np06VJGVmZqq6ulo+n0+LFi1SX1+f1q1b5/gBAQAAYHyzdaN+Exw5cpy32s9QQkKckpMnk5lN5GYf\nmUWH3Owjs+iQm31kFp2TucXK9juoAAAAwGiioAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAA\nAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwCgUVAAAABiFggoAAACjUFABAABgFAoqAAAAjEJB\nBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRUAAAAGIWCCgAAAKNQUAEAAGAUCioAAACMQkEFAACA\nUSioAAAAMAoFFQAAAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgA\nAAAwSsJYLwDjj2VZ2revxbH58vJmy+VyOTYfAAAY3yiosG3fvhZ9b+1vNSUlK+a5+g6/r4fulAoL\nixxYGQAAmAgoqIjKlJQsXTD10rFeBgAAmIBsX4Pa2dmp73znOyoqKtK1116rX/7yl+Gxjo4OLV68\nWIWFhZo3b57q6+sjHrtr1y7Nnz9fBQUFKi8vV3t7e+xHAAAAgAnFdkG94447NHnyZD377LNasWKF\nHn30Ue3YsUOStGTJEqWnp8vn82nBggVaunSpOjs7JUmHDh1SZWWlvF6vfD6fkpOTVVlZ6ezRAAAA\nYNyzVVB7e3v1+uuv67bbblNWVpauvfZalZWV6ZVXXtErr7yijo4OrVy5UjNmzFBFRYUKCgpUV1cn\nSaqtrdXs2bNVXl6u7OxsrVq1SgcPHtTu3btH5cAAAAAwPtkqqImJiUpKSpLP59PAwIDeeecdNTU1\nKScnR6+//rry8vLkdrvD+xcVFWnv3r2SpObmZpWUlETMlZubqz179jh0KAAAAJgIbBVUl8ulu+++\nW7/+9a+Vn5+vG2+8UVdffbW8Xq8CgYDS09Mj9k9JSVFXV5ckqbu7e9h4ampqeBwAAACQovgp/ra2\nNs2dO1e33HKL3nrrLd1///2aM2eOgsHgsHtZulwuWZYlServ7z/t+JmKj+d3C5ypk1k5ndlozJeQ\nYM73dbRym8jILDrkZh+ZRYfc7COz6DiVl62C2tDQoLq6Or300ktyuVzKzc1VZ2enNm7cqDlz5ujo\n0aMR+1uWpcTEREmS2+0eVkYty5LH47G1YI8nydb+cD6z0ZgvOXmyo3M6gXPNPjKLDrnZR2bRITf7\nyGxs2Cqo+/bt02c+85mId0JzcnK0adMmZWRk6O23347Yv6enR2lpaZKkjIwMBQKBYeM5OTm2Ftzb\nG9Tg4JCtx5yr4uPj5PEkOZ5Zb2/QsblOznfkyHFH54zFaOU2kZFZdMjNPjKLDrnZR2bROZlbrGwV\n1PT0dB04cEADAwNKSPjXQ9955x19+tOfVn5+vjZt2iTLssIFtrGxUcXFxZKk/Px8NTU1hecKBoNq\nbW3VsmXLbC14cHBIAwOcKHY4nZnTL1RTv6emrstkZBYdcrOPzKJDbvaR2diwdaHA3LlzlZCQoB/+\n8Id677339Je//EWbNm3SN77xDZWUlGjatGmqqqqS3+/X5s2b1dLSooULF0qSvF6vmpqaVFNTI7/f\nr+XLlysrK0ulpaWjcmAAAAAYn2wV1PPPP19PPvmkAoGAFi1apAcffFCVlZVatGiR4uLitHHjRgUC\nAXm9Xm3btk3r16/X1KlTJUmZmZmqrq6Wz+fTokWL1NfXp3Xr1o3KQQEAAGD8sv1T/NnZ2dqyZcsp\nx6ZPn66tW7eO+NiysjJt377d7lMCAADgHMK9EwAAAGAUCioAAACMQkEFAACAUSioAAAAMAoFFQAA\nAEahoAIAAMAotm8zhbPPsizt29di+3Gn+jVteXmzI35VLQAAgGkoqOPAvn0t+t7a32pKSlZM8/Qd\nfl8P3SkVFhY5tDIAAADnUVDHiSkpWbpg6qVjvQwAAIBRxzWoAAAAMAoFFQAAAEahoAIAAMAoFFQA\nAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwCgUVAAAABiF\nggoAAACjUFABAABgFAoqAAAAjEJBBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRUAAAAGIWCCgAA\nAKNQUAEAAGAUCioAAACMQkEFAACAUSioAAAAMAoFFQAAAEahoAIAAMAotgrqs88+q5kzZyonJyfi\nv7m5uZKk9vZ2LV68WIWFhZo3b57q6+sjHr9r1y7Nnz9fBQUFKi8vV3t7u3NHAgAAgAnBVkH94he/\nqPr6eu3cuVP19fV64YUXdNFFF+mb3/ymJKmyslLp6eny+XxasGCBli5dqs7OTknSoUOHVFlZKa/X\nK5/Pp+TkZFVWVjp/RAAAABjXbBVUl8ullJSU8J/nnntOknTnnXeqoaFBHR0dWrlypWbMmKGKigoV\nFBSorq5OklRbW6vZs2ervLxc2dnZWrVqlQ4ePKjdu3c7f1QAAAAYt6K+BvXYsWN6/PHHddddd+kT\nn/iEmpublZeXJ7fbHd6nqKhIe/fulSQ1NzerpKQkPJaYmKjc3Fzt2bMnhuUDAABgoom6oD799NPK\nyMjQ9ddfL0kKBAJKT0+P2CclJUVdXV2SpO7u7mHjqamp4XEAAABAkhKifWBdXZ0qKirCXweDQblc\nroh9XC6XLMuSJPX39592/EzFx597Nx5w8pjj4+OUkBDbfE5/D5xYk5NOHt+5eK5Fi8yiQ272kVl0\nyM0+MouOU3lFVVCbm5vV1dWlG2+8MbzN7Xbr2LFjEftZlqXExMTw+EfLqGVZ8ng8tp7b40mKZsnj\nmpPH7PEkKTl5csxzOMmJNY2Gc/FcixWZRYfc7COz6JCbfWQ2NqIqqDt37lRJSYmmTJkS3paRkSG/\n3x+xX09Pj9LS0sLjgUBg2HhOTo6t5+7tDWpwcCiaZY9bvb1BR+c6cuR4zHM4yYk1OSk+Pk4eT9I5\nea5Fi8yiQ272kVl0yM0+MovOydxiFfU7qFdccUXEtvz8fNXU1MiyrPBH+Y2NjSouLg6PNzU1hfcP\nBoNqbW3VsmXLbD334OCQBgbOrRPFyReGE/k5/UI19Xtq6rpMRmbRITf7yCw65GYfmY2NqC4UeOut\nt5SdnR2xrbS0VNOmTVNVVZX8fr82b96slpYWLVy4UJLk9XrV1NSkmpoa+f1+LV++XFlZWSotLY39\nKAAAADBhRFVQ//GPf+iTn/xk5ERxcdqwYYMCgYC8Xq+2bdum9evXa+rUqZKkzMxMVVdXy+fzadGi\nRerr69O6detiPwIAAABMKFF9xH/y3qYfNX36dG3dunXEx5WVlWn79u3RPCUAAADOEdw7AQAAAEah\noAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAA\nwCgUVAAAABiFggoAAACjUFABAABgFAoqAAAAjEJBBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRU\nAAAAGIWCCgAAAKNQUAEAAGAUCioAAACMQkEFAACAUSioAAAAMAoFFQAAAEahoAIAAMAoFFQAAAAY\nhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwCgUVAAAABjFdkG1\nLEv33XefSktLddVVV+mRRx4Jj3V0dGjx4sUqLCzUvHnzVF9fH/HYXbt2af78+SooKFB5ebna29tj\nPwIAAABMKLYL6o9//GM1NDToF7/4hR5++GHV1taqtrZWkrRkyRKlp6fL5/NpwYIFWrp0qTo7OyVJ\nhw4dUmVlpbxer3w+n5KTk1VZWens0QAAAGDcS7Cz87Fjx/Tb3/5WTz75pGbNmiVJ+ta3vqXXX39d\nWVlZ6ujo0G9+8xu53W5VVFSooaFBdXV1Wrp0qWprazV79myVl5dLklatWqUrr7xSu3fvVklJieMH\nBgAAgPHJVkFtbGzUlClTVFxcHN526623SpI2bdqkvLw8ud3u8FhRUZH27t0rSWpubo4ooomJicrN\nzdWePXsoqAAAAAiz9RF/e3u7MjMz9bvf/U5f+MIXdN1112nDhg0KhUIKBAJKT0+P2D8lJUVdXV2S\npO7u7mHjqamp4XEAAABAsvkO6gcffKD33ntPtbW1Wr16tQKBgO6++24lJSUpGAzK5XJF7O9yuWRZ\nliSpv7//tONnKj7+3LvxgJPHHB8fp4SE2OZz+nvgxJqcdPL4zsVzLVpkFh1ys4/MokNu9pFZdJzK\ny1ZBjY+P1/Hjx7V27VpNnTpVknTw4EE9/fTTuuqqq3T06NGI/S3LUmJioiTJ7XYPK6OWZcnj8dha\nsMeTZGv/icDJY/Z4kpScPDnmOZzkxJpGw7l4rsWKzKJDbvaRWXTIzT4yGxu2Cmp6errcbne4nErS\nxRdfrK6uLmVkZOjtt9+O2L+np0dpaWmSpIyMDAUCgWHjOTk5thbc2xvU4OCQrceMd729QUfnOnLk\neMxzOMmJNTkpPj5OHk/SOXmuRYvMokNu9pFZdMjNPjKLzsncYmWroObn5+vEiRM6cOCALrroIklS\nW1ubMjMzlZ+fr02bNsmyrPBH+Y2NjeEfqMrPz1dTU1N4rmAwqNbWVi1btszWggcHhzQwcG6dKE6+\nMJzIz+kXqqnfU1PXZTIyiw652Udm0SE3+8hsbNi6UODiiy/WNddco6qqKu3fv18vv/yyampq9NWv\nflUlJSWaNm2aqqqq5Pf7tXnzZrW0tGjhwoWSJK/Xq6amJtXU1Mjv92v58uXKyspSaWnpqBwYAAAA\nxifbV7I+/PDDuuiii/S1r31Ny5cv19e//nV97WtfU1xcnDZu3KhAICCv16tt27Zp/fr14csBMjMz\nVV1dLZ/Pp0WLFqmvr0/r1q1z/IAAAAAwvtn6iF+Szj//fK1evVqrV68eNjZ9+nRt3bp1xMeWlZVp\n+/btdp8SAAAA5xDunQAAAACjUFABAABgFAoqAAAAjEJBBQAAgFEoqAAAADAKBRUAAABGoaACAADA\nKBRUAAAAGIWCCgAAAKPY/k1SE51lWdq3r8WRufLyZsvlcjkyFwAAwLmCgvoR+/a16Htrf6spKVkx\nzXMs8K5unb9fl102M+Y1vfnm/pjnAAAAGC8oqKcwJSVLF0y9NKY5+g63a8vzrZryyj9jXk/XO7uV\nMaMk5nkAAADGAwrqKHKi6Er/KrsAAADnCn5ICgAAAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAA\nYBQKKgAAAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwCgUVAAAABiFggoAAACjUFABAABgFAoq\nAAAAjEJBBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRUAAAAGIWCCgAAAKNQUAEAAGAUCioAAACM\nYrug7tixQzNnzlROTk74v3fccYckqaOjQ4sXL1ZhYaHmzZun+vr6iMfu2rVL8+fPV0FBgcrLy9Xe\n3u7MUQAAAGDCsF1Q/X6/5s6dq/r6etXX12vnzp164IEHJElLlixRenq6fD6fFixYoKVLl6qzs1OS\ndOjQIVVWVsrr9crn8yk5OVmVlZXOHg0AAADGPdsFta2tTZdeeqkuvPBCpaSkKCUlReeff74aGhrU\n0dGhlStXasaMGaqoqFBBQYHq6uokSbW1tZo9e7bKy8uVnZ2tVatW6eDBg9q9e7fjBwUAAIDxK6qC\nevHFFw/b3tzcrLy8PLnd7vC2oqIi7d27NzxeUlISHktMTFRubq727NkTzboBAAAwQdkuqO+++65e\nfvllff7zn9f111+vn/70p/rwww8VCASUnp4esW9KSoq6urokSd3d3cPGU1NTw+MAAACAJCXY2fnv\nf/+7+vv75Xa79dhjj6mjo0MPPPCA+vv7FQwG5XK5IvZ3uVyyLEuS1N/ff9rxMxUfP7o3Hhjt+cda\nfHycEhJiO0anM3JiTU46eXwT/VxwEplFh9zsI7PokJt9ZBYdp/KyVVA/9alP6dVXX5XH45EkzZw5\nU0NDQ/rud7+rL33pS+rt7Y3Y37IsJSYmSpLcbvewMmpZVniuM+XxJNna367Rnn+seTxJSk6eHPMc\nTnJiTaNhop8Lo4HMokNu9pFZdMjNPjIbG7YKqqRhhTI7O1snTpxQamqq2traIsZ6enqUlpYmScrI\nyFAgEBg2npOTY+v5e3uDGhwcsrtsW/NPZL29QR05cjzmOZzkxJqcFB8fJ48nadTPtYmEzKJDbvaR\nWXTIzT4yi87J3GJlq6Du3LlT//3f/62XXnop/MNQra2tSk5OVnFxsX7xi1/IsqzwR/mNjY0qLi6W\nJOXn56upqSk8VzAYVGtrq5YtW2ZrwYODQxoYGL0TZaKfhE7k53RGo/09jZap6zIZmUWH3Owjs+iQ\nm31kNjZsXShQWFiopKQk/eAHP9C7776rF198UWvWrNGtt96qkpISTZs2TVVVVfL7/dq8ebNaWlq0\ncOFCSZLX61VTU5Nqamrk9/u1fPlyZWVlqbS0dFQODAAAAOOTrYI6efJkbdmyRUeOHNHChQv1ox/9\nSF/+8pf1rW99S3Fxcdq4caMCgYC8Xq+2bdum9evXa+rUqZKkzMxMVVdXy+fzadGiRerr69O6detG\n5aAAAAAwftm+BjU7O1tbtmw55dj06dO1devWER9bVlam7du3231KAAAAnEO4dwIAAACMQkEFAACA\nUSioAAAAMAoFFQAAAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgA\nAAAwCgUVAAAARqGgAgAAwCgUVAAAABiFggoAAACjUFABAABgFAoqAAAAjEJBBQAAgFEoqAAAADAK\nBRUAAABGoaACAADAKBRUAAAAGIWCCgAAAKNQUAEAAGAUCioAAACMQkEFAACAUSioAAAAMAoFFQAA\nAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgAAAAwCgUVAAAARom6\noFZUVGj58uXhrzs6OrR48WIVFhZq3rx5qq+vj9h/165dmj9/vgoKClReXq729vboVw0AAIAJK6qC\n+vzzz+ull16K2FZZWan09HT5fD4tWLBAS5cuVWdnpyTp0KFDqqyslNfrlc/nU3JysiorK2NfPQAA\nACYc2wX12LFjWrNmjS6//PLwtoaGBrW3t2vlypWaMWOGKioqVFBQoLq6OklSbW2tZs+erfLycmVn\nZ2vVqlU6ePCgdu/e7dyRAAAAYEKwXVAffPBB3XTTTcrOzg5va25uVl5entxud3hbUVGR9u7dGx4v\nKSkJjyUmJio3N1d79uyJZe0AAACYgGwV1IaGBjU2Ng77eD4QCCg9PT1iW0pKirq6uiRJ3d3dw8ZT\nU1PD4wAAAMBJCWe6o2VZuvfee3XPPffI5XJFjAWDwWHbXC6XLMuSJPX395923I74+NG98cBozz/W\n4uPjlJAQ2zE6nZETa3LSyeOb6OeCk8gsOuRmH5lFh9zsI7PoOJXXGRfU6upqzZo1S5/73OeGjbnd\nbh07dixim2VZSkxMDI9/tIxaliWPx2N7wR5Pku3HmDT/WPN4kpScPDnmOZzkxJpGw0Q/F0YDmUWH\n3Owjs+iQm31kNjbOuKD+8Y9/1OHDh1VYWChJ+vDDDyVJf/rTn/Sd73xHfr8/Yv+enh6lpaVJkjIy\nMhQIBIaN5+Tk2F5wb29Qg4NDth9nZ/6JrLc3qCNHjsc8h5OcWJOT4uPj5PEkjfq5NpGQWXTIzT4y\niw652Udm0TmZW6zOuKA+9dRTGhgYCH+9Zs0aSdJ3v/tdHTx4UJs3b5ZlWeGP8hsbG1VcXCxJys/P\nV1NTU/ixwWBQra2tWrZsme0FDw4OaWBg9E6UiX4SOpGf0xmN9vc0Wqauy2RkFh1ys4/MokNu9pHZ\n2DjjCwWmTZum6dOnh/9MnjxZkydP1vTp01VaWqpp06apqqpKfr9fmzdvVktLixYuXChJ8nq9ampq\nUk1Njfx+v5YvX66srCyVlpaO2oEBAABgfHLkSta4uDht2LBBgUBAXq9X27Zt0/r16zV16lRJUmZm\npqqrq+Xz+bRo0SL19fVp3bp1Tjw1AAAAJpgz/oj/o1atWhXx9fTp07V169YR9y8rK9P27dujfToA\nAACcI7h3AgAAAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwCgUVAAAABiFggoAAACjUFABAABg\nFAoqAAAAjEJBBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRUAAAAGIWCCgAAAKNQUAEAAGAUCioA\nAACMQkEFAACAUSioAAAAMAoFFQAAAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxC\nQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwCgUVAAAABiFggoAAACjUFABAABgFAoqAAAAjEJBBQAA\ngFEoqAAAADCK7YL6/vvv65ZbblFhYaHmzp2rLVu2hMc6Ojq0ePFiFRYWat68eaqvr4947K5duzR/\n/nwVFBSovLxc7e3tsR8BAAAAJhRbBTUUCqmiokKpqal67rnndO+992rjxo16/vnnJUlLlixRenq6\nfD6fFixYoKVLl6qzs1OSdOjQIVVWVsrr9crn8yk5OVmVlZXOHxEAAADGNVsFtaenR7m5ubrnnnuU\nlZWlq6++WnPmzFFjY6NeeeUVdXR0aOXKlZoxY4YqKipUUFCguro6SVJtba1mz56t8vJyZWdna9Wq\nVTp48KB27949KgcGAACA8clWQU1LS9PatWt13nnnSZIaGxv12muvqbS0VK+//rry8vLkdrvD+xcV\nFWnv3r2SpObmZpWUlITHEhMTlZubqz179jhxHAAAAJggov4hqblz5+rmm29WQUGBbrjhBgUCAaWn\np0fsk5KSoq6uLklSd3f3sPHU1NTwOAAAACBJCdE+sLq6Wj09Pbr33nv1k5/8RMFgUC6XK2Ifl8sl\ny7IkSf39/acdP1Px8aN744HRnn+sxcfHKSEhtmN0OiMn1uSkk8c30c8FJ5FZdMjNPjKLDrnZR2bR\ncSqvqAtqXl6eJKmqqkp33XWXFi5cqN7e3oh9LMtSYmKiJMntdg8ro5ZlyePx2Hpejycp2iUbMf9Y\n83iSlJw8OeY5nOTEmkbDRD8XRgOZRYfc7COz6JCbfWQ2NmwV1MOHD2vPnj267rrrwtsuueQSffjh\nh0pLS1NbW1vE/j09PUpLS5MkZWRkKBAIDBvPycmxteDe3qAGB4dsPcbu/BNZb29QR44cj3kOJzmx\nJifFx8fJ40ka9XNtIiGz6JCbfWQWHXKzj8yiczK3WNkqqB0dHVq2bJlefPHF8PWkLS0tSklJUVFR\nkbZs2SLLssIf5Tc2Nqq4uFiSlJ+fr6ampvBcwWBQra2tWrZsma0FDw4OaWBg9E6UiX4SOpGf0xmN\n9vc0Wqauy2RkFh1ys4/MokNu9pHZ2LB1ocDs2bM1a9YsrVixQm1tbXrxxRf18MMP67bbblNJSYmm\nTZumqqoq+f1+bd68WS0tLVq4cKEkyev1qqmpSTU1NfL7/Vq+fLmysrJUWlo6KgcGAACA8clWQY2L\ni9OGDRsvLgWTAAAXtUlEQVR03nnn6ctf/rJ+9KMf6Rvf+IZuvvlmxcXFaePGjQoEAvJ6vdq2bZvW\nr1+vqVOnSpIyMzNVXV0tn8+nRYsWqa+vT+vWrRuVgwIAAMD4ZfuHpNLS0vSzn/3slGPTp0/X1q1b\nR3xsWVmZtm/fbvcpAQAAcA7h3gkAAAAwCgUVAAAARon6PqimGRoa0tBQ7D9lNzg46MBqAAAAEK0J\nU1AfXPszNb39j5jn6e54WymX/r/YFwQAAICoTJiCOjQkTbn4/8U8T9+Hro/fCQAAAKOGa1ABAABg\nFAoqAAAAjEJBBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRUAAAAGIWCCgAAAKNQUAEAAGAUCioA\nAACMQkEFAACAUSioAAAAMAoFFQAAAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAAAIxC\nQQUAAIBRKKgAAAAwSsJYLwBnz9DggN58c3/M8zgxBwAAwEgoqOeQ40cPacvzf9eUV/4Z0zxd7+xW\nxowSh1YFAAAQiYJ6jpmSkqULpl4a0xx9h9sdWo15LMtSc/M+eTxJ6u0NanBwKKb58vJmy+VyObQ6\nAADODRRU4N/s29ei7639raakZMU8V9/h9/XQnVJhYZEDKwMA4NxBQQU+wol3mQEAQPT4KX4AAAAY\nhYIKAAAAo1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgAAAAwiq2C2tXVpdtvv12f/exndc0112j16tWy\nLEuS1NHRocWLF6uwsFDz5s1TfX19xGN37dql+fPnq6CgQOXl5Wpvn7g3ewcAAED0bBXU22+/XSdO\nnNDTTz+ttWvX6oUXXtBjjz0mSVqyZInS09Pl8/m0YMECLV26VJ2dnZKkQ4cOqbKyUl6vVz6fT8nJ\nyaqsrHT+aAAAADDunXFBfeedd9Tc3KxVq1YpOztbRUVFuv322/WHP/xBr7zyijo6OrRy5UrNmDFD\nFRUVKigoUF1dnSSptrZWs2fPVnl5ubKzs7Vq1SodPHhQu3fvHrUDAwAAwPh0xgU1LS1Njz/+uC68\n8MKI7X19fXr99deVl5cnt9sd3l5UVKS9e/dKkpqbm1VSUhIeS0xMVG5urvbs2RPr+gEAADDBnHFB\nnTJliq688srw16FQSE899ZTmzJmjQCCg9PT0iP1TUlLU1dUlSeru7h42npqaGh4HAAAATkqI9oEP\nPfSQ3njjDdXV1emJJ56Qy+WKGHe5XOEfoOrv7z/tuB3x8afu1JPiJklDtqeDAeLj45SQYMYNJUY6\nv2KZz5RjG00nc3M6v4mO3Owjs+iQm31kFh2n8oqqoK5Zs0Zbt27Vo48+qksuuURut1vHjh2L2Mey\nLCUmJkqS3G73sDJqWZY8Ho/t5/Z4kk653e3+hDRgezoYwONJUnLy5LFehqSRz69Y5jPl2M4Gp/M7\nV5CbfWQWHXKzj8zGhu2Cev/99+uZZ57RmjVrdN1110mSMjIy5Pf7I/br6elRWlpaeDwQCAwbz8nJ\nsb3g3t6gBgeHv1V64sSHtueCGXp7gzpy5PhYL0PSv9bi9HymHNtoio+Pk8eTNOLrE6dGbvaRWXTI\nzT4yi87J3GJlq6CuW7dOzzzzjB555BFdf/314e35+fmqqamRZVnhj/IbGxtVXFwcHm9qagrvHwwG\n1draqmXLltle8ODgkAYGhp8ooaGQ7blghpG+p2PB6b+ETDq2s+FcO16nkJt9ZBYdcrOPzMbGGV8o\n0NbWpo0bN6qiokKFhYXq6ekJ/yktLdW0adNUVVUlv9+vzZs3q6WlRQsXLpQkeb1eNTU1qaamRn6/\nX8uXL1dWVpZKS0tH7cAAAAAwPp1xQf3zn/+soaEhbdy4UWVlZSorK9NVV12lsrIyxcXFaf369QoE\nAvJ6vdq2bZvWr1+vqVOnSpIyMzNVXV0tn8+nRYsWqa+vT+vWrRu1gwIAAMD4dcYf8VdUVKiiomLE\n8aysLG3dunXE8bKyMm3fvt3e6gAAAHDO4d4JAAAAMAoFFQAAAEahoAIAAMAoFFQAAAAYhYIKAAAA\no1BQAQAAYBQKKgAAAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwCgUVAAAABiFggoAAACjUFAB\nAABgFAoqAAAAjEJBBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRUAAAAGIWCCgAAAKNQUAEAAGAU\nCioAAACMQkEFAACAUSioAAAAMAoFFQAAAEahoAIAAMAoFFQAAAAYhYIKAAAAo1BQAQAAYBQKKgAA\nAIxCQQUAAIBRKKgAAAAwCgUVAAAARqGgAgAAwChRF1TLsjR//nzt3r07vK2jo0OLFy9WYWGh5s2b\np/r6+ojH7Nq1S/Pnz1dBQYHKy8vV3t4e/coBAAAwIUVVUC3L0p133im/3x+xvbKyUunp6fL5fFqw\nYIGWLl2qzs5OSdKhQ4dUWVkpr9crn8+n5ORkVVZWxn4EAAAAmFBsF9S2tjb9x3/8hzo6OiK2NzQ0\nqL29XStXrtSMGTNUUVGhgoIC1dXVSZJqa2s1e/ZslZeXKzs7W6tWrdLBgwcj3oEFAAAAbBfUv/71\nr5ozZ46eeeYZhUKh8Pbm5mbl5eXJ7XaHtxUVFWnv3r3h8ZKSkvBYYmKicnNztWfPnljWDwAAgAkm\nwe4DvvKVr5xyeyAQUHp6esS2lJQUdXV1SZK6u7uHjaempobHAQAAACmKgjqSYDAol8sVsc3lcsmy\nLElSf3//acfPVHz8qd/0nRQ3SRqyNRUMER8fp4QEM24oMdL5Fct8phzbaDqZm9P5TXTkZh+ZRYfc\n7COz6DiVl2MF1e1269ixYxHbLMtSYmJiePyjZdSyLHk8HlvP4/EkjfD8n5AGbE0FQ3g8SUpOnjzW\ny5A08vkVy3ymHNvZ4HR+5wpys4/MokNu9pHZ2HCsoGZkZAz7qf6enh6lpaWFxwOBwLDxnJwcW8/T\n2xvU4ODwt0pPnPjQ5ophit7eoI4cOT7Wy5D0r7U4PZ8pxzaa4uPj5PEkjfj6xKmRm31kFh1ys4/M\nonMyt1g5VlDz8/NVU1Mjy7LCH+U3NjaquLg4PN7U1BTePxgMqrW1VcuWLbP1PIODQxoYGH6ihIZC\np9gb48FI39Ox4PRfQiYd29lwrh2vU8jNPjKLDrnZR2Zjw7ELK0pLSzVt2jRVVVXJ7/dr8+bNamlp\n0cKFCyVJXq9XTU1Nqqmpkd/v1/Lly5WVlaXS0lKnlgAAAIAJIKaCOmnSpP+bKC5OGzZsUCAQkNfr\n1bZt27R+/XpNnTpVkpSZmanq6mr5fD4tWrRIfX19WrduXWyrBwAAwIQT00f8b7zxRsTX06dP19at\nW0fcv6ysTNu3b4/lKQEAADDBce8EAAAAGIWCCgAAAKNQUAEAAGAUCioAAACMQkEFAACAUSioAAAA\nMAoFFQAAAEZx7FedAtEYGhzQm2/ud2SuvLzZ4V+zCwAAxi8KKsbU8aOHtOX5v2vKK/+MaZ6+w+/r\noTulwsIih1YGAADGCgUVY25KSpYumHrpWC8DNliWpX37WsJfx8fHyeNJUm9vUIODQ7bm4p1vAMBH\nUVAB2LZvX4u+t/a3mpKSFdM8vPMNADgVCiqAqPDONwBgtFBQAcN99OP0WPBxOgBgPKCgAobj43QA\nwLmGggqMA3ycDgA4l3CjfgAAABiFd1AxITh1w3+nfmkAAACIHgUVE4JTN/zveme3MmaUOLQqAAAQ\nDQoqJgwnrtPsO9zu0GoAAEC0uAYVAAAARqGgAgAAwCgUVAAAABiFggoAAACjUFABAABgFAoqAAAA\njEJBBQAAgFEoqAAAADAKBRUAAABG4TdJAaNkaHBAb765P+Z5nJgDAIDxhIIKjJLjRw9py/N/15RX\n/hnTPF3v7FbGjBKHVgUAgPkoqMAompKSpQumXhrTHH2H2x1aDQAA4wPXoAIAAMAovIMKnCOcuiZW\n4rpYAMDooqAC5winromVuC4WADC6KKjAOcSJa2IlrosFAIyus3oNqmVZWrFihUpKSlRWVqYnnnji\nbD49AAAAxoGz+g7qgw8+qNbWVm3dulUdHR36/ve/r8zMTN1www1ncxkAAAAw2FkrqMFgUHV1ddqy\nZYtmzpypmTNn6tvf/raeeuopCiqAmFmWpX37WhyZKy9vtlwulyNzAQDsO2sFdf/+/RocHFRBQUF4\nW1FRkTZt2nS2lgBgAtu3r0XfW/tbTUnJimmevsPv66E7pcLCIodWhtOx+w+L+Pg4eTxJ6u0NanBw\nKGKMf1gAE8dZK6iBQEAXXHCBEhL+7ylTUlJ04sQJHTlyRMnJyWdrKQAmKKd+CAyn5+S71W++uV9b\nnm/lHxYAIpzVj/g/+i/bk19blnXG88THn/rnuibFTZKGTjlkW9/h92Oe44NjnZJCsS/GwblMm8fJ\nuUybx8m5TJvHybn6Dr+vt9+eMuLr2o63337Tkdeuk2v6OHFxk3T++Yn65z/7NTTkzPfmbNi//w2t\nfXK7zvOkxzzXPw69qbSsyx1Y1b/OgbPxfRuPxuu5NpacyOyKK869fzA59RqcFAqFzsqZun37dv34\nxz/Wzp07w9va2to0b948vfrqq/J4PGdjGQAAADDcWfunZkZGho4ePaqhof97m7Onp0eJiYmUUwAA\nAISdtYKak5OjhIQE7d27N7zttdde06xZs87WEgAAADAOnLWCmpiYqJtuukn33HOPWlpatGPHDj3x\nxBP65je/ebaWAAAAgHHgrF2DKkn9/f2677779Kc//UlTpkzRt7/9bX39618/W08PAACAceCsFlQA\nAADg43A/DgAAABiFggoAAACjUFABAABgFAoqAAAAjEJBBQAAgFHGRUG1LEsrVqxQSUmJysrK9MQT\nT4z1koxlWZbmz5+v3bt3h7d1dHRo8eLFKiws1Lx581RfXz+GKzRLV1eXbr/9dn32s5/VNddco9Wr\nV8uyLEnkNpL3339ft9xyiwoLCzV37lxt2bIlPEZmH6+iokLLly8Pf01mI9uxY4dmzpypnJyc8H/v\nuOMOSeR2OpZl6b777lNpaamuuuoqPfLII+Exchvu2WefHXaezZw5U7m5uZKk9vZ2MhtBZ2envvOd\n76ioqEjXXnutfvnLX4bHYj3XxkVBffDBB9Xa2qqtW7fqnnvu0bp16/Q///M/Y70s41iWpTvvvFN+\nvz9ie2VlpdLT0+Xz+bRgwQItXbpUnZ2dY7RKs9x+++06ceKEnn76aa1du1YvvPCCHnvsMUnSkiVL\nyO0jQqGQKioqlJqaqueee0733nuvNm7cqOeff14SmX2c559/Xi+99FLENl6fI/P7/Zo7d67q6+tV\nX1+vnTt36oEHHpDEuXY6P/7xj9XQ0KBf/OIXevjhh1VbW6va2lpJ5HYqX/ziF8PnV319vV544QVd\ndNFF4V8kxGt0ZHfccYcmT56sZ599VitWrNCjjz6qHTt2SHLgXAsZ7oMPPghdfvnlod27d4e3bdiw\nIfT1r399DFdlHr/fH7rppptCN910U2jmzJmhv/71r6FQKBTatWtXqLCwMNTf3x/et7y8PFRdXT1W\nSzVGW1tbaObMmaHDhw+Ht/3hD38IXX311aGGhgZyO4Xu7u7Qf/3Xf4WOHz8e3rZ06dLQfffdR2Yf\n4+jRo6FrrrkmtGjRolBVVVUoFOL1+XHuuuuu0Nq1a4dtJ7eRHT16NJSXlxfx/8zNmzeHVqxYwWv0\nDP385z8P3XDDDSHLsjjXTuPYsWOhyy67LPT222+Hty1btix0//33O3KuGf8O6v79+zU4OKiCgoLw\ntqKiIjU3N4/hqszz17/+VXPmzNEzzzyj0L/97oXm5mbl5eXJ7XaHtxUVFWnv3r1jsUyjpKWl6fHH\nH9eFF14Ysb2vr0+vv/46uZ1CWlqa1q5dq/POO0+S1NjYqNdee02lpaVk9jEefPBB3XTTTcrOzg5v\n4/V5em1tbbr44ouHbSe3kTU2NmrKlCkqLi4Ob7v11lv1wAMP8Bo9A8eOHdPjjz+uu+66S5/4xCc4\n104jMTFRSUlJ8vl8GhgY0DvvvKOmpibl5OQ4cq4ZX1ADgYAuuOACJSQkhLelpKToxIkTOnLkyBiu\nzCxf+cpX9P3vfz/iZJD+lV96enrEtpSUFHV1dZ3N5RlpypQpuvLKK8Nfh0IhPfXUU5ozZw65nYG5\nc+fq5ptvVkFBgW644QYyO42GhgY1NjaqsrIyYjuZnd67776rl19+WZ///Od1/fXX66c//ak+/PBD\ncjuN9vZ2ZWZm6ne/+52+8IUv6LrrrtOGDRsUCoXI7Qw8/fTTysjI0PXXXy+J1+jpuFwu3X333fr1\nr3+t/Px83Xjjjbr66qvl9XodyS3h43cZW8FgUC6XK2Lbya9P/jALRjZSfmQ33EMPPaQ33nhDdXV1\neuKJJ8jtY1RXV6unp0f33nuvfvKTn3CujcCyLN1777265557huVDZiP7+9//rv7+frndbj322GPq\n6OjQAw88oP7+fnI7jQ8++EDvvfeeamtrtXr1agUCAd19991KSkoitzNQV1enioqK8NdkdnptbW2a\nO3eubrnlFr311lu6//77NWfOHEdyM76gut3uYQd08uukpKSxWNK44na7dezYsYhtlmUpMTFxjFZk\npjVr1mjr1q169NFHdckll5DbGcjLy5MkVVVV6a677tLChQvV29sbsQ+Z/avIz5o1S5/73OeGjXGe\njexTn/qUXn31VXk8HknSzJkzNTQ0pO9+97v60pe+xLk2gvj4eB0/flxr167V1KlTJUkHDx7U008/\nrauuukpHjx6N2J/c/k9zc7O6urp04403hrfxGh1ZQ0OD6urq9NJLL8nlcik3N1ednZ3auHGj5syZ\nE/O5ZvxH/BkZGTp69KiGhobC23p6epSYmBj+iwsjy8jIUCAQiNjW09OjtLS0MVqRee6//3798pe/\n1Jo1a3TddddJIreRHD58OPwTmiddcskl+vDDD5WWlkZmp/DHP/5Rf/7zn1VYWKjCwkJt27ZN27Zt\n0xVXXKGpU6eS2Wl89O/47OxsnThxQqmpqeQ2gvT0dLnd7nA5laSLL75YXV1d/L32MXbu3KmSkhJN\nmTIlvI3MRrZv3z595jOfiXinNCcnR4cOHXIkN+MLak5OjhISEiIurH3ttdc0a9asMVzV+JGfn6/W\n1taId6EbGxsjfujsXLZu3To988wzeuSRR/SFL3whvJ3cTq2jo0PLli1Td3d3eFtLS4tSUlJUVFSk\nffv2kdlHPPXUU9q2bZt+//vf6/e//73mzp2ruXPn6rnnntPll1/OeTaCnTt36rOf/axOnDgR3tba\n2qrk5GQVFxdzro0gPz9fJ06c0IEDB8Lb2tralJmZqfz8fHI7jebmZl1xxRUR2/h/wcjS09N14MAB\nDQwMhLe98847+vSnP+3IuWZ8QU1MTNRNN92ke+65Ry0tLdqxY4eeeOKJ8P3JcHqlpaWaNm2aqqqq\n5Pf7tXnzZrW0tGjhwoVjvbQx19bWpo0bN6qiokKFhYXq6ekJ/yG3U5s9e7ZmzZqlFStWqK2tTS++\n+KIefvhh3XbbbSopKSGzU5g2bZqmT58e/jN58mRNnjxZ06dP5zw7jcLCQiUlJekHP/iB3n33Xb34\n4otas2aNbr31Vs6107j44ot1zTXXqKqqSvv379fLL7+smpoaffWrXyW3j/HWW29F3GVD4v+hpzN3\n7lwlJCTohz/8od577z395S9/0aZNm/SNb3zDmXPNuTtijZ5gMBiqqqoKFRYWhq6++urQr371q7Fe\nktH+/T6ooVAo9P7774duvvnm0OWXXx6aN29eqKGhYQxXZ45NmzaFZs6cGfHnsssuC82cOTMUCoVC\nBw4cILdT6O7uDi1btixUXFwcKisrC23atCk8xrn28aqqqsL3QQ2FyOx0/H5/6Fvf+lboiiuuCJWV\nlYXWr18fHiO3kfX19YW+//3vh6644orQlVdeGdqwYUN4jNxGlp+fH9q5c+ew7WQ2spOv0eLi4tAN\nN9wQ0c9izW1SKPRvN80EAAAAxpjxH/EDAADg3EJBBQAAgFEoqAAAADAKBRUAAABGoaACAADAKBRU\nAAAAGIWCCgAAAKNQUAEAAGAUCioAAACMQkEFAACAUSioAAAAMMr/B1QlyvF9ZBmlAAAAAElFTkSu\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcbaad30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data['metro_time'].hist(bins=30)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Посмотрим квартиры, которые находятся от метро в 40 минутах"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "769     149796933\n",
       "1508    148356553\n",
       "2110    150396925\n",
       "3084    149626095\n",
       "3459    150818815\n",
       "4014    148445047\n",
       "4348    149544711\n",
       "4393    148848792\n",
       "Name: link, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.ix[data['metro_time']>40,'link']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Все норм, это не выбросы"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "11.065747613997878"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['metro_time'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data['metro_time_non'] = 0\n",
    "data.loc[data['metro_time'].isnull() ,'metro_time_non'] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data['metro_time'] = data['metro_time'].fillna(data['metro_time'].mean()) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xcf2e160>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArIAAAHcCAYAAAAqb0k3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xce25f60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data['rooms'].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       150090567\n",
       "422      32637436\n",
       "1336    149212457\n",
       "3298      2914190\n",
       "Name: link, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.ix[data['rooms'].isnull(),'link']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data['rooms'] = data['rooms'].fillna(6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xd0ac780>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAqgAAAHcCAYAAAAa41gWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKKgAAAAwCgUVAAAARvn/AIswKW3DrLSMAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd2085f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data['total_floors'].hist(bins=50) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data['total_floors_non'] = 0\n",
    "data.loc[data['total_floors'].isnull() ,'total_floors_non'] = 1\n",
    "data['total_floors'] = data['total_floors'].fillna(data['total_floors'].mean()) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data['living_space_non'] = 0\n",
    "data.loc[data['living_space'].isnull() ,'living_space_non'] = 1\n",
    "data['living_space'] = data['living_space'].fillna(data['living_space'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data['cooking_space_non'] = 0\n",
    "data.loc[data['cooking_space'].isnull() ,'cooking_space_non'] = 1\n",
    "data['cooking_space'] = data['cooking_space'].fillna(data['cooking_space'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "District  12\n",
      "Price  1801\n",
      "Tel  2\n",
      "combined_bs  5\n",
      "cooking_space  188\n",
      "dist  2733\n",
      "floor  41\n",
      "house_type  7\n",
      "lat  2677\n",
      "link  4397\n",
      "living_space  473\n",
      "lon  2663\n",
      "metro_station  175\n",
      "metro_time  37\n",
      "new  2\n",
      "rooms  6\n",
      "separate_bs  5\n",
      "total_floors  56\n",
      "total_space  666\n",
      "walk  2\n",
      "metro_time_non  2\n",
      "total_floors_non  2\n",
      "living_space_non  2\n",
      "cooking_space_non  2\n"
     ]
    }
   ],
   "source": [
    "def num_cat(data):\n",
    "    for name in data.columns:\n",
    "        print(str(name) + '  '+ str(data[name].nunique()))\n",
    "#Количество уникальных значений\n",
    "num_cat(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0L"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull().sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>District</th>\n",
       "      <th>Price</th>\n",
       "      <th>Tel</th>\n",
       "      <th>combined_bs</th>\n",
       "      <th>cooking_space</th>\n",
       "      <th>dist</th>\n",
       "      <th>floor</th>\n",
       "      <th>house_type</th>\n",
       "      <th>lat</th>\n",
       "      <th>link</th>\n",
       "      <th>...</th>\n",
       "      <th>new</th>\n",
       "      <th>rooms</th>\n",
       "      <th>separate_bs</th>\n",
       "      <th>total_floors</th>\n",
       "      <th>total_space</th>\n",
       "      <th>walk</th>\n",
       "      <th>metro_time_non</th>\n",
       "      <th>total_floors_non</th>\n",
       "      <th>living_space_non</th>\n",
       "      <th>cooking_space_non</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>528391080</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>2.720885</td>\n",
       "      <td>8</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.749603</td>\n",
       "      <td>150090567</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.00000</td>\n",
       "      <td>399.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>311298500</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>3.512113</td>\n",
       "      <td>5</td>\n",
       "      <td>монолитный</td>\n",
       "      <td>55.757526</td>\n",
       "      <td>150453732</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.00000</td>\n",
       "      <td>200.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>14000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>8.900000</td>\n",
       "      <td>6.482659</td>\n",
       "      <td>5</td>\n",
       "      <td>кирпичный</td>\n",
       "      <td>55.743831</td>\n",
       "      <td>150613273</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.00000</td>\n",
       "      <td>75.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>4403057</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>5.864250</td>\n",
       "      <td>3</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.775753</td>\n",
       "      <td>149031353</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.96535</td>\n",
       "      <td>22.7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>4814581</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>5.903571</td>\n",
       "      <td>2</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.776294</td>\n",
       "      <td>148567438</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.00000</td>\n",
       "      <td>17.3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   District      Price  Tel  combined_bs  cooking_space      dist  floor  \\\n",
       "0         0  528391080    0            0       9.011864  2.720885      8   \n",
       "1         0  311298500    1            3       9.011864  3.512113      5   \n",
       "2         0   14000000    1            0       8.900000  6.482659      5   \n",
       "3         0    4403057    0            1       9.011864  5.864250      3   \n",
       "4         0    4814581    0            1       9.011864  5.903571      2   \n",
       "\n",
       "   house_type        lat       link        ...          new  rooms  \\\n",
       "0        Else  55.749603  150090567        ...            0    6.0   \n",
       "1  монолитный  55.757526  150453732        ...            0    5.0   \n",
       "2   кирпичный  55.743831  150613273        ...            0    3.0   \n",
       "3        Else  55.775753  149031353        ...            1    1.0   \n",
       "4        Else  55.776294  148567438        ...            0    1.0   \n",
       "\n",
       "  separate_bs  total_floors  total_space  walk  metro_time_non  \\\n",
       "0           0       9.00000        399.0   1.0               0   \n",
       "1           0       7.00000        200.0   1.0               0   \n",
       "2           1       5.00000         75.5   0.0               0   \n",
       "3           0      13.96535         22.7   1.0               0   \n",
       "4           0       5.00000         17.3   1.0               0   \n",
       "\n",
       "   total_floors_non  living_space_non  cooking_space_non  \n",
       "0                 0                 1                  1  \n",
       "1                 0                 0                  1  \n",
       "2                 0                 0                  0  \n",
       "3                 1                 0                  1  \n",
       "4                 0                 1                  1  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xcb920b8>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArQAAAHcCAYAAAAncTlwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqLWD2AEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd3771d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(data['Price']**0.25).hist(bins=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>District</th>\n",
       "      <th>Price</th>\n",
       "      <th>Tel</th>\n",
       "      <th>combined_bs</th>\n",
       "      <th>cooking_space</th>\n",
       "      <th>dist</th>\n",
       "      <th>floor</th>\n",
       "      <th>lat</th>\n",
       "      <th>link</th>\n",
       "      <th>living_space</th>\n",
       "      <th>lon</th>\n",
       "      <th>metro_time</th>\n",
       "      <th>new</th>\n",
       "      <th>rooms</th>\n",
       "      <th>separate_bs</th>\n",
       "      <th>total_floors</th>\n",
       "      <th>total_space</th>\n",
       "      <th>walk</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4397.000000</td>\n",
       "      <td>4.397000e+03</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>3616.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>4.397000e+03</td>\n",
       "      <td>3953.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>3772.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>4393.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>4329.000000</td>\n",
       "      <td>4397.000000</td>\n",
       "      <td>2357.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>4.505572</td>\n",
       "      <td>9.910454e+06</td>\n",
       "      <td>0.344326</td>\n",
       "      <td>0.571071</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>17.160947</td>\n",
       "      <td>7.084603</td>\n",
       "      <td>55.727852</td>\n",
       "      <td>1.419186e+08</td>\n",
       "      <td>28.411485</td>\n",
       "      <td>37.565066</td>\n",
       "      <td>11.065748</td>\n",
       "      <td>0.292700</td>\n",
       "      <td>1.582518</td>\n",
       "      <td>0.389356</td>\n",
       "      <td>13.965350</td>\n",
       "      <td>47.985490</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>3.151581</td>\n",
       "      <td>1.900881e+07</td>\n",
       "      <td>0.475202</td>\n",
       "      <td>0.558465</td>\n",
       "      <td>4.194752</td>\n",
       "      <td>10.517870</td>\n",
       "      <td>5.920011</td>\n",
       "      <td>0.107955</td>\n",
       "      <td>2.785309e+07</td>\n",
       "      <td>17.604585</td>\n",
       "      <td>0.147533</td>\n",
       "      <td>6.342385</td>\n",
       "      <td>0.455053</td>\n",
       "      <td>0.768864</td>\n",
       "      <td>0.535675</td>\n",
       "      <td>8.352073</td>\n",
       "      <td>25.924701</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.380000e+06</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.151341</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>55.214892</td>\n",
       "      <td>1.861247e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>36.848282</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>5.290000e+06</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>9.691749</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>55.651029</td>\n",
       "      <td>1.488710e+08</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>37.475528</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>6.603800e+06</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>8.250000</td>\n",
       "      <td>16.057230</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>55.738262</td>\n",
       "      <td>1.498933e+08</td>\n",
       "      <td>22.500000</td>\n",
       "      <td>37.561293</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>41.800000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.000000</td>\n",
       "      <td>9.000000e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>22.285357</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>55.811295</td>\n",
       "      <td>1.504884e+08</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>37.656092</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>11.000000</td>\n",
       "      <td>5.283911e+08</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>55.500000</td>\n",
       "      <td>91.077575</td>\n",
       "      <td>46.000000</td>\n",
       "      <td>56.002813</td>\n",
       "      <td>1.508224e+08</td>\n",
       "      <td>300.000000</td>\n",
       "      <td>37.940921</td>\n",
       "      <td>72.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>86.000000</td>\n",
       "      <td>505.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          District         Price          Tel  combined_bs  cooking_space  \\\n",
       "count  4397.000000  4.397000e+03  4397.000000  4397.000000    3616.000000   \n",
       "mean      4.505572  9.910454e+06     0.344326     0.571071       9.011864   \n",
       "std       3.151581  1.900881e+07     0.475202     0.558465       4.194752   \n",
       "min       0.000000  1.380000e+06     0.000000     0.000000       1.000000   \n",
       "25%       2.000000  5.290000e+06     0.000000     0.000000       6.000000   \n",
       "50%       5.000000  6.603800e+06     0.000000     1.000000       8.250000   \n",
       "75%       7.000000  9.000000e+06     1.000000     1.000000      10.000000   \n",
       "max      11.000000  5.283911e+08     1.000000     4.000000      55.500000   \n",
       "\n",
       "              dist        floor          lat          link  living_space  \\\n",
       "count  4397.000000  4397.000000  4397.000000  4.397000e+03   3953.000000   \n",
       "mean     17.160947     7.084603    55.727852  1.419186e+08     28.411485   \n",
       "std      10.517870     5.920011     0.107955  2.785309e+07     17.604585   \n",
       "min       1.151341     1.000000    55.214892  1.861247e+06      1.000000   \n",
       "25%       9.691749     3.000000    55.651029  1.488710e+08     19.000000   \n",
       "50%      16.057230     5.000000    55.738262  1.498933e+08     22.500000   \n",
       "75%      22.285357    10.000000    55.811295  1.504884e+08     32.000000   \n",
       "max      91.077575    46.000000    56.002813  1.508224e+08    300.000000   \n",
       "\n",
       "               lon   metro_time          new        rooms  separate_bs  \\\n",
       "count  4397.000000  3772.000000  4397.000000  4393.000000  4397.000000   \n",
       "mean     37.565066    11.065748     0.292700     1.582518     0.389356   \n",
       "std       0.147533     6.342385     0.455053     0.768864     0.535675   \n",
       "min      36.848282     1.000000     0.000000     1.000000     0.000000   \n",
       "25%      37.475528     6.000000     0.000000     1.000000     0.000000   \n",
       "50%      37.561293    10.000000     0.000000     1.000000     0.000000   \n",
       "75%      37.656092    15.000000     1.000000     2.000000     1.000000   \n",
       "max      37.940921    72.000000     1.000000     5.000000     4.000000   \n",
       "\n",
       "       total_floors  total_space    walk  \n",
       "count   4329.000000  4397.000000  2357.0  \n",
       "mean      13.965350    47.985490     1.0  \n",
       "std        8.352073    25.924701     0.0  \n",
       "min        1.000000    13.000000     1.0  \n",
       "25%        8.000000    36.000000     1.0  \n",
       "50%       12.000000    41.800000     1.0  \n",
       "75%       17.000000    53.000000     1.0  \n",
       "max       86.000000   505.000000     1.0  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xe6162b0>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArsAAAIrCAYAAADvBVHBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xe69e0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "corr = abs(data.corr())\n",
    "sns.heatmap(corr, \n",
    "            xticklabels=corr.columns.values,\n",
    "            yticklabels=corr.columns.values, annot=False, cmap='YlGnBu')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Сохраним обработанные данные"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data.to_csv('Data_cian_full_finished.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Загрузим обработанные данные"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data = pd.read_csv('Data_cian_full_finished.csv',index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>District</th>\n",
       "      <th>Price</th>\n",
       "      <th>Tel</th>\n",
       "      <th>combined_bs</th>\n",
       "      <th>cooking_space</th>\n",
       "      <th>dist</th>\n",
       "      <th>floor</th>\n",
       "      <th>house_type</th>\n",
       "      <th>lat</th>\n",
       "      <th>link</th>\n",
       "      <th>...</th>\n",
       "      <th>new</th>\n",
       "      <th>rooms</th>\n",
       "      <th>separate_bs</th>\n",
       "      <th>total_floors</th>\n",
       "      <th>total_space</th>\n",
       "      <th>walk</th>\n",
       "      <th>metro_time_non</th>\n",
       "      <th>total_floors_non</th>\n",
       "      <th>living_space_non</th>\n",
       "      <th>cooking_space_non</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>528391080</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>2.720885</td>\n",
       "      <td>8</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.749603</td>\n",
       "      <td>150090567</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.00000</td>\n",
       "      <td>399.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>311298500</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>3.512113</td>\n",
       "      <td>5</td>\n",
       "      <td>монолитный</td>\n",
       "      <td>55.757526</td>\n",
       "      <td>150453732</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.00000</td>\n",
       "      <td>200.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>14000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>8.900000</td>\n",
       "      <td>6.482659</td>\n",
       "      <td>5</td>\n",
       "      <td>кирпичный</td>\n",
       "      <td>55.743831</td>\n",
       "      <td>150613273</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.00000</td>\n",
       "      <td>75.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>4403057</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>5.864250</td>\n",
       "      <td>3</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.775753</td>\n",
       "      <td>149031353</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.96535</td>\n",
       "      <td>22.7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>4814581</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>9.011864</td>\n",
       "      <td>5.903571</td>\n",
       "      <td>2</td>\n",
       "      <td>Else</td>\n",
       "      <td>55.776294</td>\n",
       "      <td>148567438</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.00000</td>\n",
       "      <td>17.3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   District      Price  Tel  combined_bs  cooking_space      dist  floor  \\\n",
       "0         0  528391080    0            0       9.011864  2.720885      8   \n",
       "1         0  311298500    1            3       9.011864  3.512113      5   \n",
       "2         0   14000000    1            0       8.900000  6.482659      5   \n",
       "3         0    4403057    0            1       9.011864  5.864250      3   \n",
       "4         0    4814581    0            1       9.011864  5.903571      2   \n",
       "\n",
       "   house_type        lat       link        ...          new  rooms  \\\n",
       "0        Else  55.749603  150090567        ...            0    6.0   \n",
       "1  монолитный  55.757526  150453732        ...            0    5.0   \n",
       "2   кирпичный  55.743831  150613273        ...            0    3.0   \n",
       "3        Else  55.775753  149031353        ...            1    1.0   \n",
       "4        Else  55.776294  148567438        ...            0    1.0   \n",
       "\n",
       "  separate_bs  total_floors  total_space  walk  metro_time_non  \\\n",
       "0           0       9.00000        399.0   1.0               0   \n",
       "1           0       7.00000        200.0   1.0               0   \n",
       "2           1       5.00000         75.5   0.0               0   \n",
       "3           0      13.96535         22.7   1.0               0   \n",
       "4           0       5.00000         17.3   1.0               0   \n",
       "\n",
       "   total_floors_non  living_space_non  cooking_space_non  \n",
       "0                 0                 1                  1  \n",
       "1                 0                 0                  1  \n",
       "2                 0                 0                  0  \n",
       "3                 1                 0                  1  \n",
       "4                 0                 1                  1  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "District               int64\n",
       "Price                  int64\n",
       "Tel                    int64\n",
       "combined_bs            int64\n",
       "cooking_space        float64\n",
       "dist                 float64\n",
       "floor                  int64\n",
       "house_type            object\n",
       "lat                  float64\n",
       "link                   int64\n",
       "living_space         float64\n",
       "lon                  float64\n",
       "metro_station         object\n",
       "metro_time           float64\n",
       "new                    int64\n",
       "rooms                float64\n",
       "separate_bs            int64\n",
       "total_floors         float64\n",
       "total_space          float64\n",
       "walk                 float64\n",
       "metro_time_non         int64\n",
       "total_floors_non       int64\n",
       "living_space_non       int64\n",
       "cooking_space_non      int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Линейные Модели"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Ridge, LinearRegression, Lasso\n",
    "from sklearn.grid_search import GridSearchCV\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.metrics import make_scorer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>блочный</th>\n",
       "      <th>кирпично-монолитный</th>\n",
       "      <th>кирпичный</th>\n",
       "      <th>монолитный</th>\n",
       "      <th>панельныйдом</th>\n",
       "      <th>сталинский</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   блочный  кирпично-монолитный  кирпичный  монолитный  панельныйдом  \\\n",
       "0        0                    0          0           0             0   \n",
       "1        0                    0          0           1             0   \n",
       "2        0                    0          1           0             0   \n",
       "3        0                    0          0           0             0   \n",
       "4        0                    0          0           0             0   \n",
       "\n",
       "   сталинский  \n",
       "0           0  \n",
       "1           0  \n",
       "2           0  \n",
       "3           0  \n",
       "4           0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaled_data = []\n",
    "\n",
    "f_cat = ['metro_station','house_type']\n",
    "for f in f_cat:\n",
    "    dummy = pd.get_dummies(data[f].astype('category'), drop_first=True)\n",
    "\n",
    "dummy.head()    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "dummy.columns = [ u'блочный', u'кирпично-монолитный', u'кирпичный', u'монолитный', u'панельныйдом', u'сталинский']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y = 1.0*data['Price']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>District</th>\n",
       "      <th>Tel</th>\n",
       "      <th>combined_bs</th>\n",
       "      <th>cooking_space</th>\n",
       "      <th>dist</th>\n",
       "      <th>floor</th>\n",
       "      <th>house_type</th>\n",
       "      <th>lat</th>\n",
       "      <th>living_space</th>\n",
       "      <th>lon</th>\n",
       "      <th>...</th>\n",
       "      <th>new</th>\n",
       "      <th>rooms</th>\n",
       "      <th>separate_bs</th>\n",
       "      <th>total_floors</th>\n",
       "      <th>total_space</th>\n",
       "      <th>walk</th>\n",
       "      <th>metro_time_non</th>\n",
       "      <th>total_floors_non</th>\n",
       "      <th>living_space_non</th>\n",
       "      <th>cooking_space_non</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>-0.724671</td>\n",
       "      <td>-1.022689</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.373064</td>\n",
       "      <td>0.154645</td>\n",
       "      <td>Else</td>\n",
       "      <td>0.201502</td>\n",
       "      <td>-2.189742e-12</td>\n",
       "      <td>0.222387</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.643293</td>\n",
       "      <td>5.659146</td>\n",
       "      <td>-0.726935</td>\n",
       "      <td>-5.992254e-01</td>\n",
       "      <td>13.541311</td>\n",
       "      <td>0.930326</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>2.983815</td>\n",
       "      <td>2.151735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>1.379937</td>\n",
       "      <td>4.349786</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.297828</td>\n",
       "      <td>-0.352168</td>\n",
       "      <td>монолитный</td>\n",
       "      <td>0.274902</td>\n",
       "      <td>8.183860e+00</td>\n",
       "      <td>0.173308</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.643293</td>\n",
       "      <td>4.376900</td>\n",
       "      <td>-0.726935</td>\n",
       "      <td>-8.405882e-01</td>\n",
       "      <td>5.864361</td>\n",
       "      <td>0.930326</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>-0.335141</td>\n",
       "      <td>2.151735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>1.379937</td>\n",
       "      <td>-1.022689</td>\n",
       "      <td>-2.941087e-02</td>\n",
       "      <td>-1.015367</td>\n",
       "      <td>-0.352168</td>\n",
       "      <td>кирпичный</td>\n",
       "      <td>0.148029</td>\n",
       "      <td>1.347426e+00</td>\n",
       "      <td>0.778009</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.643293</td>\n",
       "      <td>1.812408</td>\n",
       "      <td>1.140081</td>\n",
       "      <td>-1.081951e+00</td>\n",
       "      <td>1.061445</td>\n",
       "      <td>-1.074892</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>-0.335141</td>\n",
       "      <td>-0.464741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>-0.724671</td>\n",
       "      <td>0.768136</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.074170</td>\n",
       "      <td>-0.690044</td>\n",
       "      <td>Else</td>\n",
       "      <td>0.443760</td>\n",
       "      <td>-3.422103e-01</td>\n",
       "      <td>0.724476</td>\n",
       "      <td>...</td>\n",
       "      <td>1.554501</td>\n",
       "      <td>-0.752084</td>\n",
       "      <td>-0.726935</td>\n",
       "      <td>-5.936638e-12</td>\n",
       "      <td>-0.975454</td>\n",
       "      <td>0.930326</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>7.978832</td>\n",
       "      <td>-0.335141</td>\n",
       "      <td>2.151735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>-0.724671</td>\n",
       "      <td>0.768136</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.070431</td>\n",
       "      <td>-0.858982</td>\n",
       "      <td>Else</td>\n",
       "      <td>0.448772</td>\n",
       "      <td>-2.189742e-12</td>\n",
       "      <td>0.726001</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.643293</td>\n",
       "      <td>-0.752084</td>\n",
       "      <td>-0.726935</td>\n",
       "      <td>-1.081951e+00</td>\n",
       "      <td>-1.183774</td>\n",
       "      <td>0.930326</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>2.983815</td>\n",
       "      <td>2.151735</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   District       Tel  combined_bs  cooking_space      dist     floor  \\\n",
       "0 -1.429785 -0.724671    -1.022689  -6.692589e-13 -1.373064  0.154645   \n",
       "1 -1.429785  1.379937     4.349786  -6.692589e-13 -1.297828 -0.352168   \n",
       "2 -1.429785  1.379937    -1.022689  -2.941087e-02 -1.015367 -0.352168   \n",
       "3 -1.429785 -0.724671     0.768136  -6.692589e-13 -1.074170 -0.690044   \n",
       "4 -1.429785 -0.724671     0.768136  -6.692589e-13 -1.070431 -0.858982   \n",
       "\n",
       "   house_type       lat  living_space       lon        ...               new  \\\n",
       "0        Else  0.201502 -2.189742e-12  0.222387        ...         -0.643293   \n",
       "1  монолитный  0.274902  8.183860e+00  0.173308        ...         -0.643293   \n",
       "2   кирпичный  0.148029  1.347426e+00  0.778009        ...         -0.643293   \n",
       "3        Else  0.443760 -3.422103e-01  0.724476        ...          1.554501   \n",
       "4        Else  0.448772 -2.189742e-12  0.726001        ...         -0.643293   \n",
       "\n",
       "      rooms  separate_bs  total_floors  total_space      walk  metro_time_non  \\\n",
       "0  5.659146    -0.726935 -5.992254e-01    13.541311  0.930326       -0.407056   \n",
       "1  4.376900    -0.726935 -8.405882e-01     5.864361  0.930326       -0.407056   \n",
       "2  1.812408     1.140081 -1.081951e+00     1.061445 -1.074892       -0.407056   \n",
       "3 -0.752084    -0.726935 -5.936638e-12    -0.975454  0.930326       -0.407056   \n",
       "4 -0.752084    -0.726935 -1.081951e+00    -1.183774  0.930326       -0.407056   \n",
       "\n",
       "   total_floors_non  living_space_non  cooking_space_non  \n",
       "0         -0.125332          2.983815           2.151735  \n",
       "1         -0.125332         -0.335141           2.151735  \n",
       "2         -0.125332         -0.335141          -0.464741  \n",
       "3          7.978832         -0.335141           2.151735  \n",
       "4         -0.125332          2.983815           2.151735  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "scaler = StandardScaler()\n",
    "data.drop(['link','Price'], axis=1, inplace=True)\n",
    "\n",
    "f_num = [f for f in data.columns if f not in f_cat]\n",
    "data[f_num] = scaler.fit_transform(data[f_num])\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4397, 22)\n",
      "(4397, 6)\n"
     ]
    }
   ],
   "source": [
    "print data.shape\n",
    "print dummy.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data = pd.merge(data, dummy, left_index=True, right_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4397, 28)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>District</th>\n",
       "      <th>Tel</th>\n",
       "      <th>combined_bs</th>\n",
       "      <th>cooking_space</th>\n",
       "      <th>dist</th>\n",
       "      <th>floor</th>\n",
       "      <th>house_type</th>\n",
       "      <th>lat</th>\n",
       "      <th>living_space</th>\n",
       "      <th>lon</th>\n",
       "      <th>...</th>\n",
       "      <th>metro_time_non</th>\n",
       "      <th>total_floors_non</th>\n",
       "      <th>living_space_non</th>\n",
       "      <th>cooking_space_non</th>\n",
       "      <th>блочный</th>\n",
       "      <th>кирпично-монолитный</th>\n",
       "      <th>кирпичный</th>\n",
       "      <th>монолитный</th>\n",
       "      <th>панельныйдом</th>\n",
       "      <th>сталинский</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>-0.724671</td>\n",
       "      <td>-1.022689</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.373064</td>\n",
       "      <td>0.154645</td>\n",
       "      <td>Else</td>\n",
       "      <td>0.201502</td>\n",
       "      <td>-2.189742e-12</td>\n",
       "      <td>0.222387</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>2.983815</td>\n",
       "      <td>2.151735</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>1.379937</td>\n",
       "      <td>4.349786</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.297828</td>\n",
       "      <td>-0.352168</td>\n",
       "      <td>монолитный</td>\n",
       "      <td>0.274902</td>\n",
       "      <td>8.183860e+00</td>\n",
       "      <td>0.173308</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>-0.335141</td>\n",
       "      <td>2.151735</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>1.379937</td>\n",
       "      <td>-1.022689</td>\n",
       "      <td>-2.941087e-02</td>\n",
       "      <td>-1.015367</td>\n",
       "      <td>-0.352168</td>\n",
       "      <td>кирпичный</td>\n",
       "      <td>0.148029</td>\n",
       "      <td>1.347426e+00</td>\n",
       "      <td>0.778009</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>-0.335141</td>\n",
       "      <td>-0.464741</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>-0.724671</td>\n",
       "      <td>0.768136</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.074170</td>\n",
       "      <td>-0.690044</td>\n",
       "      <td>Else</td>\n",
       "      <td>0.443760</td>\n",
       "      <td>-3.422103e-01</td>\n",
       "      <td>0.724476</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>7.978832</td>\n",
       "      <td>-0.335141</td>\n",
       "      <td>2.151735</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.429785</td>\n",
       "      <td>-0.724671</td>\n",
       "      <td>0.768136</td>\n",
       "      <td>-6.692589e-13</td>\n",
       "      <td>-1.070431</td>\n",
       "      <td>-0.858982</td>\n",
       "      <td>Else</td>\n",
       "      <td>0.448772</td>\n",
       "      <td>-2.189742e-12</td>\n",
       "      <td>0.726001</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.407056</td>\n",
       "      <td>-0.125332</td>\n",
       "      <td>2.983815</td>\n",
       "      <td>2.151735</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   District       Tel  combined_bs  cooking_space      dist     floor  \\\n",
       "0 -1.429785 -0.724671    -1.022689  -6.692589e-13 -1.373064  0.154645   \n",
       "1 -1.429785  1.379937     4.349786  -6.692589e-13 -1.297828 -0.352168   \n",
       "2 -1.429785  1.379937    -1.022689  -2.941087e-02 -1.015367 -0.352168   \n",
       "3 -1.429785 -0.724671     0.768136  -6.692589e-13 -1.074170 -0.690044   \n",
       "4 -1.429785 -0.724671     0.768136  -6.692589e-13 -1.070431 -0.858982   \n",
       "\n",
       "   house_type       lat  living_space       lon     ...     metro_time_non  \\\n",
       "0        Else  0.201502 -2.189742e-12  0.222387     ...          -0.407056   \n",
       "1  монолитный  0.274902  8.183860e+00  0.173308     ...          -0.407056   \n",
       "2   кирпичный  0.148029  1.347426e+00  0.778009     ...          -0.407056   \n",
       "3        Else  0.443760 -3.422103e-01  0.724476     ...          -0.407056   \n",
       "4        Else  0.448772 -2.189742e-12  0.726001     ...          -0.407056   \n",
       "\n",
       "   total_floors_non  living_space_non  cooking_space_non  блочный  \\\n",
       "0         -0.125332          2.983815           2.151735        0   \n",
       "1         -0.125332         -0.335141           2.151735        0   \n",
       "2         -0.125332         -0.335141          -0.464741        0   \n",
       "3          7.978832         -0.335141           2.151735        0   \n",
       "4         -0.125332          2.983815           2.151735        0   \n",
       "\n",
       "   кирпично-монолитный  кирпичный  монолитный  панельныйдом  сталинский  \n",
       "0                    0          0           0             0           0  \n",
       "1                    0          0           1             0           0  \n",
       "2                    0          1           0             0           0  \n",
       "3                    0          0           0             0           0  \n",
       "4                    0          0           0             0           0  \n",
       "\n",
       "[5 rows x 28 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print data.shape\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data.drop(f_cat, axis=1, inplace=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "clf = LinearRegression()\n",
    "GS = GridSearchCV(clf, cv=5,param_grid={'fit_intercept':[True,False]}, verbose=0,scoring=make_scorer(score_func=mean_absolute_error, greater_is_better=False))\n",
    "                   #param_grid={\"C\": [1e0, 1e1, 1e2, 1e3],\n",
    "                    #           \"gamma\": np.logspace(-2, 2, 5)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False),\n",
       "       fit_params={}, iid=True, n_jobs=1,\n",
       "       param_grid={'fit_intercept': [True, False]},\n",
       "       pre_dispatch='2*n_jobs', refit=True,\n",
       "       scoring=make_scorer(mean_absolute_error, greater_is_better=False),\n",
       "       verbose=0)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.fit(data, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[mean: -5649984.32244, std: 1773474.39683, params: {'fit_intercept': True},\n",
       " mean: -8227444.48453, std: 2558452.70650, params: {'fit_intercept': False}]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.grid_scores_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2376.969567\n"
     ]
    }
   ],
   "source": [
    "score = np.sqrt(GS.best_score_*(-1.0))\n",
    "print score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf = Ridge()\n",
    "\n",
    "params={\"alpha\": np.logspace(-5, 5, 50)}\n",
    "            \n",
    "GS = GridSearchCV(clf, cv=5,param_grid=params, \\\n",
    "                  verbose=0,scoring=make_scorer(score_func=mean_absolute_error, greater_is_better=False),n_jobs=2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=Ridge(alpha=1.0, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "   normalize=False, random_state=None, solver='auto', tol=0.001),\n",
       "       fit_params={}, iid=True, n_jobs=2,\n",
       "       param_grid={'alpha': array([  1.00000e-05,   1.59986e-05,   2.55955e-05,   4.09492e-05,\n",
       "         6.55129e-05,   1.04811e-04,   1.67683e-04,   2.68270e-04,\n",
       "         4.29193e-04,   6.86649e-04,   1.09854e-03,   1.75751e-03,\n",
       "         2.81177e-03,   4.49843e-03,   7.19686e-03,   1.15140e-02,\n",
       "         1....    9.54095e+03,   1.52642e+04,   2.44205e+04,   3.90694e+04,\n",
       "         6.25055e+04,   1.00000e+05])},\n",
       "       pre_dispatch='2*n_jobs', refit=True,\n",
       "       scoring=make_scorer(mean_absolute_error, greater_is_better=False),\n",
       "       verbose=0)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.fit(data, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'alpha': 9540.9547634999635}"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-4579352.472349102"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "score = np.sqrt(GS.best_score_*(-1.0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2139.9421656552081"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "clf = RandomForestRegressor(n_estimators=100, random_state=9)\n",
    "\n",
    "params={\"max_features\": [1.0,0.9,0.8,0.7,0.6],\n",
    "       \"min_samples_leaf\": [1,2,3,4,5,10,20,30,40,50]}\n",
    "            \n",
    "GS = GridSearchCV(clf, cv=5,param_grid=params, \\\n",
    "                  verbose=0,scoring=make_scorer(score_func=mean_absolute_error, greater_is_better=False),n_jobs=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "G_fit = GS.fit(data, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_features': 0.6, 'min_samples_leaf': 3}"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1751.7447247362022"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sqrt(GS.best_score_*(-1.0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "clf = RandomForestRegressor(n_estimators=1000, random_state=9)\n",
    "\n",
    "params={\"max_features\": [1.0,0.9,0.8,0.7,0.6,0.5,0.4],\n",
    "       \"min_samples_leaf\": [1,2,3,4,5,10,20,30,40,50]}\n",
    "            \n",
    "GS = GridSearchCV(clf, cv=5,param_grid=params, \\\n",
    "                  verbose=0,scoring=make_scorer(score_func=mean_absolute_error, greater_is_better=False),n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "G_fit = GS.fit(data, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[mean: -3161291.74697, std: 2103708.59707, params: {'max_features': 1.0, 'min_samples_leaf': 1},\n",
       " mean: -3125999.28810, std: 2173887.10376, params: {'max_features': 1.0, 'min_samples_leaf': 2},\n",
       " mean: -3108385.56475, std: 2215749.90877, params: {'max_features': 1.0, 'min_samples_leaf': 3},\n",
       " mean: -3118967.60669, std: 2239439.49723, params: {'max_features': 1.0, 'min_samples_leaf': 4},\n",
       " mean: -3128977.58883, std: 2265012.66412, params: {'max_features': 1.0, 'min_samples_leaf': 5},\n",
       " mean: -3153453.22268, std: 2362603.69047, params: {'max_features': 1.0, 'min_samples_leaf': 10},\n",
       " mean: -3238903.63083, std: 2413396.49078, params: {'max_features': 1.0, 'min_samples_leaf': 20},\n",
       " mean: -3486781.52324, std: 2434279.80618, params: {'max_features': 1.0, 'min_samples_leaf': 30},\n",
       " mean: -3642345.30916, std: 2507935.48117, params: {'max_features': 1.0, 'min_samples_leaf': 40},\n",
       " mean: -3690978.04450, std: 2540185.64266, params: {'max_features': 1.0, 'min_samples_leaf': 50},\n",
       " mean: -3143093.39768, std: 2105200.37162, params: {'max_features': 0.9, 'min_samples_leaf': 1},\n",
       " mean: -3107564.38563, std: 2169324.59429, params: {'max_features': 0.9, 'min_samples_leaf': 2},\n",
       " mean: -3099271.95397, std: 2218539.48424, params: {'max_features': 0.9, 'min_samples_leaf': 3},\n",
       " mean: -3102660.89948, std: 2253846.93096, params: {'max_features': 0.9, 'min_samples_leaf': 4},\n",
       " mean: -3112649.18496, std: 2272750.65825, params: {'max_features': 0.9, 'min_samples_leaf': 5},\n",
       " mean: -3140693.43486, std: 2381954.77741, params: {'max_features': 0.9, 'min_samples_leaf': 10},\n",
       " mean: -3263738.24815, std: 2429850.25526, params: {'max_features': 0.9, 'min_samples_leaf': 20},\n",
       " mean: -3486046.46825, std: 2459697.81026, params: {'max_features': 0.9, 'min_samples_leaf': 30},\n",
       " mean: -3616639.13248, std: 2533544.44690, params: {'max_features': 0.9, 'min_samples_leaf': 40},\n",
       " mean: -3653040.63786, std: 2562759.30720, params: {'max_features': 0.9, 'min_samples_leaf': 50},\n",
       " mean: -3126814.34559, std: 2110001.89649, params: {'max_features': 0.8, 'min_samples_leaf': 1},\n",
       " mean: -3090481.21173, std: 2177485.31461, params: {'max_features': 0.8, 'min_samples_leaf': 2},\n",
       " mean: -3081173.16402, std: 2215321.62194, params: {'max_features': 0.8, 'min_samples_leaf': 3},\n",
       " mean: -3087209.93503, std: 2252756.43921, params: {'max_features': 0.8, 'min_samples_leaf': 4},\n",
       " mean: -3102788.83861, std: 2280571.61478, params: {'max_features': 0.8, 'min_samples_leaf': 5},\n",
       " mean: -3142665.09730, std: 2387149.29516, params: {'max_features': 0.8, 'min_samples_leaf': 10},\n",
       " mean: -3285822.62760, std: 2451348.43706, params: {'max_features': 0.8, 'min_samples_leaf': 20},\n",
       " mean: -3478946.17217, std: 2481664.69432, params: {'max_features': 0.8, 'min_samples_leaf': 30},\n",
       " mean: -3598371.97547, std: 2556812.74793, params: {'max_features': 0.8, 'min_samples_leaf': 40},\n",
       " mean: -3639694.66216, std: 2595497.42764, params: {'max_features': 0.8, 'min_samples_leaf': 50},\n",
       " mean: -3129694.04613, std: 2125126.44280, params: {'max_features': 0.7, 'min_samples_leaf': 1},\n",
       " mean: -3081218.07391, std: 2185693.16226, params: {'max_features': 0.7, 'min_samples_leaf': 2},\n",
       " mean: -3072741.95923, std: 2232384.66159, params: {'max_features': 0.7, 'min_samples_leaf': 3},\n",
       " mean: -3074527.31256, std: 2257890.15344, params: {'max_features': 0.7, 'min_samples_leaf': 4},\n",
       " mean: -3093974.07074, std: 2298072.14164, params: {'max_features': 0.7, 'min_samples_leaf': 5},\n",
       " mean: -3138141.87427, std: 2407277.36171, params: {'max_features': 0.7, 'min_samples_leaf': 10},\n",
       " mean: -3296215.93618, std: 2463237.11623, params: {'max_features': 0.7, 'min_samples_leaf': 20},\n",
       " mean: -3485673.38726, std: 2505356.83860, params: {'max_features': 0.7, 'min_samples_leaf': 30},\n",
       " mean: -3596674.37822, std: 2571224.35124, params: {'max_features': 0.7, 'min_samples_leaf': 40},\n",
       " mean: -3637967.02380, std: 2622554.56053, params: {'max_features': 0.7, 'min_samples_leaf': 50},\n",
       " mean: -3124643.13369, std: 2139892.06986, params: {'max_features': 0.6, 'min_samples_leaf': 1},\n",
       " mean: -3084384.11392, std: 2198282.76266, params: {'max_features': 0.6, 'min_samples_leaf': 2},\n",
       " mean: -3069592.78350, std: 2238970.29691, params: {'max_features': 0.6, 'min_samples_leaf': 3},\n",
       " mean: -3085327.58516, std: 2277515.47815, params: {'max_features': 0.6, 'min_samples_leaf': 4},\n",
       " mean: -3094301.84741, std: 2319206.07456, params: {'max_features': 0.6, 'min_samples_leaf': 5},\n",
       " mean: -3164773.54320, std: 2430228.81278, params: {'max_features': 0.6, 'min_samples_leaf': 10},\n",
       " mean: -3331637.92688, std: 2506741.29570, params: {'max_features': 0.6, 'min_samples_leaf': 20},\n",
       " mean: -3507448.83990, std: 2555554.38390, params: {'max_features': 0.6, 'min_samples_leaf': 30},\n",
       " mean: -3596540.26543, std: 2605319.86196, params: {'max_features': 0.6, 'min_samples_leaf': 40},\n",
       " mean: -3639325.70277, std: 2662208.00259, params: {'max_features': 0.6, 'min_samples_leaf': 50},\n",
       " mean: -3119818.52515, std: 2140362.46454, params: {'max_features': 0.5, 'min_samples_leaf': 1},\n",
       " mean: -3074475.20802, std: 2217977.06601, params: {'max_features': 0.5, 'min_samples_leaf': 2},\n",
       " mean: -3073360.11231, std: 2258100.46410, params: {'max_features': 0.5, 'min_samples_leaf': 3},\n",
       " mean: -3088774.37846, std: 2301455.62539, params: {'max_features': 0.5, 'min_samples_leaf': 4},\n",
       " mean: -3103063.14939, std: 2328608.29866, params: {'max_features': 0.5, 'min_samples_leaf': 5},\n",
       " mean: -3187331.03113, std: 2447600.52994, params: {'max_features': 0.5, 'min_samples_leaf': 10},\n",
       " mean: -3357084.75966, std: 2543585.82709, params: {'max_features': 0.5, 'min_samples_leaf': 20},\n",
       " mean: -3509434.63469, std: 2578445.20390, params: {'max_features': 0.5, 'min_samples_leaf': 30},\n",
       " mean: -3602909.10396, std: 2635274.88519, params: {'max_features': 0.5, 'min_samples_leaf': 40},\n",
       " mean: -3640802.55186, std: 2687985.19939, params: {'max_features': 0.5, 'min_samples_leaf': 50},\n",
       " mean: -3125297.96800, std: 2196292.65780, params: {'max_features': 0.4, 'min_samples_leaf': 1},\n",
       " mean: -3083750.74620, std: 2249631.71272, params: {'max_features': 0.4, 'min_samples_leaf': 2},\n",
       " mean: -3096397.17566, std: 2292889.15783, params: {'max_features': 0.4, 'min_samples_leaf': 3},\n",
       " mean: -3096942.36965, std: 2334580.90500, params: {'max_features': 0.4, 'min_samples_leaf': 4},\n",
       " mean: -3123478.99140, std: 2376593.74710, params: {'max_features': 0.4, 'min_samples_leaf': 5},\n",
       " mean: -3213550.63413, std: 2494065.84360, params: {'max_features': 0.4, 'min_samples_leaf': 10},\n",
       " mean: -3396908.84548, std: 2578041.95007, params: {'max_features': 0.4, 'min_samples_leaf': 20},\n",
       " mean: -3528991.09141, std: 2632711.41295, params: {'max_features': 0.4, 'min_samples_leaf': 30},\n",
       " mean: -3615242.05557, std: 2685346.54066, params: {'max_features': 0.4, 'min_samples_leaf': 40},\n",
       " mean: -3652010.21715, std: 2733974.84801, params: {'max_features': 0.4, 'min_samples_leaf': 50}]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.grid_scores_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_features': 0.6, 'min_samples_leaf': 3}"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf = RandomForestRegressor(n_estimators=10000, random_state=9)\n",
    "\n",
    "params={\"max_features\": [0.6],\n",
    "       \"min_samples_leaf\": [3]}\n",
    "            \n",
    "GS = GridSearchCV(clf, cv=5,param_grid=params, \\\n",
    "                  verbose=0,scoring=make_scorer(score_func=mean_absolute_error, greater_is_better=False),n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "G_fit = GS.fit(data, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[mean: -3077370.98164, std: 2257628.88262, params: {'max_features': 0.6, 'min_samples_leaf': 3}]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.grid_scores_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Итоговый rmse для случайного леса на кросс валидации"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1754.2437064563649"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sqrt(GS.best_score_*(-1.0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}