NYC_Property_Regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# NYC Property Appraisal - Regression Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Data Analysis & Cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "pd.options.display.max_columns=30"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv('real_estate_data.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It is obvious that *'tx_price'* is the y_label we want to predict."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tx_price</th>\n",
       "      <th>beds</th>\n",
       "      <th>baths</th>\n",
       "      <th>sqft</th>\n",
       "      <th>year_built</th>\n",
       "      <th>lot_size</th>\n",
       "      <th>property_type</th>\n",
       "      <th>exterior_walls</th>\n",
       "      <th>roof</th>\n",
       "      <th>basement</th>\n",
       "      <th>restaurants</th>\n",
       "      <th>groceries</th>\n",
       "      <th>nightlife</th>\n",
       "      <th>cafes</th>\n",
       "      <th>shopping</th>\n",
       "      <th>arts_entertainment</th>\n",
       "      <th>beauty_spas</th>\n",
       "      <th>active_life</th>\n",
       "      <th>median_age</th>\n",
       "      <th>married</th>\n",
       "      <th>college_grad</th>\n",
       "      <th>property_tax</th>\n",
       "      <th>insurance</th>\n",
       "      <th>median_school</th>\n",
       "      <th>num_schools</th>\n",
       "      <th>tx_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>295850</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>584</td>\n",
       "      <td>2013</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>Wood Siding</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>107</td>\n",
       "      <td>9</td>\n",
       "      <td>30</td>\n",
       "      <td>19</td>\n",
       "      <td>89</td>\n",
       "      <td>6</td>\n",
       "      <td>47</td>\n",
       "      <td>58</td>\n",
       "      <td>33.0</td>\n",
       "      <td>65.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>234.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>216500</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>612</td>\n",
       "      <td>1965</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>Brick</td>\n",
       "      <td>Composition Shingle</td>\n",
       "      <td>1.0</td>\n",
       "      <td>105</td>\n",
       "      <td>15</td>\n",
       "      <td>6</td>\n",
       "      <td>13</td>\n",
       "      <td>87</td>\n",
       "      <td>2</td>\n",
       "      <td>26</td>\n",
       "      <td>14</td>\n",
       "      <td>39.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>69.0</td>\n",
       "      <td>169.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>279900</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>615</td>\n",
       "      <td>1963</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>Wood Siding</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>183</td>\n",
       "      <td>13</td>\n",
       "      <td>31</td>\n",
       "      <td>30</td>\n",
       "      <td>101</td>\n",
       "      <td>10</td>\n",
       "      <td>74</td>\n",
       "      <td>62</td>\n",
       "      <td>28.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>216.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>379900</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>618</td>\n",
       "      <td>2000</td>\n",
       "      <td>33541</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>Wood Siding</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>198</td>\n",
       "      <td>9</td>\n",
       "      <td>38</td>\n",
       "      <td>25</td>\n",
       "      <td>127</td>\n",
       "      <td>11</td>\n",
       "      <td>72</td>\n",
       "      <td>83</td>\n",
       "      <td>36.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>91.0</td>\n",
       "      <td>265.0</td>\n",
       "      <td>92.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>340000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>634</td>\n",
       "      <td>1992</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>Brick</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>149</td>\n",
       "      <td>7</td>\n",
       "      <td>22</td>\n",
       "      <td>20</td>\n",
       "      <td>83</td>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>73</td>\n",
       "      <td>37.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2002</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   tx_price  beds  baths  sqft  year_built  lot_size  \\\n",
       "0    295850     1      1   584        2013         0   \n",
       "1    216500     1      1   612        1965         0   \n",
       "2    279900     1      1   615        1963         0   \n",
       "3    379900     1      1   618        2000     33541   \n",
       "4    340000     1      1   634        1992         0   \n",
       "\n",
       "                   property_type exterior_walls                 roof  \\\n",
       "0  Apartment / Condo / Townhouse    Wood Siding                  NaN   \n",
       "1  Apartment / Condo / Townhouse          Brick  Composition Shingle   \n",
       "2  Apartment / Condo / Townhouse    Wood Siding                  NaN   \n",
       "3  Apartment / Condo / Townhouse    Wood Siding                  NaN   \n",
       "4  Apartment / Condo / Townhouse          Brick                  NaN   \n",
       "\n",
       "   basement  restaurants  groceries  nightlife  cafes  shopping  \\\n",
       "0       NaN          107          9         30     19        89   \n",
       "1       1.0          105         15          6     13        87   \n",
       "2       NaN          183         13         31     30       101   \n",
       "3       NaN          198          9         38     25       127   \n",
       "4       NaN          149          7         22     20        83   \n",
       "\n",
       "   arts_entertainment  beauty_spas  active_life  median_age  married  \\\n",
       "0                   6           47           58        33.0     65.0   \n",
       "1                   2           26           14        39.0     73.0   \n",
       "2                  10           74           62        28.0     15.0   \n",
       "3                  11           72           83        36.0     25.0   \n",
       "4                  10           50           73        37.0     20.0   \n",
       "\n",
       "   college_grad  property_tax  insurance  median_school  num_schools  tx_year  \n",
       "0          84.0         234.0       81.0            9.0          3.0     2013  \n",
       "1          69.0         169.0       51.0            3.0          3.0     2006  \n",
       "2          86.0         216.0       74.0            8.0          3.0     2012  \n",
       "3          91.0         265.0       92.0            9.0          3.0     2005  \n",
       "4          75.0          88.0       30.0            9.0          3.0     2002  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head() # There are three categorical columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1883, 26)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tx_price                0\n",
       "beds                    0\n",
       "baths                   0\n",
       "sqft                    0\n",
       "year_built              0\n",
       "lot_size                0\n",
       "property_type           0\n",
       "exterior_walls        223\n",
       "roof                  354\n",
       "basement              226\n",
       "restaurants             0\n",
       "groceries               0\n",
       "nightlife               0\n",
       "cafes                   0\n",
       "shopping                0\n",
       "arts_entertainment      0\n",
       "beauty_spas             0\n",
       "active_life             0\n",
       "median_age              0\n",
       "married                 0\n",
       "college_grad            0\n",
       "property_tax            0\n",
       "insurance               0\n",
       "median_school           0\n",
       "num_schools             0\n",
       "tx_year                 0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The roof and exterior_walls are full of null values and are difficult to estimate effective proxies. So they are dropped. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.drop(columns=['roof', 'exterior_walls'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ nan,   1.])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.basement.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.basement = data.basement.fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.,  1.])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.basement.unique() # No more NaN values!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualise distribution & check if any outliers or erroneous values "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x110b752e8>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.tx_price.plot.box() # transaction price between $200K to $800K (USD)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10cd2ba20>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.tx_price.plot.hist() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10bceb4e0>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.tx_year.plot.hist() # most transactions after 2000s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1113ee630>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.sqft.plot.box() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1114a9940>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.sqft.plot.hist() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1113e5c88>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.year_built.plot.box()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11198b550>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.year_built.plot.kde() # the construction years are centered around 1970 to 2010"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEOCAYAAACAfcAXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGM5JREFUeJzt3X2UVPWd5/H3RzoQSVSQbgzSJBCCiiAitEBmTiQJowjmBIKjkWNWDJwlbnSyLpkYnNkTx8mYaHazPiSuhIwk6HFBTiYZewzjw/oQTTIYGh+ISAJ9kEiLkWYBo+P40PrdP+oHFE03DVVNFfbv8zqnT9/7vb9b9S0O8Kn7u7fqKiIwM7P8HFXtBszMrDocAGZmmXIAmJllygFgZpYpB4CZWaYcAGZmmXIAWI8gabOkv6h2H2bvJQ4AsyOYpE9Kaql2H9YzOQDMzDLlALCe5ExJz0naKelHkt4vqb+keyW1pvq9kup37yDpUkmbJL0q6XlJFxdtmytpfdrvfkkfKdoWkr4saWPa95uShkv6N0l/krRCUu+i8Z+R9LSkXZJ+LWlM0bbNkv5a0lpJr0i6O/X+AeBfgRMlvZZ+Tjz8f4yWCweA9SQXA1OB4cBJwH+n8Hf8R8BHgA8D/wF8HyD9B3sLMC0ijgH+DHg6bZsJ/A0wC6gDHgeWtXu+c4HxwCTgKmBx6mEIMBqYnR5rHLAE+BIwAPgB0CipT9FjXZgebxgwBrg0Iv4dmAZsjYgPpp+t5f4hme3mALCe5PsRsSUidgDXAbMj4v9FxD9FxOsR8WqqTy7a511gtKSjI+KliFiX6l8Cvh0R6yOiDfgWMLb4KAC4ISL+lPZ5FnggIjZFxCsU3rmfkcb9Z+AHEfFERLwTEUuBNykEx263RMTW1Pu/AGO78w/GrCMOAOtJthQt/4HC1ElfST+Q9AdJfwIeA/pJ6pXeYX8euAx4SdLPJZ2S9v8IcHOastkF7AAEDC56jpeLlv+jg/UPFj3WV3c/Vnq8IUDxdM4fi5ZfL9rX7LBxAFhPMqRo+cPAVuCrwMnAxIg4FjgrbRdARNwfEWcDg4DfAT9M27cAX4qIfkU/R0fEr0voawtwXbvH6hsR7aeUOuKv67XDxgFgPcnlkuolHU9h/v5u4BgK78Z3pfo1uwdLOkHSZ9O5gDeB14B30uZFwNWSRqWxx0m6oMS+fghcJmmiCj4g6TxJxxzEvi8DAyQdV+Jzm3XKAWA9yf8BHgA2pZ9/AG4Cjga2A6uA+4rGH0XhCGErhSmeycCXASLiZ8ANwPI0dfQshROyhywimiicB/g+sBNoBi49yH1/R+Hk86Y0feSrgKzbyDeEMTPLk48AzMwy5QAwM8uUA8DMLFMOADOzTNVUu4EDqa2tjaFDh1a7DTOz95Q1a9Zsj4i6rsYd0QEwdOhQmpqaqt2Gmdl7iqQ/HMw4TwGZmWXKAfAeN3fuXAYOHMjo0aP3qX/ve9/j5JNPZtSoUVx11VUAbN68maOPPpqxY8cyduxYLrvsMgBeffXVPbWxY8dSW1vLlVdeWfHXYmaVdURPAVnXLr30Uq644gouueSSPbVHHnmEe+65h7Vr19KnTx+2bdu2Z9vw4cN5+umn93mMY445Zp/a+PHjmTVr1uFv3syqykcA73FnnXUWxx9//D612267jYULF9KnT+Hr5gcOHHjQj7dx40a2bdvGJz7xiW7t08yOPA6AHmjDhg08/vjjTJw4kcmTJ7N69eo9255//nnOOOMMJk+ezOOPP77fvsuWLePzn/88kirZsplVgaeAeqC2tjZ27tzJqlWrWL16NRdeeCGbNm1i0KBBvPDCCwwYMIA1a9Ywc+ZM1q1bx7HHHrtn3+XLl3PnnXdWsXszqxQfAfRA9fX1zJo1C0lMmDCBo446iu3bt9OnTx8GDBgAFOb5hw8fzoYNG/bs98wzz9DW1sb48eOr1bqZVZADoAeaOXMmDz/8MFCYDnrrrbeora2ltbWVd94pfN39pk2b2LhxIx/96Ef37Lds2TJmz55dlZ7NrPI8BfQeN3v2bB599FG2b99OfX091157LXPnzmXu3LmMHj2a3r17s3TpUiTx2GOP8Y1vfIOamhp69erFokWL9jmBvGLFClauXFnFV2NmlXRE3w+goaEh3gufBB668OfVbqFH2Xz9edVuwew9TdKaiGjoapyngMzMMuUAMDPLlAPAzCxTXQaApCWStkl6tl39ryT9XtI6Sd8pql8tqTltm1pUPzfVmiUt7N6XYWZmh+pgrgL6MfB94I7dBUmfAmYAYyLiTUkDU/1U4CJgFHAi8H8lnZR2uxU4G2gBVktqjIjnuuuFmJnZoekyACLiMUlD25X/C3B9RLyZxuz+trEZwPJUf15SMzAhbWuOiE0AkpansQ4AM7MqKfUcwEnAJyQ9IekXks5M9cHAlqJxLanWWX0/kuZLapLU1NraWmJ7ZmbWlVIDoAboD0wCvgasUOHbwzr6BrE4QH3/YsTiiGiIiIa6ui7vaGZmZiUq9ZPALcBPo/Apst9IeheoTfUhRePqga1pubO6mZlVQalHAP8MfBogneTtDWwHGoGLJPWRNAwYAfwGWA2MkDRMUm8KJ4oby23ezMxK1+URgKRlwCeBWkktwDXAEmBJujT0LWBOOhpYJ2kFhZO7bcDlEfFOepwrgPuBXsCSiFh3GF6PmZkdpIO5Cqizr4f8QifjrwOu66C+EvA3jZmZHSH8SWAzs0w5AMzMMuUAMDPLlAPAzCxTDgAzs0w5AMzMMuUAMDPLlAPAzCxTDgAzs0w5AMzMMuUAMDPLlAPAzCxTDgAzs0w5AMzMMuUAMDPLlAPAzCxTXQaApCWStqW7f7Xf9teSQlJtWpekWyQ1S1oraVzR2DmSNqafOd37MszM7FAdzBHAj4Fz2xclDQHOBl4oKk+jcB/gEcB84LY09ngKt5KcCEwArpHUv5zGzcysPF0GQEQ8BuzoYNONwFVAFNVmAHdEwSqgn6RBwFTgwYjYERE7gQfpIFTMzKxySjoHIOmzwIsR8Uy7TYOBLUXrLanWWb2jx54vqUlSU2trayntmZnZQTjkAJDUF/hb4Bsdbe6gFgeo71+MWBwRDRHRUFdXd6jtmZnZQSrlCGA4MAx4RtJmoB54UtKHKLyzH1I0th7YeoC6mZlVySEHQET8NiIGRsTQiBhK4T/3cRHxR6ARuCRdDTQJeCUiXgLuB86R1D+d/D0n1czMrEoO5jLQZcC/ASdLapE07wDDVwKbgGbgh8CXASJiB/BNYHX6+ftUMzOzKqnpakBEzO5i+9Ci5QAu72TcEmDJIfZnZmaHiT8JbGaWKQeAmVmmHABmZplyAJiZZcoBYGaWKQeAmVmmHABmZplyAJiZZcoBYGaWKQeAmVmmHABmZplyAJiZZcoBYGaWKQeAmVmmHABmZplyAJiZZepg7gi2RNI2Sc8W1f6HpN9JWivpZ5L6FW27WlKzpN9LmlpUPzfVmiUt7P6XYmZmh+JgjgB+DJzbrvYgMDoixgAbgKsBJJ0KXASMSvv8b0m9JPUCbgWmAacCs9NYMzOrki4DICIeA3a0qz0QEW1pdRVQn5ZnAMsj4s2IeJ7CvYEnpJ/miNgUEW8By9NYMzOrku44BzAX+Ne0PBjYUrStJdU6q+9H0nxJTZKaWltbu6E9MzPrSFkBIOlvgTbgrt2lDobFAer7FyMWR0RDRDTU1dWV056ZmR1ATak7SpoDfAaYEhG7/zNvAYYUDasHtqblzupmZlYFJR0BSDoX+Drw2Yh4vWhTI3CRpD6ShgEjgN8Aq4ERkoZJ6k3hRHFjea2bmVk5ujwCkLQM+CRQK6kFuIbCVT99gAclAayKiMsiYp2kFcBzFKaGLo+Id9LjXAHcD/QClkTEusPweszM7CB1GQARMbuD8u0HGH8dcF0H9ZXAykPqzszMDht/EtjMLFMOADOzTDkAzMwy5QAwM8uUA8DMLFMOADOzTDkAzMwy5QAwM8uUA8DMLFMOADOzTDkAzMwy5QAwM8uUA8DMLFMOADOzTDkAzMwy5QAwM8tUlwEgaYmkbZKeLaodL+lBSRvT7/6pLkm3SGqWtFbSuKJ95qTxG9P9hM3MrIoO5gjgx8C57WoLgYciYgTwUFoHmEbhPsAjgPnAbVAIDAq3kpwITACu2R0aZmZWHV0GQEQ8BuxoV54BLE3LS4GZRfU7omAV0E/SIGAq8GBE7IiIncCD7B8qZmZWQaWeAzghIl4CSL8HpvpgYEvRuJZU66y+H0nzJTVJamptbS2xPTMz60p3nwRWB7U4QH3/YsTiiGiIiIa6urpubc7MzPYqNQBeTlM7pN/bUr0FGFI0rh7YeoC6mZlVSakB0AjsvpJnDnBPUf2SdDXQJOCVNEV0P3COpP7p5O85qWZmZlVS09UAScuATwK1klooXM1zPbBC0jzgBeCCNHwlMB1oBl4HvggQETskfRNYncb9fUS0P7FsZmYV1GUARMTsTjZN6WBsAJd38jhLgCWH1J2ZmR02/iSwmVmmHABmZplyAJiZZcoBYGaWKQeAmVmmHABmZplyAJiZZcoBYGaWKQeAmVmmHABmZplyAJiZZcoBYGaWKQeAmVmmHABmZplyAJiZZcoBYGaWqbICQNJ/k7RO0rOSlkl6v6Rhkp6QtFHS3ZJ6p7F90npz2j60O16AmZmVpuQAkDQY+ArQEBGjgV7ARcANwI0RMQLYCcxLu8wDdkbEx4Ab0zgzM6uScqeAaoCjJdUAfYGXgE8DP0nblwIz0/KMtE7aPkWSynx+MzMrUckBEBEvAv+Twk3hXwJeAdYAuyKiLQ1rAQan5cHAlrRvWxo/oP3jSpovqUlSU2tra6ntmZlZF8qZAupP4V39MOBE4APAtA6Gxu5dDrBtbyFicUQ0RERDXV1dqe2ZmVkXypkC+gvg+YhojYi3gZ8Cfwb0S1NCAPXA1rTcAgwBSNuPA3aU8fxmZlaGcgLgBWCSpL5pLn8K8BzwCPCXacwc4J603JjWSdsfjoj9jgDMzKwyyjkH8ASFk7lPAr9Nj7UY+DqwQFIzhTn+29MutwMDUn0BsLCMvs3MrEw1XQ/pXERcA1zTrrwJmNDB2DeAC8p5PjMz6z7+JLCZWaYcAGZmmXIAmJllygFgZpYpB4CZWaYcAGZmmXIAmJllygFgZpYpB4CZWaYcAGZmmXIAmJllygFgZpYpB4CZWaYcAGZmmXIAmJllygFgZpapsgJAUj9JP5H0O0nrJX1c0vGSHpS0Mf3un8ZK0i2SmiWtlTSue16CmZmVotwjgJuB+yLiFOB0YD2FWz0+FBEjgIfYe+vHacCI9DMfuK3M5zYzszKUHACSjgXOIt3zNyLeiohdwAxgaRq2FJiZlmcAd0TBKqCfpEEld25mZmUp5wjgo0Ar8CNJT0n6R0kfAE6IiJcA0u+BafxgYEvR/i2ptg9J8yU1SWpqbW0toz0zMzuQcgKgBhgH3BYRZwD/zt7pno6og1rsV4hYHBENEdFQV1dXRntmZnYg5QRAC9ASEU+k9Z9QCISXd0/tpN/bisYPKdq/HthaxvObmVkZSg6AiPgjsEXSyak0BXgOaATmpNoc4J603Ahckq4GmgS8snuqyMzMKq+mzP3/CrhLUm9gE/BFCqGyQtI84AXggjR2JTAdaAZeT2PNzKxKygqAiHgaaOhg05QOxgZweTnPZ2Zm3cefBDYzy5QDwMwsUw4AM7NMOQDMzDLlADAzy5QDwMwsUw4AM7NMOQDMzDLlADAzy5QDwMwsUw4AM7NMOQDMzDLlADAzy5QDwMwsUw4AM7NMOQDMzDJVdgBI6iXpKUn3pvVhkp6QtFHS3eluYUjqk9ab0/ah5T63mZmVrjuOAP4rsL5o/QbgxogYAewE5qX6PGBnRHwMuDGNMzOzKikrACTVA+cB/5jWBXwa+EkashSYmZZnpHXS9ilpvJmZVUG5RwA3AVcB76b1AcCuiGhL6y3A4LQ8GNgCkLa/ksbvQ9J8SU2SmlpbW8tsz8zMOlNyAEj6DLAtItYUlzsYGgexbW8hYnFENEREQ11dXantmZlZF2rK2PfPgc9Kmg68HziWwhFBP0k16V1+PbA1jW8BhgAtkmqA44AdZTy/mZmVoeQjgIi4OiLqI2IocBHwcERcDDwC/GUaNge4Jy03pnXS9ocjYr8jADMzq4zD8TmArwMLJDVTmOO/PdVvBwak+gJg4WF4bjMzO0jlTAHtERGPAo+m5U3AhA7GvAFc0B3PZ2Zm5fMngc3MMuUAMDPLlAPAzCxTDgAzs0w5AMzMMuUAMDPLlAPAzCxTDgAzs0w5AMzssNiyZQuf+tSnGDlyJKNGjeLmm28G4Gtf+xqnnHIKY8aM4XOf+xy7du3as8/atWv5+Mc/zqhRozjttNN44403qtV+FhwAZnZY1NTU8N3vfpf169ezatUqbr31Vp577jnOPvtsnn32WdauXctJJ53Et7/9bQDa2tr4whe+wKJFi1i3bh2PPvoo73vf+6r8Kno2B4CZHRaDBg1i3LhxABxzzDGMHDmSF198kXPOOYeamsK30EyaNImWlhYAHnjgAcaMGcPpp58OwIABA+jVq1d1ms+EA8DMDrvNmzfz1FNPMXHixH3qS5YsYdq0aQBs2LABSUydOpVx48bxne98pxqtZqVbvgzOzKwzr732Gueffz433XQTxx577J76ddddR01NDRdffDFQmAL65S9/yerVq+nbty9Tpkxh/PjxTJkypVqt93g+AjCzw+btt9/m/PPP5+KLL2bWrFl76kuXLuXee+/lrrvuYvetwevr65k8eTK1tbX07duX6dOn8+STT1ar9Sw4AMzssIgI5s2bx8iRI1mwYMGe+n333ccNN9xAY2Mjffv23VOfOnUqa9eu5fXXX6etrY1f/OIXnHrqqdVoPRueAjKzw+JXv/oVd955J6eddhpjx44F4Fvf+hZf+cpXePPNNzn77LOBwongRYsW0b9/fxYsWMCZZ56JJKZPn855551XzZfQ46nUuzJKGgLcAXwIeBdYHBE3SzoeuBsYCmwGLoyInSoc590MTAdeBy6NiAMe3zU0NERTU1NJ/VXS0IU/r3YLPcrm6/2Pvlv93XHV7qDn+LtXqt3BQZG0JiIauhpXzhRQG/DViBgJTAIul3QqhVs9PhQRI4CH2Hvrx2nAiPQzH7itjOc2M7MylXNT+Jd2v4OPiFeB9cBgYAawNA1bCsxMyzOAO6JgFdBP0qCSOzczs7J0y0lgSUOBM4AngBMi4iUohAQwMA0bDGwp2q0l1do/1nxJTZKaWltbu6M9MzPrQNkBIOmDwD8BV0bEnw40tIPaficgImJxRDRERENdXV257ZmZWSfKCgBJ76Pwn/9dEfHTVH5599RO+r0t1VuAIUW71wNby3l+MzMrXckBkK7quR1YHxH/q2hTIzAnLc8B7imqX6KCScAru6eKzMys8sr5HMCfA/8J+K2kp1Ptb4DrgRWS5gEvABekbSspXALaTOEy0C+W8dxmZlamkgMgIn5Jx/P6APt9eUcUPnBweanPZ2Zm3ctfBWFmlikHgJlZphwAZmaZcgCYmWXKAWBmlikHgJlZphwAZmaZcgCYmWXKAWBmlikHgJlZphwAZmaZcgCYmWXKAWBmlikHgJlZphwAZmaZcgCYmWWq4gEg6VxJv5fULGlhpZ/fzMwKKhoAknoBtwLTgFOB2ZJOrWQPZmZWUOkjgAlAc0Rsioi3gOXAjAr3YGZmlHdT+FIMBrYUrbcAE4sHSJoPzE+rr0n6fYV6y0EtsL3aTXRFN1S7A6uSI//v57Wd3Qb9iPORgxlU6QDo6E8v9lmJWAwsrkw7eZHUFBEN1e7DrCP++1l5lZ4CagGGFK3XA1sr3IOZmVH5AFgNjJA0TFJv4CKgscI9mJkZFZ4Ciog2SVcA9wO9gCURsa6SPWTOU2t2JPPfzwpTRHQ9yszMehx/EtjMLFMOADOzTDkAzMwy5QAws6qRdLyk/tXuI1cOADOrKEkflrRcUivwBLBa0rZUG1rd7vLiAOjhJJ0gaZykMySdUO1+zIC7gZ8BH4qIERHxMWAQ8M8Uvh/MKsSXgfZQksYCi4DjgBdTuR7YBXw5Ip6sVm+WN0kbI2LEoW6z7ucA6KEkPQ18KSKeaFefBPwgIk6vTmeWO0nLgR3AUvZ+OeQQYA5QGxEXVqu33DgAeqgu3mU1p8Nus4pLXwMzj8JXwQ+m8CWRW4B/AW6PiDer2F5WHAA9lKRbgOHAHez7LusS4PmIuKJavZnZkcEB0INJmsa+77JagMaIWFnVxsw6IekzEXFvtfvIhQPAzI4Ykq6NiGuq3UcuHAAZkjQ/3XjHrCokncLeo9OgcF+QxohYX9XGMuPPAeTpPXNfO+t5JH2dwvX+An5D4T4hApZJWljN3nLjI4AMSfpiRPyo2n1YniRtAEZFxNvt6r2Bdf4cQOX4CCBP11a7Acvau8CJHdQHpW1WIZW+KbxViKS1nW0C/JUQVk1XAg9J2sjeS5Q/DHwM8OXJFeQpoB5K0svAVGBn+03AryOio3dgZhUh6ShgAvteorw6It6pamOZ8RFAz3Uv8MGIeLr9BkmPVr4ds70i4l1gVbX7yJ2PAMzMMuWTwGZmmXIAmJllygFgZpYpB4CZWab+P53RFrfwJxg8AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = data.basement.value_counts().plot.bar(title='basement') # most are with basement\n",
    "for p in ax.patches:\n",
    "    ax.annotate(str(p.get_height()), (p.get_x() * 1.005 + 0.175, p.get_height() * 1.005))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[['beds','baths']].apply(pd.value_counts).plot.bar(title='number of beds-baths')\n",
    "for p in ax.patches:\n",
    "    ax.annotate(str(p.get_height()), (p.get_x() * 1.005 + 0.175, p.get_height() * 1.005))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = data[['restaurants', 'groceries', 'nightlife', 'cafes', 'shopping', 'arts_entertainment', 'beauty_spas', 'active_life']].\\\n",
    "    apply((pd.value_counts)).plot.box(figsize=(18,10), title='number of facilities & shops within 1 mile')\n",
    "    # number of facilities within 1 mile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1e95c2e8>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlkAAAFACAYAAACPyWmJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3X2YXHV99/H3Nw/skkBDgOjNgyHpJdU1UXwIFmuqxiCoUaQtWiltweYOXe+y0gbbxKwWqc0W2hKp4aorGDS2GEFsb6nYSowB70DFBuTBdLV4SXgwUSIJETYEQvzef8zZOFk3u8nunp3dmffruubaOb/z8PvOZDLzmXN+c05kJpIkSRpe42pdgCRJUj0yZEmSJJXAkCVJklQCQ5YkSVIJDFmSJEklMGRJkiSVwJAl6YAiYlNEvOkgl90cEacfYN6bIuKxYazrgojYMFzbG6CvjIgXj0RfB+o3Ijoj4iMjXYOkoZlQ6wIkjV6ZOavWNQgys7XnfhF6/zkzT6xdRZIOhnuyJI0pEeGXQ0ljgiFLqnPFYbwPRsT9EbEzIm6IiOaq+e+IiHsj4smIuDMiXtFr3dOL+4dHxOqI2BERXRHxF30cAnzlgfoptrEsIn5abPe8qvYpEfG5iNgWEQ9HxIcjYlwx74KIuCMiPh4R24GPVq3390U9D0XE26raj4+ImyNie0T8ICIWVc1rioirImJLcbsqIpqq5v95RGwt5v3RAM/tzIi4PSKeioi1EXF1RPxzMe+XDpH2ej5fGxH/WTzvW4t1DztAP5+NiL+OiMnAvwPHR8TTxe34iNgVEcdULf+a4rmc2F/9ksplyJIaw3uAtwIzgVcAFwBExKuB64A/Bo4BPgXcXB06qlwKzAB+FXgL8PsH20/hfwHHAicA5wPXRMRLinkrgSnFtt8I/CHwvqp1fx34IfACYHlV2/eLbf4tsCoiopi3BngMOB44B+iIiPnFvHbgNOCVwCnAa4EPF8/HW4EPFo/vZKDPMWZVPg/cXdTwseJxHay9wJ8V674OmA/8n/5WyMxu4G3Alsw8orhtAW6j8tz3+H3gC5m55xDqkTTMDFlSY/hEZm7JzO3Av1EJGACLgE9l5l2ZuTczVwPPUgkhvb0H6MjMHZn5GPCJQ+inx0cy89nMvB24BXhPRIwHfhf4UGY+lZmbgSuBP6hab0tmrszM5zPzmaLt4cy8NjP3AquB44AXRsSLgLnAkszcnZn3Ap+u2t55wF9l5uOZuQ24rGree4DPZOZ3i0Dz0QM9oRExHTi16jF9s3jMByUz787MbxWPaTOVgPvGg12/l9UUobd4Ps8F/mmQ25I0TAxZUmP4cdX9XcARxf2TgEuKQ1ZPRsSTwIuo7AHq7Xjg0arpR/tY5kD9AOwogkuPh4ttHgscVkxXzzvhYPvKzF3F3SOKbW7PzKcOsL3j++jr+Kp5j/aadyDH0/djOigR8WsR8ZWI+HFE/AzooPJcDMaXgZdFRM9exp2Z+e1BbkvSMDFkSY3tUWB5Zh5VdZuUmWv6WHYrUP2LthcdYl9TizFFPaYDW4CfAnuoBL7qeT+qms5D6GcLcHREHHmA7W3po68txf2t7P+4pvfTz1b6fkw9uoFJPRPFHqZpVfM/CXwPODkzfwVYBgQD+6XnIjN3AzdS2Uv3B7gXSxoVDFlSY7sWaI2IX4+KyRGxoFdA6XEj8KGImBoRJwAXDaK/yyLisIj4TeAdwBeLw303Assj4siIOAlYDPzzYB5QZj4K3An8TUQ0FwP5FwLXF4usAT4cEdMi4ljgL6v6uhG4ICJeFhGTqIxDO1A/DwMbqx7TXOCdVYv8D9BcPJ8TqYz7qh7rdiTwM+DpiHgp8P6DfIg/AY6JiCm92j9HZQzcWQzyuZM0vAxZUgPLzI1UxmVdDewAfsD+g9Wr/RWVweQPAV8HbqIyfutg/bjoYwuVwNOamd8r5rVR2fPzQ2ADlQHl1x3Ctns7l8og/S3AvwKXZubaYt5fUwlH9wMPAPcUbWTmvwNXAd+g8lx8Y4B+fo/KAPztVALZ53pmZOZOKgPZP01lL1o3leevxweL9Z+iEnZvOJgHVjxna4AfFod4jy/a7wB+DtxTjPGSVGOReSh74SWpIiLeD7w3Mwc7WLvuRMRHgRdnZl+/vByJ/r8BfD4zP12L/iXtzz1Zkg5KRBwXEa+PiHHFqRcuobKXSKNARJwKvJqD3CMmqXyeOVnSwTqMymkGZgJPAl8A/rGmFQmAiFgNnA1c3OtXlZJqyMOFkiRJJfBwoSRJUgkMWZIkSSUYFWOyjj322JwxY0aty5AkSRrQ3Xff/dPMnDbQcqMiZM2YMYONGzfWugxJkqQBRcRBXULLw4WSJEklMGRJkiSVwJAlSZJUAkOWJElSCQxZkiRJJTBkSZIklcCQJUmSVIIBQ1ZEXBcRj0fEd6vajo6ItRHxYPF3atEeEfGJiPhBRNwfEa8us3hJkqTR6mD2ZH0WeGuvtqXAusw8GVhXTAO8DTi5uF0IfHJ4ypQkSRpbBgxZmflNYHuv5ncBq4v7q4Gzq9o/lxXfAo6KiOOGq1hJkqSxYrCX1XlhZm4FyMytEfGCov0E4NGq5R4r2rb23kBEXEhlbxfTp08fZBmSGkFEjHifmTnifaqx+TqvP8M98L2vV0if/4KZeU1mzsnMOdOmDXiNRUkNLDMHdTtpyVcGva400nyd15/Bhqyf9BwGLP4+XrQ/BryoarkTgS2DL0+SJGlsGmzIuhk4v7h/PvDlqvY/LH5leBqws+ewoiRJUiMZcExWRKwB3gQcGxGPAZcClwM3RsRC4BHg3cXiXwXeDvwA2AW8r4SaJUmSRr0BQ1ZmnnuAWfP7WDaBPxlqUZIkSWOdZ3yXJEkqgSFLkiSpBIYsSZKkEhiyJEmSSmDIkiRJKoEhS5IkqQSGLEmSpBIYsiRJkkpgyJIkSSqBIUuSJKkEhixJkqQSGLIkSZJKYMiSJEkqgSFLkiSpBIYsSZKkEhiyJEmSSmDIkiRJKoEhS5IkqQSGLEmSpBIYsiRJkkpgyJIkSSqBIUuSJKkEhixJkqQSGLIkSZJKYMiSJEkqgSFLkiSpBIYsSZKkEhiyJEmSSmDIkiRJKoEhS5IkqQSGLEmSpBIYsiRJkkpgyJIkSSqBIUuSJKkEhixJkqQSGLIkSZJKMKHWBUhqHKdcdis7n9kzon3OWHrLiPU15fCJ3HfpGSPWn6TRzZAlacTsfGYPmy9fUOsySjOSgU7S6OfhQkmSpBIYsiRJkkpgyJIkSSqBIUuSJKkEhixJkqQSDClkRcSfRcSmiPhuRKyJiOaImBkRd0XEgxFxQ0QcNlzFSpIkjRWDDlkRcQLwAWBOZs4GxgPvBa4APp6ZJwM7gIXDUagkSdJYMtTDhROAwyNiAjAJ2Aq8GbipmL8aOHuIfUiSJI05gw5Zmfkj4O+BR6iEq53A3cCTmfl8sdhjwAl9rR8RF0bExojYuG3btsGWIUmSNCoN5XDhVOBdwEzgeGAy8LY+Fs2+1s/MazJzTmbOmTZt2mDLkCRJGpWGcrjwdOChzNyWmXuAfwF+AziqOHwIcCKwZYg1SpIkjTlDCVmPAKdFxKSICGA+8N/AeuCcYpnzgS8PrURJkqSxZyhjsu6iMsD9HuCBYlvXAEuAxRHxA+AYYNUw1ClJkjSmTBh4kQPLzEuBS3s1/xB47VC2K0mSNNYNKWRJkqT9nXLZrex8Zs+I9jlj6S0j1teUwydy36VnjFh/Y5khS5KkYbTzmT1svnxBrcsozUgGurHOaxdKkiSVwJAlSZJUAkOWJElSCQxZkiRJJTBkSZIklcCQJUmSVAJDliRJUgkMWZIkSSUwZEmSJJXAkCVJklQCQ5YkSVIJvHahpBFzZMtSXr56aa3LKM2RLQD1e806SYfGkCVpxDzVdbkXzpXUMDxcKEmSVAJDliRJUgkMWZIkSSUwZEmSJJXAkCVJklQCQ1aDWbNmDbNnz2b8+PHMnj2bNWvW1LokSZLqkqdwaCBr1qyhvb2dVatWMXfuXDZs2MDChQsBOPfcc2tcnSRJ9cU9WQ1k+fLlrFq1innz5jFx4kTmzZvHqlWrWL58ea1LkySp7hiyGkhXVxdz587dr23u3Ll0dXXVqCJJkuqXIauBtLS0sGHDhv3aNmzYQEtLS40qkiSpfhmyGkh7ezsLFy5k/fr17Nmzh/Xr17Nw4ULa29trXZokSXXHge8NpGdwe1tbG11dXbS0tLB8+XIHvUuSVAJDVoM599xzDVWSJI0ADxc2GM+TJUnSyHBPVgPxPFmSJI0c92Q1EM+TJUnSyHFPVgPxPFmSVL4jW5by8tVLa11GaY5sAVhQ6zLGBENWA+k5T9a8efP2tXmeLEkaXk91Xc7my+s3hMxYekutSxgzDFkNpL29nfnz55OZ+9oiguuvv76GVUmSVJ8ck9VALrroIjKTiRMnAjBx4kQyk4suuqjGlUmSVH8MWQ1k+/btzJo1i+eee47M5LnnnmPWrFls37691qVJklR3DFkN5qtf/Wq/05IkaXgYshrM29/+9n6nJUnS8DBkNZCjjz6aTZs2MXv2bB555BFmz57Npk2bOProo2tdmiRJdcdfFzaQJ554gmOOOYZNmzZx0kknAZXg9cQTT9S4MkmS6o8hq8EYqCRJGhkeLmwwZ555JuPGjSMiGDduHGeeeWatS5IkqS4ZshrImWeeya233kpraytPPvkkra2t3HrrrQYtSZJKMKSQFRFHRcRNEfG9iOiKiNdFxNERsTYiHiz+Th2uYjU0a9euZdasWVx33XUcddRRXHfddcyaNYu1a9fWujRJkurOUPdk/QPwH5n5UuAUoAtYCqzLzJOBdcW0RoHMpKuri46ODrq7u+no6KCrq2u/y+xIkqThMeiQFRG/ArwBWAWQmc9l5pPAu4DVxWKrgbOHWqSGz2mnncbixYuZNGkSixcv5rTTTqt1SZIk1aWh/LrwV4FtwGci4hTgbuBi4IWZuRUgM7dGxAv6WjkiLgQuBJg+ffoQytChuPPOO4mIWpchSVLdG8rhwgnAq4FPZuargG4O4dBgZl6TmXMyc860adOGUIYkSdLoM5SQ9RjwWGbeVUzfRCV0/SQijgMo/j4+tBI13K688kq6u7u58sora12KJEl1a9AhKzN/DDwaES8pmuYD/w3cDJxftJ0PfHlIFWpYnXfeeSxbtozJkyezbNkyzjvvvFqXJElSXRrqGd/bgOsj4jDgh8D7qAS3GyNiIfAI8O4h9qFhdMcdd7B79+590zNnzqxhNZIk1a8hhazMvBeY08es+UPZrsoxefJkNm/ezMyZM/n617/O6aefzubNm5k8eXKtS5Mkqe547cIxbjC/FNy8eTMvfvGL9013d3cf0nY8r5aGYsbSW2pdQmmmHD6x1iVIGkUMWWPcYAPPjKW3sPnyBcNcjdS/kX7N+TqXVEteu1CSJKkEhixJkqQSGLIkSZJKYMiSJEkqgSFLkiSpBIYsSZKkEhiyJEmSSmDIkiRJKoEnI5UkaZh5ZQOBIUuSpGHllQ3Uw8OFkiRJJTBkSZIklcCQJUmSVAJDliRJUgkMWZIkSSUwZEmSJJXAkCVJklQCQ5YkSVIJDFmSJEklMGRJkiSVwJAlSZJUAkOWJElSCQxZkiRJJTBkSZIklcCQJUmSVAJDliRJUgkMWZIkSSUwZEmSJJXAkCVJklQCQ5YkSVIJDFmSJEklMGRJkiSVwJAlSZJUAkOWJElSCQxZkiRJJTBkSZIklcCQJUmSVIIJtS5AFadcdis7n9kzon3OWHrLiPU15fCJ3HfpGSPWnyRJtWbIGiV2PrOHzZcvqHUZpRnJQCdJ0mjg4UJJkqQSGLIkSZJKMOSQFRHjI+I7EfGVYnpmRNwVEQ9GxA0RcdjQy5QkSRpbhmNP1sVAV9X0FcDHM/NkYAewcBj6kCRJGlOGFLIi4kRgAfDpYjqANwM3FYusBs4eSh+SJElj0VD3ZF0F/AXw82L6GODJzHy+mH4MOGGIfUiSJI05gw5ZEfEO4PHMvLu6uY9F8wDrXxgRGyNi47Zt2wZbhiRJ0qg0lD1ZrwfOiojNwBeoHCa8CjgqInrOv3UisKWvlTPzmsyck5lzpk2bNoQyJEmSRp9Bh6zM/FBmnpiZM4D3At/IzPOA9cA5xWLnA18ecpWSJEljTBnnyVoCLI6IH1AZo7WqhD4kSZJGtWG5rE5m3gbcVtz/IfDa4diuJEnSWOW1C0eJI1uW8vLVS2tdRmmObIHK2T4kSWoMhqxR4qmuy71AtCRJdcRrF0qSJJXAkCVJklQCQ5YkSVIJDFmSJEklMGRJkiSVwJAlSZJUAkOWJElSCQxZkiRJJTBkSZIklcCQJUmSVAJDliRJUgm8duEoUs/X95ty+MRalyBJ0ogyZI0SI31x6BlLb6nrC1JLklRrHi6UJEkqgSFLkiSpBIYsSZKkEhiyJEmSSmDIkiRJKoEhS5IkqQSGLEmSpBIYsiRJkkpgyJIkSSqBIUuSJKkEhixJkqQSGLIkSZJKYMiSJEkqgSFLkiSpBIYsSZKkEhiyJEmSSmDIkiRJKoEhS5IkqQSGLEmSpBIYsiRJkkpgyJIkSSqBIUuSJKkEhixJkqQSGLIkSZJKYMiSJEkqgSFLkiSpBBNqXYCGJiIGv+4Vg1svMwfdpyRJjcKQNcYZeCRJGp0GfbgwIl4UEesjoisiNkXExUX70RGxNiIeLP5OHb5yJUmSxoahjMl6HrgkM1uA04A/iYiXAUuBdZl5MrCumJYkSWoogw5Zmbk1M+8p7j8FdAEnAO8CVheLrQbOHmqRkiRJY82w/LowImYArwLuAl6YmVuhEsSAFwxHH5IkSWPJkAe+R8QRwJeAP83Mnx3sr90i4kLgQoDp06cPtQxJdcxf0Uoai4a0JysiJlIJWNdn5r8UzT+JiOOK+ccBj/e1bmZek5lzMnPOtGnThlKGpDqXmSN+k6ShGsqvCwNYBXRl5oqqWTcD5xf3zwe+PPjyJOnQtbW10dzcTETQ3NxMW1tbrUuSBhQRg7o9fMU7Br2uyjWUPVmvB/4AeHNE3Fvc3g5cDrwlIh4E3lJMS9KIaGtro7Ozk46ODrq7u+no6KCzs9OgpVHPPbb1J0bDkzxnzpzcuHFjrcuQVAeam5vp6Ohg8eLF+9pWrFjBsmXL2L17dw0rk4ZXW1sb1157Lc8++yxNTU0sWrSIlStX1rqshhARd2fmnIGW89qFkurKs88+S2tr635tra2tPPvsszWqSBp+7rEdGwxZkupKU1MTnZ2d+7V1dnbS1NRUo4qk4XfttddyxRVXsHjxYiZNmsTixYu54ooruPbaa2tdmqp4uFBSXWlra+Pqq6/+pfaLLrrIQymqGxFBd3c3kyZN2te2a9cuJk+e7FirEeDhQkkN6fbbbwdg3Lhx+/3taZfqgXtsxwZDlqS68sADD3DWWWexd+9eMpO9e/dy1lln8cADD9S6NGnYLFq0iCVLlrBixQp27drFihUrWLJkCYsWLap1aari4UJJdaW/c/+Mhvc7abj468LaOdjDhYYsSXWlJ2SNHz+edevWMX/+fPbu3QsYsiQND8dkSWpoCxYsYNasWSxYsKDWpUhqUEO+QLQkjTYzZ87k5ptvpue6qDNnzuShhx6qcVWSGo0hS1LdeeSRR/Y7NDhhgm91kkaehwsl1Z29e/fudxHcnjFZUj1Zs2YNs2fPZvz48cyePZs1a9bUuiT14tc7SXVlwoQJPP/88322S/VizZo1tLe3s2rVKubOncuGDRtYuHAhAOeee26Nq1MP92RJqivPP/88U6dOJTP33aZOndpn8JLGquXLl7Nq1SrmzZvHxIkTmTdvHqtWrWL58uW1Lk1VDFmS6k7vs7t7tnfVm66uLubOnbtf29y5c+nq6qpRReqLIUtS3XnjG9/Y77Q01rW0tLBhw4b92jZs2EBLS0uNKlJfDFmS6sqECRPYsWPHfgPfd+zY4Zgs1ZX29nYWLlzI+vXr2bNnD+vXr2fhwoW0t7fXujRV8V1HUl1pbW3l6quv7rNdqhc9g9vb2tro6uqipaWF5cuXO+h9lPGyOpLqSnNzMx0dHSxevHhf24oVK1i2bBm7d++uYWWS6oXXLpTUkCKC7u5uJk2atK9t165dTJ482WsXShoWXrtQUkNqamqis7Nzv7bOzk6amppqVJGkRuWYLEl1ZdGiRSxZsgSojMPq7OxkyZIljsmSNOIMWZLqysqVKwFYtmwZl1xyCU1NTbS2tu5rl6SR4pgsSZKkQ+CYLEmSpBoyZEmSJJXAkCVJklQCQ5YkSVIJDFmSJEklMGRJkiSVwJAlSZJUAkOWJElSCQxZkiSNQW1tbTQ3NxMRNDc309bWVuuS1IshS5KkMaatrY3Ozk46Ojro7u6mo6ODzs5Og9Yo42V1JEkaY5qbm+no6GDx4sX72lasWMGyZcvYvXt3DStrDAd7WR1DliRJY0xE0N3dzaRJk/a17dq1i8mTJzMaPtfrndculCSpTjU1NdHZ2blfW2dnJ01NTTWqSH2ZUOsCJEnSoVm0aBFLliwBoLW1lc7OTpYsWUJra2uNK1M1Q5YkSWPMypUrAVi2bBmXXHIJTU1NtLa27mvX6OCYLEmSpEPgmCxJkqQaMmRJkiSVwJAlSZJUAkOWJElSCQxZkiRJJTBkSZIklcCQJUmSVIJRcZ6siNgGPFzrOhrMscBPa12EVDJf52oEvs5H3kmZOW2ghUZFyNLIi4iNB3MiNWks83WuRuDrfPTycKEkSVIJDFmSJEklMGQ1rmtqXYA0AnydqxH4Oh+lHJMlSZJUAvdkSZIklcCQJamhRMSdh7j8RyPig2XVI6l+GbLqSETcFhFzivtfjYijal2TVCsRMaHX9HiAzPyN2lQkVVQH94j4bEScU+uaBiMinq51DaOdIatOZebbM/PJWtchHaqImBER34uIT0fEdyPi+og4PSLuiIgHI+K1xe3OiPhO8fclxboXRMQXI+LfgFsj4k0RsT4iPg88UCzzdFVffx4R/xUR90fEZVXt7RHx/Yj4OvCSEX4KpJrp+TKi4WHIqrGD/ECZHBHXFR8G34mIdxXrHh4RXyg+IG4ADq/a7uaIOLa4/38j4u6I2BQRF1Yt83RELI+I+yLiWxHxwn7qfGdE3FX0//WeZSNiWkSsjYh7IuJTEfFwVb+/HxHfjoh7i3n+59XBejHwD8ArgJcCvwfMBT4ILAO+B7whM18F/CXQUbXu64DzM/PNxfRrgfbMfFl1BxFxBnByMf+VwGsi4g0R8RrgvcCrgN8GTi3lEaruRMQfFu/H90XEP0XESRGxrmhbFxHTB1j/NRFxe/F+/bWIOK5oP7XYxn9GxN9FxHeL9vHFdM8XhT/uZ9vjIuIfi8+BrxRHO84p5m2OiL+MiA3AuyNiUbHN+yLiSxExqVhuZlHDf0XEx4btiatjhqzRYaAPlHbgG5l5KjAP+LuImAy8H9iVma8AlgOvOcD2/ygzXwPMAT4QEccU7ZOBb2XmKcA3gUX91LgBOK34UPsC8BdF+6VFba8G/hWYDhARLcDvAq/PzFcCe4HzDv4pUYN7KDMfyMyfA5uAdVn5KfQDwAxgCvDF4sPm48CsqnXXZub2qulvZ+ZDffRxRnH7DnAPlf97JwO/CfxrZu7KzJ8BNw/vQ1M9iohZVN6r31y8p14MXA18rniPvh74RD/rTwRWAucU79fXUXlfB/gM0JqZr6PyXtpjIbCz+Gw4FVgUETMP0MVvU/m/83Lgf1P5MlJtd2bOzcwvAP+SmacWj6Or6Acqn1OfLPr7cb9PiACYMPAiGgEPZWbPoYx9HygR0fOBciJwVvxi8G0zlTDzBor/tJl5f0Tcf4DtfyAifqu4/yIqHyRPAM8BXyna7wbe0k+NJwI3FN+sDgN6PrTmAr9V1PAfEbGjaJ9PJfT9V0RAZS/b4wM8D1KPZ6vu/7xq+udU3rc+BqzPzN+KiBnAbVXLd/faVu/pHgH8TWZ+ar/GiD8FPLeNDtWbgZsy86cAmbk9Il5HJdwA/BPwt/2s/xJgNrC2eM8cD2yNytjaIzOz5wcbnwfeUdw/A3hF/GJM1xQq7+99famYC3yx+OLy44hY32v+DVX3Z0fEXwNHAUcAXyvaXw/8TtXjuaKfxyMMWaPFQB8oe4HfyczvV69U/Efs98MgIt4EnA68LjN3RcRtVEIawJ78xYnS9tL/62ElsCIzby62+dGeLg7UNbA6Mz/UX33SIE0BflTcv2CQ2/ga8LGIuD4zn46IE4A9VPbqfjYiLqfyf+KdwKf62Y4Elfe8gcJ5f/MD2FTsrfpFY8TUAdZpy8yv9bNM9bL9qf4y8lng7My8LyIuAN5UNc8vIIfAw4Vjw9eAtihSVUS8qmj/JsUhuIiYTeVwY29TgB1FwHopcNoga6j+UDu/qn0D8J6ihjOAnjeEdcA5EfGCYt7REXHSIPuWevtb4G8i4g4q3/gPWWbeSmWvwH8We41vorLH4B4q3+rvBb4E/L/hKVl1bh3wnp7hGBFxNHAnlfF9UHmv3tDP+t8HphV7v4iIiRExKzN3AE9FRM9793ur1vka8P7iUCMR8WvFUJK+bAB+pxib9UL2D069HUllL9pE9h/mcUevx6MBuCdrbPgYcBVwfxG0NlPZXfxJ4DPFYcJ7gW/3se5/AK3FMt8HvjXIGj5KZQzMj4pt9Bz3vwxYExG/C9wObAWeysyfRsSHqfzCaxyVPQR/Ajw8yP7VIDJzM5XDJj3TFxxg3q9VrfaRYv5nqXwL71n+NvY/lEhmHlF1/x+ojDPpXcNyfjEeRhpQZm6KiOXA7RGxl8pYvw8A10XEnwPbgPf1s/5zxWG/T0TEFCqfz1dRGZO4ELg2IrqpvJ53Fqt9msqQknuKz4ZtwNkH6OJLVIZxfBf4H+Cuqu309pFi/sNUxkEeWbRfDHw+Ii4utqcBeFkdDUlENAF7M/P54hvYJ4uB7pKkYRARR2Tm08X9pcBxmXnxYLdT7G37NpUfJjmAvUTuydJQTQduLPZWPUf/v1CUJB26BRHxISqf2Q8z+HGIXykG0h8GfMyAVT73ZGk/EdEOvLtX8xeLwyeSpFEsIl5O5Zd/1Z7NzF+FSpQSAAAAMUlEQVSvRT2NzpAlSZJUAn9dKEmSVAJDliRJUgkMWZIkSSUwZEmSJJXAkCVJklSC/w8EdSupQwofywAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[['median_age', 'married', 'college_grad']].plot.box(figsize=(10,5), title='neighborhood quality') # neighborhood quality"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1ed604a8>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[['median_school', 'num_schools']].plot.hist(figsize=(10,5), alpha=0.5, title='Schools')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1ef9f940>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.property_tax.plot.box(figsize=(5,8))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1f0bcef0>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.insurance.plot.box(figsize=(5,8))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.Feature Engineering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* One important missing feature is the age of the property!\n",
    "* Which is simply: *tx_year - year_built*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1880, 1892, 1900, 1905, 1918, 1920, 1921, 1924, 1925, 1927, 1929,\n",
       "       1930, 1932, 1935, 1936, 1937, 1938, 1939, 1940, 1941, 1942, 1943,\n",
       "       1944, 1945, 1946, 1947, 1948, 1949, 1950, 1951, 1952, 1953, 1954,\n",
       "       1955, 1956, 1957, 1958, 1959, 1960, 1961, 1962, 1963, 1964, 1965,\n",
       "       1966, 1967, 1968, 1969, 1970, 1971, 1972, 1973, 1974, 1975, 1976,\n",
       "       1977, 1978, 1979, 1980, 1981, 1982, 1983, 1984, 1985, 1986, 1987,\n",
       "       1988, 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998,\n",
       "       1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009,\n",
       "       2010, 2011, 2012, 2013, 2014, 2015])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.year_built.sort_values().unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['age'] = data.apply(lambda x: x['tx_year'] - x['year_built'], axis=1 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tx_price</th>\n",
       "      <th>beds</th>\n",
       "      <th>baths</th>\n",
       "      <th>sqft</th>\n",
       "      <th>year_built</th>\n",
       "      <th>lot_size</th>\n",
       "      <th>property_type</th>\n",
       "      <th>basement</th>\n",
       "      <th>restaurants</th>\n",
       "      <th>groceries</th>\n",
       "      <th>nightlife</th>\n",
       "      <th>cafes</th>\n",
       "      <th>shopping</th>\n",
       "      <th>arts_entertainment</th>\n",
       "      <th>beauty_spas</th>\n",
       "      <th>active_life</th>\n",
       "      <th>median_age</th>\n",
       "      <th>married</th>\n",
       "      <th>college_grad</th>\n",
       "      <th>property_tax</th>\n",
       "      <th>insurance</th>\n",
       "      <th>median_school</th>\n",
       "      <th>num_schools</th>\n",
       "      <th>tx_year</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>295850</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>584</td>\n",
       "      <td>2013</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>0.0</td>\n",
       "      <td>107</td>\n",
       "      <td>9</td>\n",
       "      <td>30</td>\n",
       "      <td>19</td>\n",
       "      <td>89</td>\n",
       "      <td>6</td>\n",
       "      <td>47</td>\n",
       "      <td>58</td>\n",
       "      <td>33.0</td>\n",
       "      <td>65.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>234.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2013</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>216500</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>612</td>\n",
       "      <td>1965</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>1.0</td>\n",
       "      <td>105</td>\n",
       "      <td>15</td>\n",
       "      <td>6</td>\n",
       "      <td>13</td>\n",
       "      <td>87</td>\n",
       "      <td>2</td>\n",
       "      <td>26</td>\n",
       "      <td>14</td>\n",
       "      <td>39.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>69.0</td>\n",
       "      <td>169.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2006</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>279900</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>615</td>\n",
       "      <td>1963</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>0.0</td>\n",
       "      <td>183</td>\n",
       "      <td>13</td>\n",
       "      <td>31</td>\n",
       "      <td>30</td>\n",
       "      <td>101</td>\n",
       "      <td>10</td>\n",
       "      <td>74</td>\n",
       "      <td>62</td>\n",
       "      <td>28.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>216.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2012</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>379900</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>618</td>\n",
       "      <td>2000</td>\n",
       "      <td>33541</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>0.0</td>\n",
       "      <td>198</td>\n",
       "      <td>9</td>\n",
       "      <td>38</td>\n",
       "      <td>25</td>\n",
       "      <td>127</td>\n",
       "      <td>11</td>\n",
       "      <td>72</td>\n",
       "      <td>83</td>\n",
       "      <td>36.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>91.0</td>\n",
       "      <td>265.0</td>\n",
       "      <td>92.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2005</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>340000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>634</td>\n",
       "      <td>1992</td>\n",
       "      <td>0</td>\n",
       "      <td>Apartment / Condo / Townhouse</td>\n",
       "      <td>0.0</td>\n",
       "      <td>149</td>\n",
       "      <td>7</td>\n",
       "      <td>22</td>\n",
       "      <td>20</td>\n",
       "      <td>83</td>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>73</td>\n",
       "      <td>37.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2002</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   tx_price  beds  baths  sqft  year_built  lot_size  \\\n",
       "0    295850     1      1   584        2013         0   \n",
       "1    216500     1      1   612        1965         0   \n",
       "2    279900     1      1   615        1963         0   \n",
       "3    379900     1      1   618        2000     33541   \n",
       "4    340000     1      1   634        1992         0   \n",
       "\n",
       "                   property_type  basement  restaurants  groceries  nightlife  \\\n",
       "0  Apartment / Condo / Townhouse       0.0          107          9         30   \n",
       "1  Apartment / Condo / Townhouse       1.0          105         15          6   \n",
       "2  Apartment / Condo / Townhouse       0.0          183         13         31   \n",
       "3  Apartment / Condo / Townhouse       0.0          198          9         38   \n",
       "4  Apartment / Condo / Townhouse       0.0          149          7         22   \n",
       "\n",
       "   cafes  shopping  arts_entertainment  beauty_spas  active_life  median_age  \\\n",
       "0     19        89                   6           47           58        33.0   \n",
       "1     13        87                   2           26           14        39.0   \n",
       "2     30       101                  10           74           62        28.0   \n",
       "3     25       127                  11           72           83        36.0   \n",
       "4     20        83                  10           50           73        37.0   \n",
       "\n",
       "   married  college_grad  property_tax  insurance  median_school  num_schools  \\\n",
       "0     65.0          84.0         234.0       81.0            9.0          3.0   \n",
       "1     73.0          69.0         169.0       51.0            3.0          3.0   \n",
       "2     15.0          86.0         216.0       74.0            8.0          3.0   \n",
       "3     25.0          91.0         265.0       92.0            9.0          3.0   \n",
       "4     20.0          75.0          88.0       30.0            9.0          3.0   \n",
       "\n",
       "   tx_year  age  \n",
       "0     2013    0  \n",
       "1     2006   41  \n",
       "2     2012   49  \n",
       "3     2005    5  \n",
       "4     2002   10  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ -8,  -7,  -4,  -2,  -1,   0,   1,   2,   3,   4,   5,   6,   7,\n",
       "         8,   9,  10,  11,  12,  13,  14,  15,  16,  17,  18,  19,  20,\n",
       "        21,  22,  23,  24,  25,  26,  27,  28,  29,  30,  31,  32,  33,\n",
       "        34,  35,  36,  37,  38,  39,  40,  41,  42,  43,  44,  45,  46,\n",
       "        47,  48,  49,  50,  51,  52,  53,  54,  55,  56,  57,  58,  59,\n",
       "        60,  61,  62,  63,  64,  65,  66,  67,  68,  69,  70,  71,  72,\n",
       "        73,  74,  75,  77,  78,  79,  80,  82,  85,  86,  89,  90,  94,\n",
       "       101, 103, 104, 110, 113, 114])"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.age.sort_values().unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>year_built</th>\n",
       "      <th>tx_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1540</th>\n",
       "      <td>2015</td>\n",
       "      <td>2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1864</th>\n",
       "      <td>2015</td>\n",
       "      <td>2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1828</th>\n",
       "      <td>2015</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1840</th>\n",
       "      <td>2014</td>\n",
       "      <td>2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1207</th>\n",
       "      <td>2014</td>\n",
       "      <td>2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1769</th>\n",
       "      <td>2014</td>\n",
       "      <td>2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2013</td>\n",
       "      <td>2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1818</th>\n",
       "      <td>2013</td>\n",
       "      <td>2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>2013</td>\n",
       "      <td>2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1809</th>\n",
       "      <td>2013</td>\n",
       "      <td>2012</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      year_built  tx_year\n",
       "1540        2015     2014\n",
       "1864        2015     2013\n",
       "1828        2015     2007\n",
       "1840        2014     2014\n",
       "1207        2014     2014\n",
       "1769        2014     2014\n",
       "0           2013     2013\n",
       "1818        2013     2013\n",
       "111         2013     2013\n",
       "1809        2013     2012"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.iloc[:,4::19].sort_values(by='year_built', ascending=False)[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some properties are sold before it finished construction.\n",
    "\n",
    "That is why we have negative *age* values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop('property_type', 1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.to_csv('data_cooked.csv', index=False) # Now, output to cleaned & engineered table for use"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.Modelling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('data_cooked.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tx_price</th>\n",
       "      <th>beds</th>\n",
       "      <th>baths</th>\n",
       "      <th>sqft</th>\n",
       "      <th>year_built</th>\n",
       "      <th>lot_size</th>\n",
       "      <th>basement</th>\n",
       "      <th>restaurants</th>\n",
       "      <th>groceries</th>\n",
       "      <th>nightlife</th>\n",
       "      <th>cafes</th>\n",
       "      <th>shopping</th>\n",
       "      <th>arts_entertainment</th>\n",
       "      <th>beauty_spas</th>\n",
       "      <th>active_life</th>\n",
       "      <th>median_age</th>\n",
       "      <th>married</th>\n",
       "      <th>college_grad</th>\n",
       "      <th>property_tax</th>\n",
       "      <th>insurance</th>\n",
       "      <th>median_school</th>\n",
       "      <th>num_schools</th>\n",
       "      <th>tx_year</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>295850</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>584</td>\n",
       "      <td>2013</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>107</td>\n",
       "      <td>9</td>\n",
       "      <td>30</td>\n",
       "      <td>19</td>\n",
       "      <td>89</td>\n",
       "      <td>6</td>\n",
       "      <td>47</td>\n",
       "      <td>58</td>\n",
       "      <td>33.0</td>\n",
       "      <td>65.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>234.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2013</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>216500</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>612</td>\n",
       "      <td>1965</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>105</td>\n",
       "      <td>15</td>\n",
       "      <td>6</td>\n",
       "      <td>13</td>\n",
       "      <td>87</td>\n",
       "      <td>2</td>\n",
       "      <td>26</td>\n",
       "      <td>14</td>\n",
       "      <td>39.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>69.0</td>\n",
       "      <td>169.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2006</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>279900</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>615</td>\n",
       "      <td>1963</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>183</td>\n",
       "      <td>13</td>\n",
       "      <td>31</td>\n",
       "      <td>30</td>\n",
       "      <td>101</td>\n",
       "      <td>10</td>\n",
       "      <td>74</td>\n",
       "      <td>62</td>\n",
       "      <td>28.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>216.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2012</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>379900</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>618</td>\n",
       "      <td>2000</td>\n",
       "      <td>33541</td>\n",
       "      <td>0.0</td>\n",
       "      <td>198</td>\n",
       "      <td>9</td>\n",
       "      <td>38</td>\n",
       "      <td>25</td>\n",
       "      <td>127</td>\n",
       "      <td>11</td>\n",
       "      <td>72</td>\n",
       "      <td>83</td>\n",
       "      <td>36.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>91.0</td>\n",
       "      <td>265.0</td>\n",
       "      <td>92.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2005</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>340000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>634</td>\n",
       "      <td>1992</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>149</td>\n",
       "      <td>7</td>\n",
       "      <td>22</td>\n",
       "      <td>20</td>\n",
       "      <td>83</td>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>73</td>\n",
       "      <td>37.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2002</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   tx_price  beds  baths  sqft  year_built  lot_size  basement  restaurants  \\\n",
       "0    295850     1      1   584        2013         0       0.0          107   \n",
       "1    216500     1      1   612        1965         0       1.0          105   \n",
       "2    279900     1      1   615        1963         0       0.0          183   \n",
       "3    379900     1      1   618        2000     33541       0.0          198   \n",
       "4    340000     1      1   634        1992         0       0.0          149   \n",
       "\n",
       "   groceries  nightlife  cafes  shopping  arts_entertainment  beauty_spas  \\\n",
       "0          9         30     19        89                   6           47   \n",
       "1         15          6     13        87                   2           26   \n",
       "2         13         31     30       101                  10           74   \n",
       "3          9         38     25       127                  11           72   \n",
       "4          7         22     20        83                  10           50   \n",
       "\n",
       "   active_life  median_age  married  college_grad  property_tax  insurance  \\\n",
       "0           58        33.0     65.0          84.0         234.0       81.0   \n",
       "1           14        39.0     73.0          69.0         169.0       51.0   \n",
       "2           62        28.0     15.0          86.0         216.0       74.0   \n",
       "3           83        36.0     25.0          91.0         265.0       92.0   \n",
       "4           73        37.0     20.0          75.0          88.0       30.0   \n",
       "\n",
       "   median_school  num_schools  tx_year  age  \n",
       "0            9.0          3.0     2013    0  \n",
       "1            3.0          3.0     2006   41  \n",
       "2            8.0          3.0     2012   49  \n",
       "3            9.0          3.0     2005    5  \n",
       "4            9.0          3.0     2002   10  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_X, test_X, train_y, test_y = train_test_split(df.drop('tx_price',1), df.tx_price, random_state=99,\n",
    "                                                   test_size=0.3, )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train_X.shape: (1318, 23) \n",
      "train_y.shape: (1318,) \n",
      "test_X.shape: (565, 23) \n",
      "test_y.shape: (565,)\n"
     ]
    }
   ],
   "source": [
    "print('train_X.shape:', train_X.shape,\n",
    "     '\\ntrain_y.shape:', train_y.shape,\n",
    "     '\\ntest_X.shape:', test_X.shape,\n",
    "     '\\ntest_y.shape:', test_y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "rrr = RandomForestRegressor(n_estimators=100)\n",
    "rrr = rrr.fit(train_X, train_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "46052.933486725669"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat_rrr = rrr.predict(test_X)\n",
    "mean_absolute_error(test_y, yhat_rrr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Tuning number of estimators via gridsearch:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_mae_rf(num_est, predictors_train, predictors_val, targ_train, targ_val):\n",
    "\n",
    "    # fitting model with input max_leaf_nodes\n",
    "    model = RandomForestRegressor(n_estimators=num_est, random_state=0)\n",
    "\n",
    "    # fitting the model with training dataset\n",
    "    model.fit(predictors_train, targ_train)\n",
    "\n",
    "    # making prediction with the test dataset\n",
    "    preds_val = model.predict(predictors_val)\n",
    "\n",
    "    # calculate and return the MAE\n",
    "    mae = mean_absolute_error(targ_val, preds_val)\n",
    "    return(mae)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_est : 2 \t\t Mean Absolute Error:  58775\n",
      "num_est : 3 \t\t Mean Absolute Error:  54786\n",
      "num_est : 4 \t\t Mean Absolute Error:  51000\n",
      "num_est : 5 \t\t Mean Absolute Error:  49162\n",
      "num_est : 6 \t\t Mean Absolute Error:  47648\n",
      "num_est : 7 \t\t Mean Absolute Error:  47152\n",
      "num_est : 8 \t\t Mean Absolute Error:  47404\n",
      "num_est : 9 \t\t Mean Absolute Error:  47464\n",
      "num_est : 10 \t\t Mean Absolute Error:  47479\n",
      "num_est : 11 \t\t Mean Absolute Error:  47503\n",
      "num_est : 12 \t\t Mean Absolute Error:  47329\n",
      "num_est : 13 \t\t Mean Absolute Error:  47571\n",
      "num_est : 14 \t\t Mean Absolute Error:  47755\n",
      "num_est : 15 \t\t Mean Absolute Error:  47719\n",
      "num_est : 16 \t\t Mean Absolute Error:  47538\n",
      "num_est : 17 \t\t Mean Absolute Error:  47702\n",
      "num_est : 18 \t\t Mean Absolute Error:  47351\n",
      "num_est : 19 \t\t Mean Absolute Error:  47075\n",
      "num_est : 20 \t\t Mean Absolute Error:  47104\n",
      "num_est : 21 \t\t Mean Absolute Error:  47142\n",
      "num_est : 22 \t\t Mean Absolute Error:  46920\n",
      "num_est : 23 \t\t Mean Absolute Error:  46916\n",
      "num_est : 24 \t\t Mean Absolute Error:  46993\n",
      "num_est : 25 \t\t Mean Absolute Error:  47145\n",
      "num_est : 26 \t\t Mean Absolute Error:  47148\n",
      "num_est : 27 \t\t Mean Absolute Error:  47146\n",
      "num_est : 28 \t\t Mean Absolute Error:  47114\n",
      "num_est : 29 \t\t Mean Absolute Error:  46932\n",
      "num_est : 30 \t\t Mean Absolute Error:  46762\n",
      "num_est : 31 \t\t Mean Absolute Error:  46806\n",
      "num_est : 32 \t\t Mean Absolute Error:  46658\n",
      "num_est : 33 \t\t Mean Absolute Error:  46618\n",
      "num_est : 34 \t\t Mean Absolute Error:  46647\n",
      "num_est : 35 \t\t Mean Absolute Error:  46628\n",
      "num_est : 36 \t\t Mean Absolute Error:  46571\n",
      "num_est : 37 \t\t Mean Absolute Error:  46608\n",
      "num_est : 38 \t\t Mean Absolute Error:  46537\n",
      "num_est : 39 \t\t Mean Absolute Error:  46427\n",
      "num_est : 40 \t\t Mean Absolute Error:  46378\n",
      "num_est : 41 \t\t Mean Absolute Error:  46398\n",
      "num_est : 42 \t\t Mean Absolute Error:  46348\n",
      "num_est : 43 \t\t Mean Absolute Error:  46414\n",
      "num_est : 44 \t\t Mean Absolute Error:  46350\n",
      "num_est : 45 \t\t Mean Absolute Error:  46373\n",
      "num_est : 46 \t\t Mean Absolute Error:  46339\n",
      "num_est : 47 \t\t Mean Absolute Error:  46452\n",
      "num_est : 48 \t\t Mean Absolute Error:  46417\n",
      "num_est : 49 \t\t Mean Absolute Error:  46452\n",
      "num_est : 50 \t\t Mean Absolute Error:  46471\n",
      "num_est : 51 \t\t Mean Absolute Error:  46459\n",
      "num_est : 52 \t\t Mean Absolute Error:  46394\n",
      "num_est : 53 \t\t Mean Absolute Error:  46470\n",
      "num_est : 54 \t\t Mean Absolute Error:  46411\n",
      "num_est : 55 \t\t Mean Absolute Error:  46501\n",
      "num_est : 56 \t\t Mean Absolute Error:  46521\n",
      "num_est : 57 \t\t Mean Absolute Error:  46561\n",
      "num_est : 58 \t\t Mean Absolute Error:  46565\n",
      "num_est : 59 \t\t Mean Absolute Error:  46606\n",
      "num_est : 60 \t\t Mean Absolute Error:  46630\n",
      "num_est : 61 \t\t Mean Absolute Error:  46553\n",
      "num_est : 62 \t\t Mean Absolute Error:  46506\n",
      "num_est : 63 \t\t Mean Absolute Error:  46591\n",
      "num_est : 64 \t\t Mean Absolute Error:  46635\n",
      "num_est : 65 \t\t Mean Absolute Error:  46702\n",
      "num_est : 66 \t\t Mean Absolute Error:  46730\n",
      "num_est : 67 \t\t Mean Absolute Error:  46759\n",
      "num_est : 68 \t\t Mean Absolute Error:  46790\n",
      "num_est : 69 \t\t Mean Absolute Error:  46770\n",
      "num_est : 70 \t\t Mean Absolute Error:  46793\n",
      "num_est : 71 \t\t Mean Absolute Error:  46779\n",
      "num_est : 72 \t\t Mean Absolute Error:  46760\n",
      "num_est : 73 \t\t Mean Absolute Error:  46787\n",
      "num_est : 74 \t\t Mean Absolute Error:  46689\n",
      "num_est : 75 \t\t Mean Absolute Error:  46650\n",
      "num_est : 76 \t\t Mean Absolute Error:  46664\n",
      "num_est : 77 \t\t Mean Absolute Error:  46617\n",
      "num_est : 78 \t\t Mean Absolute Error:  46489\n",
      "num_est : 79 \t\t Mean Absolute Error:  46512\n",
      "num_est : 80 \t\t Mean Absolute Error:  46481\n",
      "num_est : 81 \t\t Mean Absolute Error:  46490\n",
      "num_est : 82 \t\t Mean Absolute Error:  46578\n",
      "num_est : 83 \t\t Mean Absolute Error:  46646\n",
      "num_est : 84 \t\t Mean Absolute Error:  46691\n",
      "num_est : 85 \t\t Mean Absolute Error:  46662\n",
      "num_est : 86 \t\t Mean Absolute Error:  46681\n",
      "num_est : 87 \t\t Mean Absolute Error:  46726\n",
      "num_est : 88 \t\t Mean Absolute Error:  46722\n",
      "num_est : 89 \t\t Mean Absolute Error:  46722\n",
      "num_est : 90 \t\t Mean Absolute Error:  46672\n",
      "num_est : 91 \t\t Mean Absolute Error:  46692\n",
      "num_est : 92 \t\t Mean Absolute Error:  46643\n",
      "num_est : 93 \t\t Mean Absolute Error:  46654\n",
      "num_est : 94 \t\t Mean Absolute Error:  46669\n",
      "num_est : 95 \t\t Mean Absolute Error:  46685\n",
      "num_est : 96 \t\t Mean Absolute Error:  46659\n",
      "num_est : 97 \t\t Mean Absolute Error:  46668\n",
      "num_est : 98 \t\t Mean Absolute Error:  46659\n",
      "num_est : 99 \t\t Mean Absolute Error:  46601\n",
      "46.106316\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.clock()\n",
    "\n",
    "plot_mae = []\n",
    "estimator = range(2,100)\n",
    "\n",
    "for num_est in range(2,100):\n",
    "    my_mae = get_mae_rf(num_est,train_X,test_X,train_y,test_y)\n",
    "    print(\"num_est : %d \\t\\t Mean Absolute Error:  %d\" %(num_est,my_mae))\n",
    "    plot_mae.append(my_mae)\n",
    "\n",
    "print (time.clock() - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_est : 340 \t\t Mean Absolute Error:  45897\n",
      "num_est : 341 \t\t Mean Absolute Error:  45891\n",
      "num_est : 342 \t\t Mean Absolute Error:  45873\n",
      "num_est : 343 \t\t Mean Absolute Error:  45870\n",
      "num_est : 344 \t\t Mean Absolute Error:  45868\n",
      "num_est : 345 \t\t Mean Absolute Error:  45865\n",
      "num_est : 346 \t\t Mean Absolute Error:  45866\n",
      "num_est : 347 \t\t Mean Absolute Error:  45868\n",
      "num_est : 348 \t\t Mean Absolute Error:  45884\n",
      "num_est : 349 \t\t Mean Absolute Error:  45877\n",
      "31.643173999999995\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.clock()\n",
    "\n",
    "plot_mae = []\n",
    "estimator = range(2,100)\n",
    "\n",
    "for num_est in range(340,350):\n",
    "    my_mae = get_mae_rf(num_est,train_X,test_X,train_y,test_y)\n",
    "    print(\"num_est : %d \\t\\t Mean Absolute Error:  %d\" %(num_est,my_mae))\n",
    "    plot_mae.append(my_mae)\n",
    "\n",
    "print (time.clock() - start)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even if we are increasing the number of estimators, it seems that it is not getting any better then the first trial.\n",
    "\n",
    "At aounrd 350 round, it starts to get better results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "features = train_X.columns\n",
    "importances = rrr.feature_importances_\n",
    "indices = np.argsort(importances)\n",
    "\n",
    "plt.title('Feature Importances - Random Forest')\n",
    "plt.barh(range(len(indices)), importances[indices], color='b', align='center')\n",
    "plt.yticks(range(len(indices)), features[indices])\n",
    "plt.xlabel('Relative Importance')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import XGBRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgbr = XGBRegressor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "       colsample_bytree=1, gamma=0, learning_rate=0.1, max_delta_step=0,\n",
       "       max_depth=3, min_child_weight=1, missing=None, n_estimators=100,\n",
       "       n_jobs=1, nthread=None, objective='reg:linear', random_state=0,\n",
       "       reg_alpha=0, reg_lambda=1, scale_pos_weight=1, seed=None,\n",
       "       silent=True, subsample=1)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgbr.fit(train_X, train_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.03768116,  0.00724638,  0.12028985,  0.06086956,  0.03043478,\n",
       "        0.00289855,  0.01449275,  0.01594203,  0.0057971 ,  0.01884058,\n",
       "        0.01304348,  0.01594203,  0.01304348,  0.01594203,  0.01304348,\n",
       "        0.02173913,  0.03188406,  0.0768116 ,  0.17536232,  0.0173913 ,\n",
       "        0.        ,  0.2173913 ,  0.07391305], dtype=float32)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgbr.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['beds', 'baths', 'sqft', 'year_built', 'lot_size', 'basement',\n",
       "       'restaurants', 'groceries', 'nightlife', 'cafes', 'shopping',\n",
       "       'arts_entertainment', 'beauty_spas', 'active_life', 'median_age',\n",
       "       'married', 'college_grad', 'property_tax', 'insurance', 'median_school',\n",
       "       'num_schools', 'tx_year', 'age'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_X.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for y,x in sorted(zip(xgbr.feature_importances_, train_X.columns), reverse=False):\n",
    "    #print (x,y)\n",
    "    plt.barh(x,y, color='b')\n",
    "    plt.title('Feature Importances - XGB')\n",
    "    plt.xlabel('Relative Importance')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45025.599529867257"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat_xgb = xgbr.predict(test_X)\n",
    "mean_absolute_error(test_y, yhat_xgb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3LightGBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "import lightgbm as lgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "#lgr = lgb.LGBMRegressor(num_leaves=10, n_estimators=150, learning_rate=0.3, max_depth=3, min_child_samples=1)\n",
    "lgr = lgb.LGBMRegressor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\tvalid_0's l2: 2.41605e+10\n",
      "[2]\tvalid_0's l2: 2.41369e+10\n",
      "[3]\tvalid_0's l2: 2.41134e+10\n",
      "[4]\tvalid_0's l2: 2.40895e+10\n",
      "[5]\tvalid_0's l2: 2.4066e+10\n",
      "[6]\tvalid_0's l2: 2.40422e+10\n",
      "[7]\tvalid_0's l2: 2.40184e+10\n",
      "[8]\tvalid_0's l2: 2.3995e+10\n",
      "[9]\tvalid_0's l2: 2.39712e+10\n",
      "[10]\tvalid_0's l2: 2.39475e+10\n",
      "[11]\tvalid_0's l2: 2.39238e+10\n",
      "[12]\tvalid_0's l2: 2.39004e+10\n",
      "[13]\tvalid_0's l2: 2.38767e+10\n",
      "[14]\tvalid_0's l2: 2.38531e+10\n",
      "[15]\tvalid_0's l2: 2.38298e+10\n",
      "[16]\tvalid_0's l2: 2.38062e+10\n",
      "[17]\tvalid_0's l2: 2.37826e+10\n",
      "[18]\tvalid_0's l2: 2.37593e+10\n",
      "[19]\tvalid_0's l2: 2.37358e+10\n",
      "[20]\tvalid_0's l2: 2.37123e+10\n",
      "[21]\tvalid_0's l2: 2.36891e+10\n",
      "[22]\tvalid_0's l2: 2.36656e+10\n",
      "[23]\tvalid_0's l2: 2.36421e+10\n",
      "[24]\tvalid_0's l2: 2.36187e+10\n",
      "[25]\tvalid_0's l2: 2.3595e+10\n",
      "[26]\tvalid_0's l2: 2.35714e+10\n",
      "[27]\tvalid_0's l2: 2.35483e+10\n",
      "[28]\tvalid_0's l2: 2.35247e+10\n",
      "[29]\tvalid_0's l2: 2.35016e+10\n",
      "[30]\tvalid_0's l2: 2.34781e+10\n",
      "[31]\tvalid_0's l2: 2.34551e+10\n",
      "[32]\tvalid_0's l2: 2.34316e+10\n",
      "[33]\tvalid_0's l2: 2.34083e+10\n",
      "[34]\tvalid_0's l2: 2.33848e+10\n",
      "[35]\tvalid_0's l2: 2.33613e+10\n",
      "[36]\tvalid_0's l2: 2.33384e+10\n",
      "[37]\tvalid_0's l2: 2.3315e+10\n",
      "[38]\tvalid_0's l2: 2.32921e+10\n",
      "[39]\tvalid_0's l2: 2.32687e+10\n",
      "[40]\tvalid_0's l2: 2.32455e+10\n",
      "[41]\tvalid_0's l2: 2.32222e+10\n",
      "[42]\tvalid_0's l2: 2.31988e+10\n",
      "[43]\tvalid_0's l2: 2.31761e+10\n",
      "[44]\tvalid_0's l2: 2.31528e+10\n",
      "[45]\tvalid_0's l2: 2.313e+10\n",
      "[46]\tvalid_0's l2: 2.31068e+10\n",
      "[47]\tvalid_0's l2: 2.30841e+10\n",
      "[48]\tvalid_0's l2: 2.30608e+10\n",
      "[49]\tvalid_0's l2: 2.30376e+10\n",
      "[50]\tvalid_0's l2: 2.3015e+10\n",
      "[51]\tvalid_0's l2: 2.29918e+10\n",
      "[52]\tvalid_0's l2: 2.29689e+10\n",
      "[53]\tvalid_0's l2: 2.29457e+10\n",
      "[54]\tvalid_0's l2: 2.29232e+10\n",
      "[55]\tvalid_0's l2: 2.29001e+10\n",
      "[56]\tvalid_0's l2: 2.28771e+10\n",
      "[57]\tvalid_0's l2: 2.28547e+10\n",
      "[58]\tvalid_0's l2: 2.28318e+10\n",
      "[59]\tvalid_0's l2: 2.28088e+10\n",
      "[60]\tvalid_0's l2: 2.27859e+10\n",
      "[61]\tvalid_0's l2: 2.2763e+10\n",
      "[62]\tvalid_0's l2: 2.27407e+10\n",
      "[63]\tvalid_0's l2: 2.27179e+10\n",
      "[64]\tvalid_0's l2: 2.26951e+10\n",
      "[65]\tvalid_0's l2: 2.26723e+10\n",
      "[66]\tvalid_0's l2: 2.26495e+10\n",
      "[67]\tvalid_0's l2: 2.26267e+10\n",
      "[68]\tvalid_0's l2: 2.26046e+10\n",
      "[69]\tvalid_0's l2: 2.25819e+10\n",
      "[70]\tvalid_0's l2: 2.25592e+10\n",
      "[71]\tvalid_0's l2: 2.25365e+10\n",
      "[72]\tvalid_0's l2: 2.25139e+10\n",
      "[73]\tvalid_0's l2: 2.24912e+10\n",
      "[74]\tvalid_0's l2: 2.24686e+10\n",
      "[75]\tvalid_0's l2: 2.2446e+10\n",
      "[76]\tvalid_0's l2: 2.24235e+10\n",
      "[77]\tvalid_0's l2: 2.24009e+10\n",
      "[78]\tvalid_0's l2: 2.23784e+10\n",
      "[79]\tvalid_0's l2: 2.23565e+10\n",
      "[80]\tvalid_0's l2: 2.2334e+10\n",
      "[81]\tvalid_0's l2: 2.23115e+10\n",
      "[82]\tvalid_0's l2: 2.22891e+10\n",
      "[83]\tvalid_0's l2: 2.22666e+10\n",
      "[84]\tvalid_0's l2: 2.22442e+10\n",
      "[85]\tvalid_0's l2: 2.22218e+10\n",
      "[86]\tvalid_0's l2: 2.21994e+10\n",
      "[87]\tvalid_0's l2: 2.21771e+10\n",
      "[88]\tvalid_0's l2: 2.21547e+10\n",
      "[89]\tvalid_0's l2: 2.21324e+10\n",
      "[90]\tvalid_0's l2: 2.211e+10\n",
      "[91]\tvalid_0's l2: 2.20883e+10\n",
      "[92]\tvalid_0's l2: 2.20658e+10\n",
      "[93]\tvalid_0's l2: 2.20435e+10\n",
      "[94]\tvalid_0's l2: 2.20218e+10\n",
      "[95]\tvalid_0's l2: 2.19994e+10\n",
      "[96]\tvalid_0's l2: 2.19771e+10\n",
      "[97]\tvalid_0's l2: 2.19548e+10\n",
      "[98]\tvalid_0's l2: 2.19332e+10\n",
      "[99]\tvalid_0's l2: 2.19109e+10\n",
      "[100]\tvalid_0's l2: 2.18887e+10\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LGBMRegressor(boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,\n",
       "       learning_rate=0.1, max_depth=-1, min_child_samples=20,\n",
       "       min_child_weight=0.001, min_split_gain=0.0, n_estimators=100,\n",
       "       n_jobs=-1, num_leaves=31, objective=None, random_state=None,\n",
       "       reg_alpha=0.0, reg_lambda=0.0, silent=True, subsample=1.0,\n",
       "       subsample_for_bin=200000, subsample_freq=1)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgr.fit(train_X, train_y, eval_set=(test_X, test_y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "121103.29577598203"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat_lgr = lgr.predict(test_X)\n",
    "mean_absolute_error(test_y, yhat_lgr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hmm the mae is bit too high, let's do some manual tuning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LGBMRegressor(boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,\n",
       "       learning_rate=0.1, max_depth=-1, min_child_samples=20,\n",
       "       min_child_weight=0.001, min_split_gain=0.0, n_estimators=10000,\n",
       "       n_jobs=-1, num_leaves=31, objective=None, random_state=None,\n",
       "       reg_alpha=0.0, reg_lambda=0.0, silent=True, subsample=1.0,\n",
       "       subsample_for_bin=200000, subsample_freq=1)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgr = lgb.LGBMRegressor(n_estimators=10000)\n",
    "lgr.fit(train_X, train_y, eval_set=(test_X, test_y), verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "48051.45830198504"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat_lgr = lgr.predict(test_X)\n",
    "mean_absolute_error(test_y, yhat_lgr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for y,x in sorted(zip(lgr.feature_importances_, train_X.columns), reverse=False):\n",
    "    #print (x,y)\n",
    "    plt.barh(x,y, color='b')\n",
    "    plt.title('Feature Importances - Lightgbm')\n",
    "    plt.xlabel('Relative Importance')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3CatBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "from catboost import CatBoostRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "catr = CatBoostRegressor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 437495.0549052\ttest: 437723.3685829\tbest: 437723.3685829 (0)\ttotal: 63.2ms\tremaining: 1m 3s\n",
      "1:\tlearn: 425992.6066846\ttest: 426501.5594061\tbest: 426501.5594061 (1)\ttotal: 68.7ms\tremaining: 34.3s\n",
      "2:\tlearn: 414378.3162013\ttest: 415015.2880961\tbest: 415015.2880961 (2)\ttotal: 73.1ms\tremaining: 24.3s\n",
      "3:\tlearn: 403370.6254951\ttest: 404239.3898016\tbest: 404239.3898016 (3)\ttotal: 76.6ms\tremaining: 19.1s\n",
      "4:\tlearn: 392870.0151548\ttest: 393944.8265820\tbest: 393944.8265820 (4)\ttotal: 79.9ms\tremaining: 15.9s\n",
      "5:\tlearn: 382755.2764853\ttest: 384110.6477866\tbest: 384110.6477866 (5)\ttotal: 83.1ms\tremaining: 13.8s\n",
      "6:\tlearn: 372777.3817192\ttest: 374242.8287849\tbest: 374242.8287849 (6)\ttotal: 87ms\tremaining: 12.3s\n",
      "7:\tlearn: 363170.1680121\ttest: 364913.2769081\tbest: 364913.2769081 (7)\ttotal: 90.6ms\tremaining: 11.2s\n",
      "8:\tlearn: 353497.1600502\ttest: 355466.7124258\tbest: 355466.7124258 (8)\ttotal: 93.8ms\tremaining: 10.3s\n",
      "9:\tlearn: 344367.3405380\ttest: 346493.2402822\tbest: 346493.2402822 (9)\ttotal: 97ms\tremaining: 9.6s\n",
      "10:\tlearn: 335311.8672878\ttest: 337502.8034486\tbest: 337502.8034486 (10)\ttotal: 100ms\tremaining: 8.99s\n",
      "11:\tlearn: 326772.1367440\ttest: 329093.7410221\tbest: 329093.7410221 (11)\ttotal: 103ms\tremaining: 8.51s\n",
      "12:\tlearn: 318380.2667128\ttest: 320829.8055906\tbest: 320829.8055906 (12)\ttotal: 107ms\tremaining: 8.09s\n",
      "13:\tlearn: 310413.8832240\ttest: 313136.8729255\tbest: 313136.8729255 (13)\ttotal: 109ms\tremaining: 7.71s\n",
      "14:\tlearn: 302789.4178631\ttest: 305704.5817278\tbest: 305704.5817278 (14)\ttotal: 113ms\tremaining: 7.43s\n",
      "15:\tlearn: 294830.3204865\ttest: 298008.9786510\tbest: 298008.9786510 (15)\ttotal: 116ms\tremaining: 7.14s\n",
      "16:\tlearn: 287647.4389350\ttest: 290958.4617728\tbest: 290958.4617728 (16)\ttotal: 119ms\tremaining: 6.88s\n",
      "17:\tlearn: 280627.8981284\ttest: 284078.4590156\tbest: 284078.4590156 (17)\ttotal: 122ms\tremaining: 6.68s\n",
      "18:\tlearn: 273919.3327758\ttest: 277544.5348302\tbest: 277544.5348302 (18)\ttotal: 126ms\tremaining: 6.5s\n",
      "19:\tlearn: 267501.1286193\ttest: 271367.8410656\tbest: 271367.8410656 (19)\ttotal: 129ms\tremaining: 6.33s\n",
      "20:\tlearn: 261013.9524776\ttest: 264926.8019863\tbest: 264926.8019863 (20)\ttotal: 133ms\tremaining: 6.2s\n",
      "21:\tlearn: 254950.6573816\ttest: 259018.6865437\tbest: 259018.6865437 (21)\ttotal: 136ms\tremaining: 6.05s\n",
      "22:\tlearn: 248974.4985684\ttest: 253174.7941712\tbest: 253174.7941712 (22)\ttotal: 139ms\tremaining: 5.92s\n",
      "23:\tlearn: 242718.5415252\ttest: 246921.6183552\tbest: 246921.6183552 (23)\ttotal: 143ms\tremaining: 5.81s\n",
      "24:\tlearn: 236988.1258716\ttest: 241168.6328898\tbest: 241168.6328898 (24)\ttotal: 146ms\tremaining: 5.7s\n",
      "25:\tlearn: 231609.0934696\ttest: 235795.7594498\tbest: 235795.7594498 (25)\ttotal: 150ms\tremaining: 5.61s\n",
      "26:\tlearn: 226650.6694000\ttest: 230886.5683830\tbest: 230886.5683830 (26)\ttotal: 153ms\tremaining: 5.51s\n",
      "27:\tlearn: 221575.7705571\ttest: 225888.9669080\tbest: 225888.9669080 (27)\ttotal: 156ms\tremaining: 5.42s\n",
      "28:\tlearn: 216680.1025362\ttest: 221038.7724350\tbest: 221038.7724350 (28)\ttotal: 159ms\tremaining: 5.34s\n",
      "29:\tlearn: 211695.9619040\ttest: 216119.7614733\tbest: 216119.7614733 (29)\ttotal: 163ms\tremaining: 5.27s\n",
      "30:\tlearn: 206842.2520820\ttest: 211336.2391038\tbest: 211336.2391038 (30)\ttotal: 178ms\tremaining: 5.55s\n",
      "31:\tlearn: 202164.8308836\ttest: 206768.7716823\tbest: 206768.7716823 (31)\ttotal: 183ms\tremaining: 5.54s\n",
      "32:\tlearn: 197367.4932780\ttest: 201902.0706425\tbest: 201902.0706425 (32)\ttotal: 189ms\tremaining: 5.54s\n",
      "33:\tlearn: 192913.6145481\ttest: 197394.5915023\tbest: 197394.5915023 (33)\ttotal: 196ms\tremaining: 5.56s\n",
      "34:\tlearn: 188546.0837972\ttest: 192950.8324982\tbest: 192950.8324982 (34)\ttotal: 200ms\tremaining: 5.51s\n",
      "35:\tlearn: 184330.5398841\ttest: 188683.4825054\tbest: 188683.4825054 (35)\ttotal: 204ms\tremaining: 5.45s\n",
      "36:\tlearn: 180262.7546138\ttest: 184626.1797622\tbest: 184626.1797622 (36)\ttotal: 207ms\tremaining: 5.39s\n",
      "37:\tlearn: 176707.8637932\ttest: 181143.5740698\tbest: 181143.5740698 (37)\ttotal: 210ms\tremaining: 5.32s\n",
      "38:\tlearn: 172895.0442652\ttest: 177371.1965174\tbest: 177371.1965174 (38)\ttotal: 213ms\tremaining: 5.25s\n",
      "39:\tlearn: 169529.5496966\ttest: 174061.8097481\tbest: 174061.8097481 (39)\ttotal: 216ms\tremaining: 5.18s\n",
      "40:\tlearn: 165957.8048990\ttest: 170582.5964809\tbest: 170582.5964809 (40)\ttotal: 221ms\tremaining: 5.17s\n",
      "41:\tlearn: 162407.6825783\ttest: 167037.7711299\tbest: 167037.7711299 (41)\ttotal: 224ms\tremaining: 5.12s\n",
      "42:\tlearn: 159380.7841768\ttest: 164034.4281389\tbest: 164034.4281389 (42)\ttotal: 228ms\tremaining: 5.07s\n",
      "43:\tlearn: 156515.0471905\ttest: 161192.4449403\tbest: 161192.4449403 (43)\ttotal: 231ms\tremaining: 5.01s\n",
      "44:\tlearn: 153484.2435750\ttest: 158090.7066261\tbest: 158090.7066261 (44)\ttotal: 236ms\tremaining: 5s\n",
      "45:\tlearn: 150804.9663987\ttest: 155415.1661723\tbest: 155415.1661723 (45)\ttotal: 239ms\tremaining: 4.96s\n",
      "46:\tlearn: 147955.6254267\ttest: 152564.1534975\tbest: 152564.1534975 (46)\ttotal: 243ms\tremaining: 4.93s\n",
      "47:\tlearn: 145506.3475057\ttest: 150165.9468451\tbest: 150165.9468451 (47)\ttotal: 246ms\tremaining: 4.88s\n",
      "48:\tlearn: 143085.7561567\ttest: 147861.7729127\tbest: 147861.7729127 (48)\ttotal: 250ms\tremaining: 4.85s\n",
      "49:\tlearn: 140235.3123936\ttest: 144995.7078026\tbest: 144995.7078026 (49)\ttotal: 253ms\tremaining: 4.8s\n",
      "50:\tlearn: 137641.9150832\ttest: 142341.9850255\tbest: 142341.9850255 (50)\ttotal: 257ms\tremaining: 4.77s\n",
      "51:\tlearn: 135254.0725364\ttest: 139927.7869762\tbest: 139927.7869762 (51)\ttotal: 263ms\tremaining: 4.8s\n",
      "52:\tlearn: 132765.9040031\ttest: 137351.9216986\tbest: 137351.9216986 (52)\ttotal: 269ms\tremaining: 4.8s\n",
      "53:\tlearn: 130780.4688893\ttest: 135341.1038070\tbest: 135341.1038070 (53)\ttotal: 272ms\tremaining: 4.76s\n",
      "54:\tlearn: 128643.1966739\ttest: 133151.4559389\tbest: 133151.4559389 (54)\ttotal: 278ms\tremaining: 4.78s\n",
      "55:\tlearn: 126476.0315448\ttest: 130924.0612347\tbest: 130924.0612347 (55)\ttotal: 283ms\tremaining: 4.77s\n",
      "56:\tlearn: 124820.5752098\ttest: 129228.1343238\tbest: 129228.1343238 (56)\ttotal: 287ms\tremaining: 4.74s\n",
      "57:\tlearn: 122943.1769403\ttest: 127301.7170612\tbest: 127301.7170612 (57)\ttotal: 290ms\tremaining: 4.71s\n",
      "58:\tlearn: 120999.8711640\ttest: 125341.5150951\tbest: 125341.5150951 (58)\ttotal: 294ms\tremaining: 4.69s\n",
      "59:\tlearn: 119489.8066673\ttest: 123805.6859576\tbest: 123805.6859576 (59)\ttotal: 299ms\tremaining: 4.68s\n",
      "60:\tlearn: 117701.8543519\ttest: 121914.5983237\tbest: 121914.5983237 (60)\ttotal: 304ms\tremaining: 4.67s\n",
      "61:\tlearn: 116327.0949800\ttest: 120478.4382641\tbest: 120478.4382641 (61)\ttotal: 307ms\tremaining: 4.65s\n",
      "62:\tlearn: 114702.7648797\ttest: 118773.2346533\tbest: 118773.2346533 (62)\ttotal: 310ms\tremaining: 4.61s\n",
      "63:\tlearn: 113044.7711992\ttest: 117041.1593396\tbest: 117041.1593396 (63)\ttotal: 313ms\tremaining: 4.58s\n",
      "64:\tlearn: 111530.9330813\ttest: 115452.3408012\tbest: 115452.3408012 (64)\ttotal: 318ms\tremaining: 4.58s\n",
      "65:\tlearn: 110260.9158299\ttest: 114130.2768858\tbest: 114130.2768858 (65)\ttotal: 322ms\tremaining: 4.55s\n",
      "66:\tlearn: 109138.2780721\ttest: 112968.3801228\tbest: 112968.3801228 (66)\ttotal: 325ms\tremaining: 4.53s\n",
      "67:\tlearn: 108055.5763695\ttest: 111855.7786190\tbest: 111855.7786190 (67)\ttotal: 328ms\tremaining: 4.5s\n",
      "68:\tlearn: 107031.0839197\ttest: 110773.5648329\tbest: 110773.5648329 (68)\ttotal: 331ms\tremaining: 4.47s\n",
      "69:\tlearn: 105794.8100741\ttest: 109478.2882959\tbest: 109478.2882959 (69)\ttotal: 335ms\tremaining: 4.46s\n",
      "70:\tlearn: 104639.8917366\ttest: 108262.4225977\tbest: 108262.4225977 (70)\ttotal: 339ms\tremaining: 4.43s\n",
      "71:\tlearn: 103362.5208793\ttest: 106895.0467960\tbest: 106895.0467960 (71)\ttotal: 342ms\tremaining: 4.41s\n",
      "72:\tlearn: 102307.9338059\ttest: 105803.6060271\tbest: 105803.6060271 (72)\ttotal: 345ms\tremaining: 4.38s\n",
      "73:\tlearn: 100905.4712781\ttest: 104480.1731615\tbest: 104480.1731615 (73)\ttotal: 348ms\tremaining: 4.36s\n",
      "74:\tlearn: 100168.8408303\ttest: 103700.3741101\tbest: 103700.3741101 (74)\ttotal: 351ms\tremaining: 4.33s\n",
      "75:\tlearn: 99372.7071457\ttest: 102857.8246483\tbest: 102857.8246483 (75)\ttotal: 356ms\tremaining: 4.32s\n",
      "76:\tlearn: 98493.5209653\ttest: 101920.6501163\tbest: 101920.6501163 (76)\ttotal: 359ms\tremaining: 4.3s\n",
      "77:\tlearn: 97286.0463636\ttest: 100755.9199902\tbest: 100755.9199902 (77)\ttotal: 362ms\tremaining: 4.28s\n",
      "78:\tlearn: 96308.3507236\ttest: 99666.6869962\tbest: 99666.6869962 (78)\ttotal: 366ms\tremaining: 4.26s\n",
      "79:\tlearn: 95430.9772843\ttest: 98705.1508045\tbest: 98705.1508045 (79)\ttotal: 369ms\tremaining: 4.25s\n",
      "80:\tlearn: 94511.2449043\ttest: 97724.6906362\tbest: 97724.6906362 (80)\ttotal: 373ms\tremaining: 4.23s\n",
      "81:\tlearn: 93545.3561675\ttest: 96696.2462395\tbest: 96696.2462395 (81)\ttotal: 376ms\tremaining: 4.21s\n",
      "82:\tlearn: 92664.1837508\ttest: 95779.0176497\tbest: 95779.0176497 (82)\ttotal: 380ms\tremaining: 4.19s\n",
      "83:\tlearn: 91916.0313548\ttest: 95030.7831143\tbest: 95030.7831143 (83)\ttotal: 383ms\tremaining: 4.17s\n",
      "84:\tlearn: 91164.1900360\ttest: 94153.5007990\tbest: 94153.5007990 (84)\ttotal: 386ms\tremaining: 4.16s\n",
      "85:\tlearn: 90508.7900468\ttest: 93413.9921694\tbest: 93413.9921694 (85)\ttotal: 389ms\tremaining: 4.14s\n",
      "86:\tlearn: 89783.7147725\ttest: 92656.9926273\tbest: 92656.9926273 (86)\ttotal: 393ms\tremaining: 4.13s\n",
      "87:\tlearn: 89111.8427525\ttest: 91902.6037334\tbest: 91902.6037334 (87)\ttotal: 397ms\tremaining: 4.12s\n",
      "88:\tlearn: 88587.1683351\ttest: 91350.2710021\tbest: 91350.2710021 (88)\ttotal: 401ms\tremaining: 4.1s\n",
      "89:\tlearn: 87922.7247934\ttest: 90640.7683698\tbest: 90640.7683698 (89)\ttotal: 404ms\tremaining: 4.08s\n",
      "90:\tlearn: 87355.0862475\ttest: 90020.4427223\tbest: 90020.4427223 (90)\ttotal: 407ms\tremaining: 4.07s\n",
      "91:\tlearn: 86948.9152327\ttest: 89589.2178590\tbest: 89589.2178590 (91)\ttotal: 411ms\tremaining: 4.05s\n",
      "92:\tlearn: 86383.0259172\ttest: 89001.7612463\tbest: 89001.7612463 (92)\ttotal: 414ms\tremaining: 4.04s\n",
      "93:\tlearn: 85995.6284843\ttest: 88565.2181491\tbest: 88565.2181491 (93)\ttotal: 418ms\tremaining: 4.03s\n",
      "94:\tlearn: 85350.1613039\ttest: 87869.9456331\tbest: 87869.9456331 (94)\ttotal: 421ms\tremaining: 4.01s\n",
      "95:\tlearn: 84941.9803269\ttest: 87449.5981615\tbest: 87449.5981615 (95)\ttotal: 424ms\tremaining: 4s\n",
      "96:\tlearn: 84622.8742359\ttest: 87106.0715663\tbest: 87106.0715663 (96)\ttotal: 427ms\tremaining: 3.98s\n",
      "97:\tlearn: 84264.9927594\ttest: 86702.6910861\tbest: 86702.6910861 (97)\ttotal: 430ms\tremaining: 3.96s\n",
      "98:\tlearn: 83687.2279900\ttest: 86106.6586472\tbest: 86106.6586472 (98)\ttotal: 434ms\tremaining: 3.94s\n",
      "99:\tlearn: 82947.5796230\ttest: 85367.5273758\tbest: 85367.5273758 (99)\ttotal: 437ms\tremaining: 3.93s\n",
      "100:\tlearn: 82721.3852743\ttest: 85154.5792893\tbest: 85154.5792893 (100)\ttotal: 440ms\tremaining: 3.92s\n",
      "101:\tlearn: 82194.7392589\ttest: 84593.7236758\tbest: 84593.7236758 (101)\ttotal: 443ms\tremaining: 3.9s\n",
      "102:\tlearn: 81980.5258347\ttest: 84373.1421637\tbest: 84373.1421637 (102)\ttotal: 446ms\tremaining: 3.89s\n",
      "103:\tlearn: 81623.3970521\ttest: 84001.5331429\tbest: 84001.5331429 (103)\ttotal: 451ms\tremaining: 3.89s\n",
      "104:\tlearn: 81105.8995215\ttest: 83449.1648070\tbest: 83449.1648070 (104)\ttotal: 457ms\tremaining: 3.9s\n",
      "105:\tlearn: 80663.4417324\ttest: 82951.7319218\tbest: 82951.7319218 (105)\ttotal: 464ms\tremaining: 3.91s\n",
      "106:\tlearn: 80106.3835066\ttest: 82374.1790445\tbest: 82374.1790445 (106)\ttotal: 470ms\tremaining: 3.92s\n",
      "107:\tlearn: 79819.2852447\ttest: 82080.2843423\tbest: 82080.2843423 (107)\ttotal: 476ms\tremaining: 3.93s\n",
      "108:\tlearn: 79512.5980525\ttest: 81746.4383601\tbest: 81746.4383601 (108)\ttotal: 482ms\tremaining: 3.94s\n",
      "109:\tlearn: 79061.2575868\ttest: 81232.5414961\tbest: 81232.5414961 (109)\ttotal: 487ms\tremaining: 3.94s\n",
      "110:\tlearn: 78625.7478445\ttest: 80833.8557005\tbest: 80833.8557005 (110)\ttotal: 492ms\tremaining: 3.94s\n",
      "111:\tlearn: 78240.5037486\ttest: 80413.3149133\tbest: 80413.3149133 (111)\ttotal: 495ms\tremaining: 3.93s\n",
      "112:\tlearn: 77852.0307324\ttest: 80031.2720713\tbest: 80031.2720713 (112)\ttotal: 500ms\tremaining: 3.92s\n",
      "113:\tlearn: 77513.4467466\ttest: 79671.4446130\tbest: 79671.4446130 (113)\ttotal: 503ms\tremaining: 3.91s\n",
      "114:\tlearn: 76962.4697247\ttest: 79181.1899202\tbest: 79181.1899202 (114)\ttotal: 511ms\tremaining: 3.93s\n",
      "115:\tlearn: 76674.8931154\ttest: 78874.7163084\tbest: 78874.7163084 (115)\ttotal: 516ms\tremaining: 3.93s\n",
      "116:\tlearn: 76492.0213664\ttest: 78666.9293776\tbest: 78666.9293776 (116)\ttotal: 521ms\tremaining: 3.93s\n",
      "117:\tlearn: 76371.0060438\ttest: 78508.5504606\tbest: 78508.5504606 (117)\ttotal: 527ms\tremaining: 3.94s\n",
      "118:\tlearn: 76220.4661833\ttest: 78341.1076542\tbest: 78341.1076542 (118)\ttotal: 533ms\tremaining: 3.95s\n",
      "119:\tlearn: 75976.1437573\ttest: 78047.1050597\tbest: 78047.1050597 (119)\ttotal: 543ms\tremaining: 3.98s\n",
      "120:\tlearn: 75677.6474029\ttest: 77706.7630785\tbest: 77706.7630785 (120)\ttotal: 549ms\tremaining: 3.99s\n",
      "121:\tlearn: 75349.2069456\ttest: 77371.3451331\tbest: 77371.3451331 (121)\ttotal: 559ms\tremaining: 4.02s\n",
      "122:\tlearn: 75090.2680887\ttest: 77128.8593207\tbest: 77128.8593207 (122)\ttotal: 565ms\tremaining: 4.03s\n",
      "123:\tlearn: 74973.4931813\ttest: 76988.6271030\tbest: 76988.6271030 (123)\ttotal: 573ms\tremaining: 4.05s\n",
      "124:\tlearn: 74847.6982963\ttest: 76824.9142158\tbest: 76824.9142158 (124)\ttotal: 580ms\tremaining: 4.06s\n",
      "125:\tlearn: 74616.2027047\ttest: 76540.4034788\tbest: 76540.4034788 (125)\ttotal: 589ms\tremaining: 4.09s\n",
      "126:\tlearn: 74366.5352277\ttest: 76272.7767838\tbest: 76272.7767838 (126)\ttotal: 596ms\tremaining: 4.09s\n",
      "127:\tlearn: 74298.6673975\ttest: 76154.6828133\tbest: 76154.6828133 (127)\ttotal: 601ms\tremaining: 4.09s\n",
      "128:\tlearn: 74118.2651618\ttest: 76004.2693252\tbest: 76004.2693252 (128)\ttotal: 610ms\tremaining: 4.12s\n",
      "129:\tlearn: 74054.8208449\ttest: 75924.0403503\tbest: 75924.0403503 (129)\ttotal: 616ms\tremaining: 4.12s\n",
      "130:\tlearn: 74000.0079065\ttest: 75856.2987225\tbest: 75856.2987225 (130)\ttotal: 625ms\tremaining: 4.14s\n",
      "131:\tlearn: 73905.0800101\ttest: 75758.2045605\tbest: 75758.2045605 (131)\ttotal: 632ms\tremaining: 4.15s\n",
      "132:\tlearn: 73772.1948959\ttest: 75596.2716443\tbest: 75596.2716443 (132)\ttotal: 640ms\tremaining: 4.17s\n",
      "133:\tlearn: 73548.7939721\ttest: 75360.0424607\tbest: 75360.0424607 (133)\ttotal: 648ms\tremaining: 4.19s\n",
      "134:\tlearn: 73175.9219480\ttest: 75033.8686878\tbest: 75033.8686878 (134)\ttotal: 656ms\tremaining: 4.21s\n",
      "135:\tlearn: 73077.5625514\ttest: 74926.6063290\tbest: 74926.6063290 (135)\ttotal: 662ms\tremaining: 4.21s\n",
      "136:\tlearn: 72867.3438483\ttest: 74704.2100950\tbest: 74704.2100950 (136)\ttotal: 671ms\tremaining: 4.23s\n",
      "137:\tlearn: 72793.1467904\ttest: 74596.5572176\tbest: 74596.5572176 (137)\ttotal: 677ms\tremaining: 4.23s\n",
      "138:\tlearn: 72667.7208714\ttest: 74494.3067190\tbest: 74494.3067190 (138)\ttotal: 686ms\tremaining: 4.25s\n",
      "139:\tlearn: 72531.8904822\ttest: 74359.1047737\tbest: 74359.1047737 (139)\ttotal: 691ms\tremaining: 4.25s\n",
      "140:\tlearn: 72400.1671583\ttest: 74172.8823362\tbest: 74172.8823362 (140)\ttotal: 697ms\tremaining: 4.25s\n",
      "141:\tlearn: 72357.0213994\ttest: 74120.2337947\tbest: 74120.2337947 (141)\ttotal: 701ms\tremaining: 4.24s\n",
      "142:\tlearn: 72032.6993740\ttest: 73866.5251290\tbest: 73866.5251290 (142)\ttotal: 706ms\tremaining: 4.23s\n",
      "143:\tlearn: 71844.4132395\ttest: 73730.5226832\tbest: 73730.5226832 (143)\ttotal: 713ms\tremaining: 4.24s\n",
      "144:\tlearn: 71732.7307754\ttest: 73610.1548556\tbest: 73610.1548556 (144)\ttotal: 719ms\tremaining: 4.24s\n",
      "145:\tlearn: 71546.2726456\ttest: 73410.5545558\tbest: 73410.5545558 (145)\ttotal: 726ms\tremaining: 4.24s\n",
      "146:\tlearn: 71341.5447261\ttest: 73245.4665276\tbest: 73245.4665276 (146)\ttotal: 735ms\tremaining: 4.27s\n",
      "147:\tlearn: 71114.7039091\ttest: 73050.8325368\tbest: 73050.8325368 (147)\ttotal: 742ms\tremaining: 4.27s\n",
      "148:\tlearn: 71031.0604418\ttest: 72959.6409025\tbest: 72959.6409025 (148)\ttotal: 750ms\tremaining: 4.28s\n",
      "149:\tlearn: 70910.9341960\ttest: 72862.2241312\tbest: 72862.2241312 (149)\ttotal: 756ms\tremaining: 4.29s\n",
      "150:\tlearn: 70898.1010082\ttest: 72840.5252494\tbest: 72840.5252494 (150)\ttotal: 762ms\tremaining: 4.29s\n",
      "151:\tlearn: 70596.4200557\ttest: 72535.0245076\tbest: 72535.0245076 (151)\ttotal: 769ms\tremaining: 4.29s\n",
      "152:\tlearn: 70503.8504527\ttest: 72423.9480335\tbest: 72423.9480335 (152)\ttotal: 776ms\tremaining: 4.29s\n",
      "153:\tlearn: 70377.0273220\ttest: 72282.7918144\tbest: 72282.7918144 (153)\ttotal: 783ms\tremaining: 4.3s\n",
      "154:\tlearn: 70260.5111972\ttest: 72146.5659541\tbest: 72146.5659541 (154)\ttotal: 788ms\tremaining: 4.29s\n",
      "155:\tlearn: 70181.1035146\ttest: 72055.7315585\tbest: 72055.7315585 (155)\ttotal: 796ms\tremaining: 4.31s\n",
      "156:\tlearn: 70074.4637384\ttest: 71988.8411911\tbest: 71988.8411911 (156)\ttotal: 802ms\tremaining: 4.3s\n",
      "157:\tlearn: 69935.5452394\ttest: 71809.3052571\tbest: 71809.3052571 (157)\ttotal: 806ms\tremaining: 4.29s\n",
      "158:\tlearn: 69820.6340242\ttest: 71704.5027880\tbest: 71704.5027880 (158)\ttotal: 813ms\tremaining: 4.3s\n",
      "159:\tlearn: 69789.8320078\ttest: 71679.5289626\tbest: 71679.5289626 (159)\ttotal: 819ms\tremaining: 4.3s\n",
      "160:\tlearn: 69651.2223850\ttest: 71528.2694797\tbest: 71528.2694797 (160)\ttotal: 825ms\tremaining: 4.3s\n",
      "161:\tlearn: 69611.1544291\ttest: 71461.8941305\tbest: 71461.8941305 (161)\ttotal: 833ms\tremaining: 4.31s\n",
      "162:\tlearn: 69569.0985278\ttest: 71394.7112934\tbest: 71394.7112934 (162)\ttotal: 836ms\tremaining: 4.29s\n",
      "163:\tlearn: 69265.1571433\ttest: 71195.2397534\tbest: 71195.2397534 (163)\ttotal: 841ms\tremaining: 4.29s\n",
      "164:\tlearn: 68960.2672145\ttest: 70989.1937231\tbest: 70989.1937231 (164)\ttotal: 848ms\tremaining: 4.29s\n",
      "165:\tlearn: 68942.5463182\ttest: 70959.8672657\tbest: 70959.8672657 (165)\ttotal: 855ms\tremaining: 4.29s\n",
      "166:\tlearn: 68864.9742613\ttest: 70880.9601993\tbest: 70880.9601993 (166)\ttotal: 862ms\tremaining: 4.3s\n",
      "167:\tlearn: 68747.5967245\ttest: 70742.2855779\tbest: 70742.2855779 (167)\ttotal: 867ms\tremaining: 4.29s\n",
      "168:\tlearn: 68670.9022908\ttest: 70672.5322650\tbest: 70672.5322650 (168)\ttotal: 872ms\tremaining: 4.29s\n",
      "169:\tlearn: 68601.5104462\ttest: 70591.3338410\tbest: 70591.3338410 (169)\ttotal: 878ms\tremaining: 4.29s\n",
      "170:\tlearn: 68532.5941115\ttest: 70501.0906046\tbest: 70501.0906046 (170)\ttotal: 884ms\tremaining: 4.29s\n",
      "171:\tlearn: 68422.0233338\ttest: 70394.1244223\tbest: 70394.1244223 (171)\ttotal: 890ms\tremaining: 4.29s\n",
      "172:\tlearn: 68369.4457950\ttest: 70372.6143539\tbest: 70372.6143539 (172)\ttotal: 895ms\tremaining: 4.28s\n",
      "173:\tlearn: 68243.8091432\ttest: 70233.8363163\tbest: 70233.8363163 (173)\ttotal: 901ms\tremaining: 4.28s\n",
      "174:\tlearn: 68218.7557550\ttest: 70203.8228729\tbest: 70203.8228729 (174)\ttotal: 907ms\tremaining: 4.28s\n",
      "175:\tlearn: 67887.4747620\ttest: 69979.5985426\tbest: 69979.5985426 (175)\ttotal: 913ms\tremaining: 4.28s\n",
      "176:\tlearn: 67645.1941362\ttest: 69799.7417332\tbest: 69799.7417332 (176)\ttotal: 921ms\tremaining: 4.28s\n",
      "177:\tlearn: 67402.6292061\ttest: 69640.1564834\tbest: 69640.1564834 (177)\ttotal: 929ms\tremaining: 4.29s\n",
      "178:\tlearn: 67304.1545366\ttest: 69539.8887795\tbest: 69539.8887795 (178)\ttotal: 936ms\tremaining: 4.29s\n",
      "179:\tlearn: 67221.3753923\ttest: 69483.2184313\tbest: 69483.2184313 (179)\ttotal: 945ms\tremaining: 4.3s\n",
      "180:\tlearn: 67204.7294996\ttest: 69480.1531058\tbest: 69480.1531058 (180)\ttotal: 952ms\tremaining: 4.31s\n",
      "181:\tlearn: 67000.6622089\ttest: 69302.4378367\tbest: 69302.4378367 (181)\ttotal: 960ms\tremaining: 4.31s\n",
      "182:\tlearn: 66950.0516222\ttest: 69233.1922902\tbest: 69233.1922902 (182)\ttotal: 966ms\tremaining: 4.31s\n",
      "183:\tlearn: 66911.5776830\ttest: 69190.0175744\tbest: 69190.0175744 (183)\ttotal: 974ms\tremaining: 4.32s\n",
      "184:\tlearn: 66849.2824282\ttest: 69144.2992143\tbest: 69144.2992143 (184)\ttotal: 979ms\tremaining: 4.31s\n",
      "185:\tlearn: 66628.0819079\ttest: 68972.4672103\tbest: 68972.4672103 (185)\ttotal: 987ms\tremaining: 4.32s\n",
      "186:\tlearn: 66610.0952791\ttest: 68947.7579906\tbest: 68947.7579906 (186)\ttotal: 992ms\tremaining: 4.31s\n",
      "187:\tlearn: 66354.0286356\ttest: 68769.2639206\tbest: 68769.2639206 (187)\ttotal: 996ms\tremaining: 4.3s\n",
      "188:\tlearn: 66330.2805774\ttest: 68731.9730613\tbest: 68731.9730613 (188)\ttotal: 1s\tremaining: 4.3s\n",
      "189:\tlearn: 66322.7163036\ttest: 68719.2965644\tbest: 68719.2965644 (189)\ttotal: 1.01s\tremaining: 4.3s\n",
      "190:\tlearn: 66290.6502954\ttest: 68681.2526979\tbest: 68681.2526979 (190)\ttotal: 1.01s\tremaining: 4.28s\n",
      "191:\tlearn: 66233.1500631\ttest: 68643.7868397\tbest: 68643.7868397 (191)\ttotal: 1.02s\tremaining: 4.29s\n",
      "192:\tlearn: 66178.3825734\ttest: 68595.7988624\tbest: 68595.7988624 (192)\ttotal: 1.02s\tremaining: 4.28s\n",
      "193:\tlearn: 65953.4544416\ttest: 68440.7648669\tbest: 68440.7648669 (193)\ttotal: 1.03s\tremaining: 4.27s\n",
      "194:\tlearn: 65895.6604218\ttest: 68364.0107155\tbest: 68364.0107155 (194)\ttotal: 1.03s\tremaining: 4.28s\n",
      "195:\tlearn: 65892.6517682\ttest: 68345.8202930\tbest: 68345.8202930 (195)\ttotal: 1.04s\tremaining: 4.26s\n",
      "196:\tlearn: 65829.2571738\ttest: 68244.8360112\tbest: 68244.8360112 (196)\ttotal: 1.05s\tremaining: 4.26s\n",
      "197:\tlearn: 65618.1659006\ttest: 68093.0703970\tbest: 68093.0703970 (197)\ttotal: 1.05s\tremaining: 4.27s\n",
      "198:\tlearn: 65572.5117814\ttest: 68035.0383778\tbest: 68035.0383778 (198)\ttotal: 1.06s\tremaining: 4.26s\n",
      "199:\tlearn: 65516.4020509\ttest: 68026.6234828\tbest: 68026.6234828 (199)\ttotal: 1.06s\tremaining: 4.26s\n",
      "200:\tlearn: 65406.6378253\ttest: 67944.2155426\tbest: 67944.2155426 (200)\ttotal: 1.07s\tremaining: 4.26s\n",
      "201:\tlearn: 65361.2185310\ttest: 67925.9297351\tbest: 67925.9297351 (201)\ttotal: 1.08s\tremaining: 4.25s\n",
      "202:\tlearn: 65284.1702276\ttest: 67871.3567226\tbest: 67871.3567226 (202)\ttotal: 1.08s\tremaining: 4.25s\n",
      "203:\tlearn: 65172.3365886\ttest: 67819.5786565\tbest: 67819.5786565 (203)\ttotal: 1.09s\tremaining: 4.25s\n",
      "204:\tlearn: 65083.8632291\ttest: 67776.8401742\tbest: 67776.8401742 (204)\ttotal: 1.09s\tremaining: 4.24s\n",
      "205:\tlearn: 64997.4733062\ttest: 67752.5105994\tbest: 67752.5105994 (205)\ttotal: 1.1s\tremaining: 4.24s\n",
      "206:\tlearn: 64917.4158952\ttest: 67678.1478057\tbest: 67678.1478057 (206)\ttotal: 1.11s\tremaining: 4.24s\n",
      "207:\tlearn: 64855.5369993\ttest: 67641.0982090\tbest: 67641.0982090 (207)\ttotal: 1.11s\tremaining: 4.25s\n",
      "208:\tlearn: 64844.8772705\ttest: 67638.7144170\tbest: 67638.7144170 (208)\ttotal: 1.13s\tremaining: 4.28s\n",
      "209:\tlearn: 64638.0199409\ttest: 67458.9952439\tbest: 67458.9952439 (209)\ttotal: 1.14s\tremaining: 4.28s\n",
      "210:\tlearn: 64555.6888726\ttest: 67399.1644823\tbest: 67399.1644823 (210)\ttotal: 1.15s\tremaining: 4.28s\n",
      "211:\tlearn: 64346.0321469\ttest: 67225.7292374\tbest: 67225.7292374 (211)\ttotal: 1.15s\tremaining: 4.29s\n",
      "212:\tlearn: 64284.6896213\ttest: 67204.3393981\tbest: 67204.3393981 (212)\ttotal: 1.16s\tremaining: 4.28s\n",
      "213:\tlearn: 64269.6545582\ttest: 67193.3992095\tbest: 67193.3992095 (213)\ttotal: 1.17s\tremaining: 4.29s\n",
      "214:\tlearn: 64233.0867565\ttest: 67165.4946589\tbest: 67165.4946589 (214)\ttotal: 1.17s\tremaining: 4.29s\n",
      "215:\tlearn: 64192.0075668\ttest: 67142.1547996\tbest: 67142.1547996 (215)\ttotal: 1.18s\tremaining: 4.29s\n",
      "216:\tlearn: 64138.4373714\ttest: 67077.9565449\tbest: 67077.9565449 (216)\ttotal: 1.19s\tremaining: 4.28s\n",
      "217:\tlearn: 64074.5201283\ttest: 67003.8707050\tbest: 67003.8707050 (217)\ttotal: 1.19s\tremaining: 4.28s\n",
      "218:\tlearn: 64045.0833497\ttest: 66970.8789472\tbest: 66970.8789472 (218)\ttotal: 1.2s\tremaining: 4.28s\n",
      "219:\tlearn: 64033.2652843\ttest: 66955.3397490\tbest: 66955.3397490 (219)\ttotal: 1.2s\tremaining: 4.27s\n",
      "220:\tlearn: 64017.0973655\ttest: 66942.2923100\tbest: 66942.2923100 (220)\ttotal: 1.21s\tremaining: 4.27s\n",
      "221:\tlearn: 63946.6880834\ttest: 66915.1674691\tbest: 66915.1674691 (221)\ttotal: 1.22s\tremaining: 4.26s\n",
      "222:\tlearn: 63903.7352122\ttest: 66865.4241652\tbest: 66865.4241652 (222)\ttotal: 1.22s\tremaining: 4.25s\n",
      "223:\tlearn: 63848.9102454\ttest: 66816.1184357\tbest: 66816.1184357 (223)\ttotal: 1.23s\tremaining: 4.25s\n",
      "224:\tlearn: 63803.0441611\ttest: 66775.4517518\tbest: 66775.4517518 (224)\ttotal: 1.23s\tremaining: 4.25s\n",
      "225:\tlearn: 63630.2801301\ttest: 66648.5421025\tbest: 66648.5421025 (225)\ttotal: 1.24s\tremaining: 4.25s\n",
      "226:\tlearn: 63625.5207462\ttest: 66640.9602006\tbest: 66640.9602006 (226)\ttotal: 1.25s\tremaining: 4.24s\n",
      "227:\tlearn: 63607.3338108\ttest: 66626.0162970\tbest: 66626.0162970 (227)\ttotal: 1.25s\tremaining: 4.23s\n",
      "228:\tlearn: 63560.8034097\ttest: 66578.9707897\tbest: 66578.9707897 (228)\ttotal: 1.26s\tremaining: 4.23s\n",
      "229:\tlearn: 63502.0569952\ttest: 66537.1072140\tbest: 66537.1072140 (229)\ttotal: 1.26s\tremaining: 4.22s\n",
      "230:\tlearn: 63457.9387369\ttest: 66508.3719643\tbest: 66508.3719643 (230)\ttotal: 1.27s\tremaining: 4.22s\n",
      "231:\tlearn: 63409.3966200\ttest: 66467.0735268\tbest: 66467.0735268 (231)\ttotal: 1.27s\tremaining: 4.21s\n",
      "232:\tlearn: 63221.8689750\ttest: 66287.1315032\tbest: 66287.1315032 (232)\ttotal: 1.28s\tremaining: 4.21s\n",
      "233:\tlearn: 63183.6679393\ttest: 66253.3210006\tbest: 66253.3210006 (233)\ttotal: 1.28s\tremaining: 4.21s\n",
      "234:\tlearn: 63140.9693722\ttest: 66250.0959461\tbest: 66250.0959461 (234)\ttotal: 1.29s\tremaining: 4.2s\n",
      "235:\tlearn: 63084.3200509\ttest: 66182.6102327\tbest: 66182.6102327 (235)\ttotal: 1.29s\tremaining: 4.19s\n",
      "236:\tlearn: 62919.6789362\ttest: 66047.7237200\tbest: 66047.7237200 (236)\ttotal: 1.3s\tremaining: 4.19s\n",
      "237:\tlearn: 62832.6607529\ttest: 65996.8447112\tbest: 65996.8447112 (237)\ttotal: 1.31s\tremaining: 4.18s\n",
      "238:\tlearn: 62814.5603100\ttest: 65984.0652938\tbest: 65984.0652938 (238)\ttotal: 1.31s\tremaining: 4.18s\n",
      "239:\tlearn: 62647.5600854\ttest: 65864.4574179\tbest: 65864.4574179 (239)\ttotal: 1.32s\tremaining: 4.17s\n",
      "240:\tlearn: 62608.3468352\ttest: 65853.6703365\tbest: 65853.6703365 (240)\ttotal: 1.32s\tremaining: 4.17s\n",
      "241:\tlearn: 62469.5366155\ttest: 65787.3907398\tbest: 65787.3907398 (241)\ttotal: 1.33s\tremaining: 4.17s\n",
      "242:\tlearn: 62407.2182780\ttest: 65744.0485101\tbest: 65744.0485101 (242)\ttotal: 1.34s\tremaining: 4.17s\n",
      "243:\tlearn: 62373.7413368\ttest: 65740.9108228\tbest: 65740.9108228 (243)\ttotal: 1.34s\tremaining: 4.17s\n",
      "244:\tlearn: 62370.3438385\ttest: 65724.4493869\tbest: 65724.4493869 (244)\ttotal: 1.35s\tremaining: 4.17s\n",
      "245:\tlearn: 62325.0049966\ttest: 65706.7140608\tbest: 65706.7140608 (245)\ttotal: 1.36s\tremaining: 4.17s\n",
      "246:\tlearn: 62222.9036955\ttest: 65603.2458517\tbest: 65603.2458517 (246)\ttotal: 1.37s\tremaining: 4.19s\n",
      "247:\tlearn: 62110.1393582\ttest: 65467.7778124\tbest: 65467.7778124 (247)\ttotal: 1.38s\tremaining: 4.19s\n",
      "248:\tlearn: 62056.1041761\ttest: 65445.2634782\tbest: 65445.2634782 (248)\ttotal: 1.39s\tremaining: 4.19s\n",
      "249:\tlearn: 62021.0983042\ttest: 65430.6098886\tbest: 65430.6098886 (249)\ttotal: 1.39s\tremaining: 4.18s\n",
      "250:\tlearn: 61860.5792032\ttest: 65277.8620804\tbest: 65277.8620804 (250)\ttotal: 1.4s\tremaining: 4.18s\n",
      "251:\tlearn: 61855.4627336\ttest: 65255.7141072\tbest: 65255.7141072 (251)\ttotal: 1.41s\tremaining: 4.17s\n",
      "252:\tlearn: 61840.1413391\ttest: 65248.2272423\tbest: 65248.2272423 (252)\ttotal: 1.42s\tremaining: 4.18s\n",
      "253:\tlearn: 61830.4496176\ttest: 65254.5075757\tbest: 65248.2272423 (252)\ttotal: 1.42s\tremaining: 4.18s\n",
      "254:\tlearn: 61759.8194990\ttest: 65200.2255138\tbest: 65200.2255138 (254)\ttotal: 1.43s\tremaining: 4.18s\n",
      "255:\tlearn: 61743.4169197\ttest: 65195.8378828\tbest: 65195.8378828 (255)\ttotal: 1.44s\tremaining: 4.19s\n",
      "256:\tlearn: 61692.1246606\ttest: 65161.0391909\tbest: 65161.0391909 (256)\ttotal: 1.45s\tremaining: 4.19s\n",
      "257:\tlearn: 61676.6231702\ttest: 65160.5471489\tbest: 65160.5471489 (257)\ttotal: 1.46s\tremaining: 4.19s\n",
      "258:\tlearn: 61605.0464304\ttest: 65109.6551198\tbest: 65109.6551198 (258)\ttotal: 1.46s\tremaining: 4.19s\n",
      "259:\tlearn: 61557.5236260\ttest: 65066.7146274\tbest: 65066.7146274 (259)\ttotal: 1.47s\tremaining: 4.18s\n",
      "260:\tlearn: 61512.2906060\ttest: 65028.3199527\tbest: 65028.3199527 (260)\ttotal: 1.47s\tremaining: 4.17s\n",
      "261:\tlearn: 61439.0353545\ttest: 64944.3762365\tbest: 64944.3762365 (261)\ttotal: 1.48s\tremaining: 4.17s\n",
      "262:\tlearn: 61345.1182032\ttest: 64868.6538703\tbest: 64868.6538703 (262)\ttotal: 1.49s\tremaining: 4.16s\n",
      "263:\tlearn: 61201.9679640\ttest: 64811.3964426\tbest: 64811.3964426 (263)\ttotal: 1.49s\tremaining: 4.16s\n",
      "264:\tlearn: 61180.4292655\ttest: 64799.0867491\tbest: 64799.0867491 (264)\ttotal: 1.5s\tremaining: 4.15s\n",
      "265:\tlearn: 61139.9232501\ttest: 64757.2831836\tbest: 64757.2831836 (265)\ttotal: 1.5s\tremaining: 4.14s\n",
      "266:\tlearn: 61107.5406514\ttest: 64741.7983138\tbest: 64741.7983138 (266)\ttotal: 1.51s\tremaining: 4.15s\n",
      "267:\tlearn: 61099.1117889\ttest: 64739.6086149\tbest: 64739.6086149 (267)\ttotal: 1.52s\tremaining: 4.15s\n",
      "268:\tlearn: 61054.8167034\ttest: 64699.9304324\tbest: 64699.9304324 (268)\ttotal: 1.53s\tremaining: 4.15s\n",
      "269:\tlearn: 60965.8396675\ttest: 64683.3715576\tbest: 64683.3715576 (269)\ttotal: 1.53s\tremaining: 4.15s\n",
      "270:\tlearn: 60905.3571760\ttest: 64651.0389698\tbest: 64651.0389698 (270)\ttotal: 1.54s\tremaining: 4.13s\n",
      "271:\tlearn: 60781.3432267\ttest: 64592.9031179\tbest: 64592.9031179 (271)\ttotal: 1.54s\tremaining: 4.12s\n",
      "272:\tlearn: 60686.8795460\ttest: 64549.1298737\tbest: 64549.1298737 (272)\ttotal: 1.54s\tremaining: 4.11s\n",
      "273:\tlearn: 60677.6544508\ttest: 64552.1426282\tbest: 64549.1298737 (272)\ttotal: 1.55s\tremaining: 4.1s\n",
      "274:\tlearn: 60645.3414565\ttest: 64529.7263225\tbest: 64529.7263225 (274)\ttotal: 1.55s\tremaining: 4.09s\n",
      "275:\tlearn: 60622.5272312\ttest: 64514.8933377\tbest: 64514.8933377 (275)\ttotal: 1.55s\tremaining: 4.08s\n",
      "276:\tlearn: 60517.9963166\ttest: 64411.8515842\tbest: 64411.8515842 (276)\ttotal: 1.56s\tremaining: 4.06s\n",
      "277:\tlearn: 60474.5100681\ttest: 64377.1255444\tbest: 64377.1255444 (277)\ttotal: 1.56s\tremaining: 4.05s\n",
      "278:\tlearn: 60342.0622120\ttest: 64249.3254194\tbest: 64249.3254194 (278)\ttotal: 1.56s\tremaining: 4.04s\n",
      "279:\tlearn: 60302.8215431\ttest: 64248.7582008\tbest: 64248.7582008 (279)\ttotal: 1.57s\tremaining: 4.03s\n",
      "280:\tlearn: 60141.7285230\ttest: 64144.5552640\tbest: 64144.5552640 (280)\ttotal: 1.57s\tremaining: 4.02s\n",
      "281:\tlearn: 60085.1683895\ttest: 64096.7189918\tbest: 64096.7189918 (281)\ttotal: 1.57s\tremaining: 4.01s\n",
      "282:\tlearn: 60045.0681481\ttest: 64057.4809737\tbest: 64057.4809737 (282)\ttotal: 1.58s\tremaining: 4s\n",
      "283:\tlearn: 60026.2757415\ttest: 64047.6743350\tbest: 64047.6743350 (283)\ttotal: 1.58s\tremaining: 3.99s\n",
      "284:\tlearn: 59996.7078459\ttest: 64013.8710013\tbest: 64013.8710013 (284)\ttotal: 1.58s\tremaining: 3.97s\n",
      "285:\tlearn: 59880.1828040\ttest: 63963.5051350\tbest: 63963.5051350 (285)\ttotal: 1.59s\tremaining: 3.96s\n",
      "286:\tlearn: 59749.3029024\ttest: 63876.6001454\tbest: 63876.6001454 (286)\ttotal: 1.59s\tremaining: 3.95s\n",
      "287:\tlearn: 59722.2386507\ttest: 63865.8038346\tbest: 63865.8038346 (287)\ttotal: 1.59s\tremaining: 3.94s\n",
      "288:\tlearn: 59705.7975963\ttest: 63854.9326417\tbest: 63854.9326417 (288)\ttotal: 1.59s\tremaining: 3.93s\n",
      "289:\tlearn: 59665.3335714\ttest: 63830.6738053\tbest: 63830.6738053 (289)\ttotal: 1.6s\tremaining: 3.92s\n",
      "290:\tlearn: 59663.4549997\ttest: 63829.8699315\tbest: 63829.8699315 (290)\ttotal: 1.6s\tremaining: 3.9s\n",
      "291:\tlearn: 59618.8778370\ttest: 63797.3111687\tbest: 63797.3111687 (291)\ttotal: 1.6s\tremaining: 3.89s\n",
      "292:\tlearn: 59503.6394260\ttest: 63795.4494901\tbest: 63795.4494901 (292)\ttotal: 1.61s\tremaining: 3.88s\n",
      "293:\tlearn: 59468.4130147\ttest: 63774.8721434\tbest: 63774.8721434 (293)\ttotal: 1.61s\tremaining: 3.87s\n",
      "294:\tlearn: 59451.3892964\ttest: 63758.8216357\tbest: 63758.8216357 (294)\ttotal: 1.61s\tremaining: 3.86s\n",
      "295:\tlearn: 59392.3523946\ttest: 63726.0557590\tbest: 63726.0557590 (295)\ttotal: 1.62s\tremaining: 3.85s\n",
      "296:\tlearn: 59359.1224562\ttest: 63698.1526541\tbest: 63698.1526541 (296)\ttotal: 1.63s\tremaining: 3.85s\n",
      "297:\tlearn: 59347.9871828\ttest: 63691.3048112\tbest: 63691.3048112 (297)\ttotal: 1.63s\tremaining: 3.84s\n",
      "298:\tlearn: 59315.4283794\ttest: 63649.4701183\tbest: 63649.4701183 (298)\ttotal: 1.64s\tremaining: 3.84s\n",
      "299:\tlearn: 59255.8402606\ttest: 63657.8415387\tbest: 63649.4701183 (298)\ttotal: 1.64s\tremaining: 3.83s\n",
      "300:\tlearn: 59225.7776779\ttest: 63652.7728251\tbest: 63649.4701183 (298)\ttotal: 1.65s\tremaining: 3.83s\n",
      "301:\tlearn: 59178.0107680\ttest: 63594.2110001\tbest: 63594.2110001 (301)\ttotal: 1.66s\tremaining: 3.83s\n",
      "302:\tlearn: 59153.3473068\ttest: 63599.4404329\tbest: 63594.2110001 (301)\ttotal: 1.66s\tremaining: 3.82s\n",
      "303:\tlearn: 59064.8796593\ttest: 63555.0653795\tbest: 63555.0653795 (303)\ttotal: 1.67s\tremaining: 3.81s\n",
      "304:\tlearn: 59038.8930633\ttest: 63523.9048502\tbest: 63523.9048502 (304)\ttotal: 1.67s\tremaining: 3.81s\n",
      "305:\tlearn: 59012.5444732\ttest: 63514.1725165\tbest: 63514.1725165 (305)\ttotal: 1.68s\tremaining: 3.8s\n",
      "306:\tlearn: 58913.5065643\ttest: 63450.9504440\tbest: 63450.9504440 (306)\ttotal: 1.68s\tremaining: 3.8s\n",
      "307:\tlearn: 58898.1608295\ttest: 63446.1450401\tbest: 63446.1450401 (307)\ttotal: 1.69s\tremaining: 3.79s\n",
      "308:\tlearn: 58877.6755474\ttest: 63447.7020735\tbest: 63446.1450401 (307)\ttotal: 1.69s\tremaining: 3.79s\n",
      "309:\tlearn: 58803.1872731\ttest: 63441.0888422\tbest: 63441.0888422 (309)\ttotal: 1.7s\tremaining: 3.78s\n",
      "310:\tlearn: 58722.9271313\ttest: 63383.6701393\tbest: 63383.6701393 (310)\ttotal: 1.71s\tremaining: 3.78s\n",
      "311:\tlearn: 58699.9303791\ttest: 63355.7304114\tbest: 63355.7304114 (311)\ttotal: 1.71s\tremaining: 3.77s\n",
      "312:\tlearn: 58653.5609602\ttest: 63335.0572364\tbest: 63335.0572364 (312)\ttotal: 1.71s\tremaining: 3.76s\n",
      "313:\tlearn: 58599.7517764\ttest: 63318.1480085\tbest: 63318.1480085 (313)\ttotal: 1.72s\tremaining: 3.75s\n",
      "314:\tlearn: 58490.1127809\ttest: 63308.7387162\tbest: 63308.7387162 (314)\ttotal: 1.72s\tremaining: 3.75s\n",
      "315:\tlearn: 58457.9402768\ttest: 63268.4312943\tbest: 63268.4312943 (315)\ttotal: 1.73s\tremaining: 3.74s\n",
      "316:\tlearn: 58457.3586284\ttest: 63268.6341857\tbest: 63268.4312943 (315)\ttotal: 1.73s\tremaining: 3.73s\n",
      "317:\tlearn: 58452.1322155\ttest: 63257.8127097\tbest: 63257.8127097 (317)\ttotal: 1.74s\tremaining: 3.72s\n",
      "318:\tlearn: 58438.1118715\ttest: 63252.9586472\tbest: 63252.9586472 (318)\ttotal: 1.74s\tremaining: 3.72s\n",
      "319:\tlearn: 58437.3275986\ttest: 63257.4581269\tbest: 63252.9586472 (318)\ttotal: 1.75s\tremaining: 3.72s\n",
      "320:\tlearn: 58402.6306664\ttest: 63217.9925338\tbest: 63217.9925338 (320)\ttotal: 1.76s\tremaining: 3.72s\n",
      "321:\tlearn: 58284.9389971\ttest: 63187.7108657\tbest: 63187.7108657 (321)\ttotal: 1.76s\tremaining: 3.72s\n",
      "322:\tlearn: 58248.5661291\ttest: 63177.2730553\tbest: 63177.2730553 (322)\ttotal: 1.77s\tremaining: 3.72s\n",
      "323:\tlearn: 58221.7002613\ttest: 63172.8791308\tbest: 63172.8791308 (323)\ttotal: 1.78s\tremaining: 3.71s\n",
      "324:\tlearn: 58189.4075410\ttest: 63165.1331525\tbest: 63165.1331525 (324)\ttotal: 1.79s\tremaining: 3.71s\n",
      "325:\tlearn: 58139.9906804\ttest: 63173.2824262\tbest: 63165.1331525 (324)\ttotal: 1.8s\tremaining: 3.72s\n",
      "326:\tlearn: 58035.3139141\ttest: 63071.6467920\tbest: 63071.6467920 (326)\ttotal: 1.81s\tremaining: 3.72s\n",
      "327:\tlearn: 57989.3657333\ttest: 63073.0404654\tbest: 63071.6467920 (326)\ttotal: 1.82s\tremaining: 3.72s\n",
      "328:\tlearn: 57905.4741402\ttest: 63036.7552586\tbest: 63036.7552586 (328)\ttotal: 1.82s\tremaining: 3.72s\n",
      "329:\tlearn: 57881.1331527\ttest: 63033.0354385\tbest: 63033.0354385 (329)\ttotal: 1.83s\tremaining: 3.72s\n",
      "330:\tlearn: 57879.7335045\ttest: 63031.4968925\tbest: 63031.4968925 (330)\ttotal: 1.84s\tremaining: 3.72s\n",
      "331:\tlearn: 57840.9155441\ttest: 62983.4914061\tbest: 62983.4914061 (331)\ttotal: 1.85s\tremaining: 3.72s\n",
      "332:\tlearn: 57800.8785003\ttest: 62973.7338229\tbest: 62973.7338229 (332)\ttotal: 1.86s\tremaining: 3.72s\n",
      "333:\tlearn: 57781.6343753\ttest: 62956.3279113\tbest: 62956.3279113 (333)\ttotal: 1.86s\tremaining: 3.72s\n",
      "334:\tlearn: 57735.3620292\ttest: 62923.8001262\tbest: 62923.8001262 (334)\ttotal: 1.87s\tremaining: 3.72s\n",
      "335:\tlearn: 57656.5103851\ttest: 62879.5636926\tbest: 62879.5636926 (335)\ttotal: 1.88s\tremaining: 3.72s\n",
      "336:\tlearn: 57636.6339628\ttest: 62873.1248338\tbest: 62873.1248338 (336)\ttotal: 1.89s\tremaining: 3.71s\n",
      "337:\tlearn: 57629.5120704\ttest: 62881.3401887\tbest: 62873.1248338 (336)\ttotal: 1.9s\tremaining: 3.71s\n",
      "338:\tlearn: 57602.6182293\ttest: 62883.2449011\tbest: 62873.1248338 (336)\ttotal: 1.9s\tremaining: 3.71s\n",
      "339:\tlearn: 57572.4603011\ttest: 62846.2813707\tbest: 62846.2813707 (339)\ttotal: 1.91s\tremaining: 3.71s\n",
      "340:\tlearn: 57546.8795660\ttest: 62833.5852231\tbest: 62833.5852231 (340)\ttotal: 1.92s\tremaining: 3.71s\n",
      "341:\tlearn: 57515.4854500\ttest: 62832.0669418\tbest: 62832.0669418 (341)\ttotal: 1.92s\tremaining: 3.7s\n",
      "342:\tlearn: 57450.5274016\ttest: 62805.8662771\tbest: 62805.8662771 (342)\ttotal: 1.93s\tremaining: 3.69s\n",
      "343:\tlearn: 57378.8660305\ttest: 62755.8892511\tbest: 62755.8892511 (343)\ttotal: 1.93s\tremaining: 3.69s\n",
      "344:\tlearn: 57290.4174077\ttest: 62717.3120909\tbest: 62717.3120909 (344)\ttotal: 1.94s\tremaining: 3.69s\n",
      "345:\tlearn: 57249.1050332\ttest: 62697.9799209\tbest: 62697.9799209 (345)\ttotal: 1.95s\tremaining: 3.68s\n",
      "346:\tlearn: 57234.2627003\ttest: 62692.7395119\tbest: 62692.7395119 (346)\ttotal: 1.95s\tremaining: 3.68s\n",
      "347:\tlearn: 57181.9978350\ttest: 62665.3181789\tbest: 62665.3181789 (347)\ttotal: 1.96s\tremaining: 3.67s\n",
      "348:\tlearn: 57158.9946642\ttest: 62664.5642952\tbest: 62664.5642952 (348)\ttotal: 1.97s\tremaining: 3.67s\n",
      "349:\tlearn: 57130.6150131\ttest: 62652.7513294\tbest: 62652.7513294 (349)\ttotal: 1.97s\tremaining: 3.66s\n",
      "350:\tlearn: 57027.5598429\ttest: 62584.6710500\tbest: 62584.6710500 (350)\ttotal: 1.98s\tremaining: 3.65s\n",
      "351:\tlearn: 56999.0976433\ttest: 62585.0551283\tbest: 62584.6710500 (350)\ttotal: 1.98s\tremaining: 3.65s\n",
      "352:\tlearn: 56991.3136547\ttest: 62581.5484805\tbest: 62581.5484805 (352)\ttotal: 1.99s\tremaining: 3.64s\n",
      "353:\tlearn: 56952.0692719\ttest: 62591.9903649\tbest: 62581.5484805 (352)\ttotal: 1.99s\tremaining: 3.64s\n",
      "354:\tlearn: 56872.3964850\ttest: 62550.1796237\tbest: 62550.1796237 (354)\ttotal: 2s\tremaining: 3.63s\n",
      "355:\tlearn: 56826.7299836\ttest: 62537.1123279\tbest: 62537.1123279 (355)\ttotal: 2.01s\tremaining: 3.63s\n",
      "356:\tlearn: 56792.9097987\ttest: 62501.6534881\tbest: 62501.6534881 (356)\ttotal: 2.02s\tremaining: 3.63s\n",
      "357:\tlearn: 56699.3189454\ttest: 62427.1785130\tbest: 62427.1785130 (357)\ttotal: 2.02s\tremaining: 3.63s\n",
      "358:\tlearn: 56657.3301604\ttest: 62425.6947531\tbest: 62425.6947531 (358)\ttotal: 2.03s\tremaining: 3.62s\n",
      "359:\tlearn: 56614.5238906\ttest: 62421.6422553\tbest: 62421.6422553 (359)\ttotal: 2.03s\tremaining: 3.61s\n",
      "360:\tlearn: 56564.3173225\ttest: 62424.2701791\tbest: 62421.6422553 (359)\ttotal: 2.04s\tremaining: 3.61s\n",
      "361:\tlearn: 56564.1271030\ttest: 62424.4402262\tbest: 62421.6422553 (359)\ttotal: 2.04s\tremaining: 3.6s\n",
      "362:\tlearn: 56563.7940106\ttest: 62424.4528069\tbest: 62421.6422553 (359)\ttotal: 2.05s\tremaining: 3.59s\n",
      "363:\tlearn: 56545.0452692\ttest: 62418.0864047\tbest: 62418.0864047 (363)\ttotal: 2.05s\tremaining: 3.59s\n",
      "364:\tlearn: 56510.8910888\ttest: 62423.7729292\tbest: 62418.0864047 (363)\ttotal: 2.06s\tremaining: 3.58s\n",
      "365:\tlearn: 56486.2796570\ttest: 62403.9314176\tbest: 62403.9314176 (365)\ttotal: 2.06s\tremaining: 3.58s\n",
      "366:\tlearn: 56452.9404530\ttest: 62411.0016335\tbest: 62403.9314176 (365)\ttotal: 2.07s\tremaining: 3.57s\n",
      "367:\tlearn: 56419.0766074\ttest: 62404.6999562\tbest: 62403.9314176 (365)\ttotal: 2.08s\tremaining: 3.58s\n",
      "368:\tlearn: 56394.2367155\ttest: 62395.4144614\tbest: 62395.4144614 (368)\ttotal: 2.09s\tremaining: 3.57s\n",
      "369:\tlearn: 56279.9762293\ttest: 62358.1767622\tbest: 62358.1767622 (369)\ttotal: 2.09s\tremaining: 3.57s\n",
      "370:\tlearn: 56244.7135261\ttest: 62349.6985730\tbest: 62349.6985730 (370)\ttotal: 2.1s\tremaining: 3.56s\n",
      "371:\tlearn: 56227.3649352\ttest: 62336.7869476\tbest: 62336.7869476 (371)\ttotal: 2.11s\tremaining: 3.56s\n",
      "372:\tlearn: 56190.6683372\ttest: 62326.7857399\tbest: 62326.7857399 (372)\ttotal: 2.11s\tremaining: 3.55s\n",
      "373:\tlearn: 56181.4416411\ttest: 62311.8942554\tbest: 62311.8942554 (373)\ttotal: 2.12s\tremaining: 3.54s\n",
      "374:\tlearn: 56180.1271578\ttest: 62310.5737195\tbest: 62310.5737195 (374)\ttotal: 2.12s\tremaining: 3.54s\n",
      "375:\tlearn: 56156.4800752\ttest: 62304.8387171\tbest: 62304.8387171 (375)\ttotal: 2.13s\tremaining: 3.54s\n",
      "376:\tlearn: 56155.1407788\ttest: 62305.3369683\tbest: 62304.8387171 (375)\ttotal: 2.14s\tremaining: 3.54s\n",
      "377:\tlearn: 56106.1239695\ttest: 62282.8442732\tbest: 62282.8442732 (377)\ttotal: 2.15s\tremaining: 3.53s\n",
      "378:\tlearn: 56099.6044523\ttest: 62280.4577966\tbest: 62280.4577966 (378)\ttotal: 2.15s\tremaining: 3.52s\n",
      "379:\tlearn: 56068.9407895\ttest: 62282.7468693\tbest: 62280.4577966 (378)\ttotal: 2.15s\tremaining: 3.51s\n",
      "380:\tlearn: 56028.3068414\ttest: 62262.0665943\tbest: 62262.0665943 (380)\ttotal: 2.16s\tremaining: 3.5s\n",
      "381:\tlearn: 56022.9481993\ttest: 62255.2413976\tbest: 62255.2413976 (381)\ttotal: 2.16s\tremaining: 3.5s\n",
      "382:\tlearn: 56019.4898483\ttest: 62254.7325200\tbest: 62254.7325200 (382)\ttotal: 2.16s\tremaining: 3.49s\n",
      "383:\tlearn: 55974.9458180\ttest: 62244.0927585\tbest: 62244.0927585 (383)\ttotal: 2.17s\tremaining: 3.48s\n",
      "384:\tlearn: 55949.0233666\ttest: 62245.4725190\tbest: 62244.0927585 (383)\ttotal: 2.17s\tremaining: 3.48s\n",
      "385:\tlearn: 55915.1307852\ttest: 62233.6142174\tbest: 62233.6142174 (385)\ttotal: 2.18s\tremaining: 3.47s\n",
      "386:\tlearn: 55865.8655631\ttest: 62226.7393031\tbest: 62226.7393031 (386)\ttotal: 2.19s\tremaining: 3.46s\n",
      "387:\tlearn: 55825.1480746\ttest: 62220.3193920\tbest: 62220.3193920 (387)\ttotal: 2.19s\tremaining: 3.46s\n",
      "388:\tlearn: 55763.4540665\ttest: 62221.2180070\tbest: 62220.3193920 (387)\ttotal: 2.2s\tremaining: 3.46s\n",
      "389:\tlearn: 55718.8503448\ttest: 62211.5893506\tbest: 62211.5893506 (389)\ttotal: 2.21s\tremaining: 3.45s\n",
      "390:\tlearn: 55718.6783240\ttest: 62213.6940070\tbest: 62211.5893506 (389)\ttotal: 2.21s\tremaining: 3.45s\n",
      "391:\tlearn: 55707.7379752\ttest: 62212.9144862\tbest: 62211.5893506 (389)\ttotal: 2.22s\tremaining: 3.45s\n",
      "392:\tlearn: 55694.0249787\ttest: 62203.5140682\tbest: 62203.5140682 (392)\ttotal: 2.23s\tremaining: 3.44s\n",
      "393:\tlearn: 55689.8470433\ttest: 62204.7880937\tbest: 62203.5140682 (392)\ttotal: 2.24s\tremaining: 3.44s\n",
      "394:\tlearn: 55687.4841070\ttest: 62205.9004594\tbest: 62203.5140682 (392)\ttotal: 2.24s\tremaining: 3.44s\n",
      "395:\tlearn: 55685.7096441\ttest: 62208.7213882\tbest: 62203.5140682 (392)\ttotal: 2.25s\tremaining: 3.43s\n",
      "396:\tlearn: 55580.5081977\ttest: 62178.9783220\tbest: 62178.9783220 (396)\ttotal: 2.26s\tremaining: 3.43s\n",
      "397:\tlearn: 55569.6451297\ttest: 62178.8827594\tbest: 62178.8827594 (397)\ttotal: 2.26s\tremaining: 3.42s\n",
      "398:\tlearn: 55472.5654284\ttest: 62104.6129163\tbest: 62104.6129163 (398)\ttotal: 2.27s\tremaining: 3.42s\n",
      "399:\tlearn: 55463.0980551\ttest: 62097.6259792\tbest: 62097.6259792 (399)\ttotal: 2.27s\tremaining: 3.41s\n",
      "400:\tlearn: 55461.9574958\ttest: 62098.0519750\tbest: 62097.6259792 (399)\ttotal: 2.28s\tremaining: 3.41s\n",
      "401:\tlearn: 55394.2568995\ttest: 62071.9412460\tbest: 62071.9412460 (401)\ttotal: 2.29s\tremaining: 3.4s\n",
      "402:\tlearn: 55354.2742436\ttest: 62057.1235537\tbest: 62057.1235537 (402)\ttotal: 2.29s\tremaining: 3.4s\n",
      "403:\tlearn: 55346.0482227\ttest: 62062.7778202\tbest: 62057.1235537 (402)\ttotal: 2.3s\tremaining: 3.39s\n",
      "404:\tlearn: 55317.4139817\ttest: 62059.6801006\tbest: 62057.1235537 (402)\ttotal: 2.3s\tremaining: 3.38s\n",
      "405:\tlearn: 55283.0741719\ttest: 62076.4282670\tbest: 62057.1235537 (402)\ttotal: 2.31s\tremaining: 3.38s\n",
      "406:\tlearn: 55281.3440995\ttest: 62077.8450393\tbest: 62057.1235537 (402)\ttotal: 2.32s\tremaining: 3.38s\n",
      "407:\tlearn: 55281.3193126\ttest: 62078.1439093\tbest: 62057.1235537 (402)\ttotal: 2.32s\tremaining: 3.37s\n",
      "408:\tlearn: 55258.5034919\ttest: 62071.5396671\tbest: 62057.1235537 (402)\ttotal: 2.33s\tremaining: 3.36s\n",
      "409:\tlearn: 55180.2033378\ttest: 62048.7242479\tbest: 62048.7242479 (409)\ttotal: 2.33s\tremaining: 3.35s\n",
      "410:\tlearn: 55179.0918970\ttest: 62048.9167612\tbest: 62048.7242479 (409)\ttotal: 2.34s\tremaining: 3.35s\n",
      "411:\tlearn: 55143.6323470\ttest: 62059.5036273\tbest: 62048.7242479 (409)\ttotal: 2.34s\tremaining: 3.35s\n",
      "412:\tlearn: 55111.6042701\ttest: 62057.8588962\tbest: 62048.7242479 (409)\ttotal: 2.35s\tremaining: 3.34s\n",
      "413:\tlearn: 55104.3790636\ttest: 62070.7850514\tbest: 62048.7242479 (409)\ttotal: 2.35s\tremaining: 3.33s\n",
      "414:\tlearn: 55068.5366673\ttest: 62040.7614351\tbest: 62040.7614351 (414)\ttotal: 2.36s\tremaining: 3.33s\n",
      "415:\tlearn: 55066.8723107\ttest: 62040.4798467\tbest: 62040.4798467 (415)\ttotal: 2.37s\tremaining: 3.32s\n",
      "416:\tlearn: 55066.5154481\ttest: 62041.1009732\tbest: 62040.4798467 (415)\ttotal: 2.37s\tremaining: 3.32s\n",
      "417:\tlearn: 55037.6426928\ttest: 62021.4597716\tbest: 62021.4597716 (417)\ttotal: 2.38s\tremaining: 3.31s\n",
      "418:\tlearn: 55007.2047924\ttest: 62013.1507445\tbest: 62013.1507445 (418)\ttotal: 2.38s\tremaining: 3.3s\n",
      "419:\tlearn: 54965.5296322\ttest: 61992.9034469\tbest: 61992.9034469 (419)\ttotal: 2.38s\tremaining: 3.29s\n",
      "420:\tlearn: 54868.3309684\ttest: 61970.5567307\tbest: 61970.5567307 (420)\ttotal: 2.38s\tremaining: 3.28s\n",
      "421:\tlearn: 54842.8913119\ttest: 61964.1390398\tbest: 61964.1390398 (421)\ttotal: 2.39s\tremaining: 3.28s\n",
      "422:\tlearn: 54842.0426661\ttest: 61966.2393359\tbest: 61964.1390398 (421)\ttotal: 2.4s\tremaining: 3.27s\n",
      "423:\tlearn: 54740.7564286\ttest: 61903.4073477\tbest: 61903.4073477 (423)\ttotal: 2.4s\tremaining: 3.26s\n",
      "424:\tlearn: 54694.0988578\ttest: 61892.4152031\tbest: 61892.4152031 (424)\ttotal: 2.41s\tremaining: 3.26s\n",
      "425:\tlearn: 54655.8821000\ttest: 61882.7892103\tbest: 61882.7892103 (425)\ttotal: 2.42s\tremaining: 3.25s\n",
      "426:\tlearn: 54620.6856575\ttest: 61862.0650605\tbest: 61862.0650605 (426)\ttotal: 2.42s\tremaining: 3.25s\n",
      "427:\tlearn: 54614.0872948\ttest: 61868.8397560\tbest: 61862.0650605 (426)\ttotal: 2.43s\tremaining: 3.24s\n",
      "428:\tlearn: 54606.5613214\ttest: 61881.4390810\tbest: 61862.0650605 (426)\ttotal: 2.43s\tremaining: 3.24s\n",
      "429:\tlearn: 54586.4269966\ttest: 61865.9767837\tbest: 61862.0650605 (426)\ttotal: 2.44s\tremaining: 3.23s\n",
      "430:\tlearn: 54560.0947922\ttest: 61834.0822859\tbest: 61834.0822859 (430)\ttotal: 2.44s\tremaining: 3.23s\n",
      "431:\tlearn: 54526.7815974\ttest: 61825.2126633\tbest: 61825.2126633 (431)\ttotal: 2.45s\tremaining: 3.22s\n",
      "432:\tlearn: 54450.5332385\ttest: 61776.0037834\tbest: 61776.0037834 (432)\ttotal: 2.46s\tremaining: 3.21s\n",
      "433:\tlearn: 54444.7179183\ttest: 61784.4454985\tbest: 61776.0037834 (432)\ttotal: 2.46s\tremaining: 3.21s\n",
      "434:\tlearn: 54415.4283482\ttest: 61777.2490206\tbest: 61776.0037834 (432)\ttotal: 2.47s\tremaining: 3.2s\n",
      "435:\tlearn: 54401.0172350\ttest: 61762.8745487\tbest: 61762.8745487 (435)\ttotal: 2.47s\tremaining: 3.2s\n",
      "436:\tlearn: 54314.6897237\ttest: 61731.5319202\tbest: 61731.5319202 (436)\ttotal: 2.48s\tremaining: 3.19s\n",
      "437:\tlearn: 54312.1352403\ttest: 61731.6556660\tbest: 61731.5319202 (436)\ttotal: 2.48s\tremaining: 3.19s\n",
      "438:\tlearn: 54306.2303272\ttest: 61733.3061464\tbest: 61731.5319202 (436)\ttotal: 2.49s\tremaining: 3.18s\n",
      "439:\tlearn: 54295.7176601\ttest: 61750.0891966\tbest: 61731.5319202 (436)\ttotal: 2.5s\tremaining: 3.18s\n",
      "440:\tlearn: 54279.9395765\ttest: 61748.1728929\tbest: 61731.5319202 (436)\ttotal: 2.5s\tremaining: 3.17s\n",
      "441:\tlearn: 54277.4845015\ttest: 61748.2737600\tbest: 61731.5319202 (436)\ttotal: 2.51s\tremaining: 3.17s\n",
      "442:\tlearn: 54256.9149402\ttest: 61748.1332727\tbest: 61731.5319202 (436)\ttotal: 2.51s\tremaining: 3.16s\n",
      "443:\tlearn: 54250.4547067\ttest: 61751.0663848\tbest: 61731.5319202 (436)\ttotal: 2.52s\tremaining: 3.16s\n",
      "444:\tlearn: 54184.7289779\ttest: 61724.3322091\tbest: 61724.3322091 (444)\ttotal: 2.52s\tremaining: 3.15s\n",
      "445:\tlearn: 54160.8624927\ttest: 61727.6503558\tbest: 61724.3322091 (444)\ttotal: 2.53s\tremaining: 3.14s\n",
      "446:\tlearn: 54158.3806771\ttest: 61727.0605414\tbest: 61724.3322091 (444)\ttotal: 2.53s\tremaining: 3.14s\n",
      "447:\tlearn: 54086.3612272\ttest: 61677.5774484\tbest: 61677.5774484 (447)\ttotal: 2.54s\tremaining: 3.13s\n",
      "448:\tlearn: 54084.0867544\ttest: 61677.5101717\tbest: 61677.5101717 (448)\ttotal: 2.54s\tremaining: 3.12s\n",
      "449:\tlearn: 54027.9942745\ttest: 61634.6089394\tbest: 61634.6089394 (449)\ttotal: 2.55s\tremaining: 3.11s\n",
      "450:\tlearn: 53988.0041900\ttest: 61632.6879381\tbest: 61632.6879381 (450)\ttotal: 2.56s\tremaining: 3.11s\n",
      "451:\tlearn: 53957.6684654\ttest: 61605.6004233\tbest: 61605.6004233 (451)\ttotal: 2.56s\tremaining: 3.1s\n",
      "452:\tlearn: 53900.8084416\ttest: 61589.0818688\tbest: 61589.0818688 (452)\ttotal: 2.57s\tremaining: 3.1s\n",
      "453:\tlearn: 53868.8817336\ttest: 61570.3533928\tbest: 61570.3533928 (453)\ttotal: 2.58s\tremaining: 3.1s\n",
      "454:\tlearn: 53842.0732606\ttest: 61569.8266169\tbest: 61569.8266169 (454)\ttotal: 2.59s\tremaining: 3.1s\n",
      "455:\tlearn: 53827.1788222\ttest: 61573.3033735\tbest: 61569.8266169 (454)\ttotal: 2.59s\tremaining: 3.1s\n",
      "456:\tlearn: 53816.1063417\ttest: 61568.1793712\tbest: 61568.1793712 (456)\ttotal: 2.6s\tremaining: 3.09s\n",
      "457:\tlearn: 53756.7292057\ttest: 61551.8379526\tbest: 61551.8379526 (457)\ttotal: 2.61s\tremaining: 3.09s\n",
      "458:\tlearn: 53711.0632056\ttest: 61530.4050761\tbest: 61530.4050761 (458)\ttotal: 2.62s\tremaining: 3.09s\n",
      "459:\tlearn: 53696.2541716\ttest: 61523.9138974\tbest: 61523.9138974 (459)\ttotal: 2.63s\tremaining: 3.08s\n",
      "460:\tlearn: 53633.7638375\ttest: 61498.7933279\tbest: 61498.7933279 (460)\ttotal: 2.63s\tremaining: 3.08s\n",
      "461:\tlearn: 53631.3550987\ttest: 61499.4224595\tbest: 61498.7933279 (460)\ttotal: 2.64s\tremaining: 3.08s\n",
      "462:\tlearn: 53623.2186178\ttest: 61503.7789522\tbest: 61498.7933279 (460)\ttotal: 2.65s\tremaining: 3.07s\n",
      "463:\tlearn: 53602.9544712\ttest: 61519.1945572\tbest: 61498.7933279 (460)\ttotal: 2.66s\tremaining: 3.07s\n",
      "464:\tlearn: 53584.6768815\ttest: 61509.2975995\tbest: 61498.7933279 (460)\ttotal: 2.66s\tremaining: 3.06s\n",
      "465:\tlearn: 53536.3267020\ttest: 61502.3079161\tbest: 61498.7933279 (460)\ttotal: 2.67s\tremaining: 3.06s\n",
      "466:\tlearn: 53524.1791029\ttest: 61498.0196828\tbest: 61498.0196828 (466)\ttotal: 2.68s\tremaining: 3.06s\n",
      "467:\tlearn: 53518.8648985\ttest: 61499.4338384\tbest: 61498.0196828 (466)\ttotal: 2.69s\tremaining: 3.05s\n",
      "468:\tlearn: 53486.9014504\ttest: 61493.1781457\tbest: 61493.1781457 (468)\ttotal: 2.69s\tremaining: 3.05s\n",
      "469:\tlearn: 53477.3316519\ttest: 61496.2073434\tbest: 61493.1781457 (468)\ttotal: 2.7s\tremaining: 3.04s\n",
      "470:\tlearn: 53452.5095064\ttest: 61499.5085140\tbest: 61493.1781457 (468)\ttotal: 2.7s\tremaining: 3.04s\n",
      "471:\tlearn: 53400.8493620\ttest: 61512.8799774\tbest: 61493.1781457 (468)\ttotal: 2.71s\tremaining: 3.03s\n",
      "472:\tlearn: 53383.4826886\ttest: 61509.0220159\tbest: 61493.1781457 (468)\ttotal: 2.72s\tremaining: 3.03s\n",
      "473:\tlearn: 53353.8261229\ttest: 61514.2550357\tbest: 61493.1781457 (468)\ttotal: 2.73s\tremaining: 3.03s\n",
      "474:\tlearn: 53328.3859948\ttest: 61484.1771666\tbest: 61484.1771666 (474)\ttotal: 2.73s\tremaining: 3.02s\n",
      "475:\tlearn: 53314.8216597\ttest: 61467.4645000\tbest: 61467.4645000 (475)\ttotal: 2.74s\tremaining: 3.02s\n",
      "476:\tlearn: 53298.3724095\ttest: 61467.5020966\tbest: 61467.4645000 (475)\ttotal: 2.75s\tremaining: 3.01s\n",
      "477:\tlearn: 53271.1652281\ttest: 61477.4827524\tbest: 61467.4645000 (475)\ttotal: 2.75s\tremaining: 3.01s\n",
      "478:\tlearn: 53242.6011892\ttest: 61488.2916734\tbest: 61467.4645000 (475)\ttotal: 2.76s\tremaining: 3s\n",
      "479:\tlearn: 53224.2636917\ttest: 61489.9253960\tbest: 61467.4645000 (475)\ttotal: 2.77s\tremaining: 3s\n",
      "480:\tlearn: 53184.2446724\ttest: 61500.7415528\tbest: 61467.4645000 (475)\ttotal: 2.77s\tremaining: 2.99s\n",
      "481:\tlearn: 53147.5423603\ttest: 61510.6051492\tbest: 61467.4645000 (475)\ttotal: 2.78s\tremaining: 2.98s\n",
      "482:\tlearn: 53125.2658791\ttest: 61518.9049493\tbest: 61467.4645000 (475)\ttotal: 2.78s\tremaining: 2.98s\n",
      "483:\tlearn: 53119.1415075\ttest: 61514.4399851\tbest: 61467.4645000 (475)\ttotal: 2.79s\tremaining: 2.97s\n",
      "484:\tlearn: 53077.2210669\ttest: 61495.7225069\tbest: 61467.4645000 (475)\ttotal: 2.8s\tremaining: 2.97s\n",
      "485:\tlearn: 53022.1568379\ttest: 61452.2577373\tbest: 61452.2577373 (485)\ttotal: 2.8s\tremaining: 2.96s\n",
      "486:\tlearn: 52989.3432487\ttest: 61425.3976616\tbest: 61425.3976616 (486)\ttotal: 2.81s\tremaining: 2.96s\n",
      "487:\tlearn: 52961.6476677\ttest: 61422.5809676\tbest: 61422.5809676 (487)\ttotal: 2.81s\tremaining: 2.95s\n",
      "488:\tlearn: 52892.7687775\ttest: 61426.4360242\tbest: 61422.5809676 (487)\ttotal: 2.82s\tremaining: 2.95s\n",
      "489:\tlearn: 52876.5539831\ttest: 61423.9450034\tbest: 61422.5809676 (487)\ttotal: 2.83s\tremaining: 2.94s\n",
      "490:\tlearn: 52872.7414647\ttest: 61416.8282395\tbest: 61416.8282395 (490)\ttotal: 2.83s\tremaining: 2.94s\n",
      "491:\tlearn: 52807.5260365\ttest: 61389.7050977\tbest: 61389.7050977 (491)\ttotal: 2.84s\tremaining: 2.93s\n",
      "492:\tlearn: 52767.2539236\ttest: 61382.1303869\tbest: 61382.1303869 (492)\ttotal: 2.85s\tremaining: 2.93s\n",
      "493:\tlearn: 52765.9685452\ttest: 61378.6646119\tbest: 61378.6646119 (493)\ttotal: 2.85s\tremaining: 2.92s\n",
      "494:\tlearn: 52755.6959942\ttest: 61383.0033951\tbest: 61378.6646119 (493)\ttotal: 2.86s\tremaining: 2.92s\n",
      "495:\tlearn: 52712.1962906\ttest: 61399.1483578\tbest: 61378.6646119 (493)\ttotal: 2.87s\tremaining: 2.91s\n",
      "496:\tlearn: 52702.8031315\ttest: 61395.5712585\tbest: 61378.6646119 (493)\ttotal: 2.87s\tremaining: 2.91s\n",
      "497:\tlearn: 52686.6350785\ttest: 61392.8108166\tbest: 61378.6646119 (493)\ttotal: 2.88s\tremaining: 2.9s\n",
      "498:\tlearn: 52666.9638132\ttest: 61399.4672437\tbest: 61378.6646119 (493)\ttotal: 2.88s\tremaining: 2.9s\n",
      "499:\tlearn: 52624.9827055\ttest: 61396.3254881\tbest: 61378.6646119 (493)\ttotal: 2.89s\tremaining: 2.89s\n",
      "500:\tlearn: 52610.9425853\ttest: 61397.1246430\tbest: 61378.6646119 (493)\ttotal: 2.9s\tremaining: 2.89s\n",
      "501:\tlearn: 52563.7378959\ttest: 61352.5362662\tbest: 61352.5362662 (501)\ttotal: 2.9s\tremaining: 2.88s\n",
      "502:\tlearn: 52547.7850579\ttest: 61350.8583939\tbest: 61350.8583939 (502)\ttotal: 2.91s\tremaining: 2.87s\n",
      "503:\tlearn: 52517.6538842\ttest: 61334.5546616\tbest: 61334.5546616 (503)\ttotal: 2.91s\tremaining: 2.87s\n",
      "504:\tlearn: 52496.6787730\ttest: 61326.7302358\tbest: 61326.7302358 (504)\ttotal: 2.92s\tremaining: 2.86s\n",
      "505:\tlearn: 52475.2163580\ttest: 61327.5466381\tbest: 61326.7302358 (504)\ttotal: 2.93s\tremaining: 2.86s\n",
      "506:\tlearn: 52430.9926048\ttest: 61348.4118418\tbest: 61326.7302358 (504)\ttotal: 2.93s\tremaining: 2.85s\n",
      "507:\tlearn: 52399.4561442\ttest: 61331.8012428\tbest: 61326.7302358 (504)\ttotal: 2.94s\tremaining: 2.85s\n",
      "508:\tlearn: 52372.0878572\ttest: 61339.9021192\tbest: 61326.7302358 (504)\ttotal: 2.94s\tremaining: 2.84s\n",
      "509:\tlearn: 52331.7940760\ttest: 61348.1757671\tbest: 61326.7302358 (504)\ttotal: 2.95s\tremaining: 2.83s\n",
      "510:\tlearn: 52307.4768996\ttest: 61335.8647020\tbest: 61326.7302358 (504)\ttotal: 2.96s\tremaining: 2.83s\n",
      "511:\tlearn: 52292.7777423\ttest: 61340.3451240\tbest: 61326.7302358 (504)\ttotal: 2.96s\tremaining: 2.82s\n",
      "512:\tlearn: 52288.1912573\ttest: 61335.6652248\tbest: 61326.7302358 (504)\ttotal: 2.97s\tremaining: 2.82s\n",
      "513:\tlearn: 52281.6749820\ttest: 61332.9403332\tbest: 61326.7302358 (504)\ttotal: 2.98s\tremaining: 2.81s\n",
      "514:\tlearn: 52259.1310007\ttest: 61337.6933770\tbest: 61326.7302358 (504)\ttotal: 2.98s\tremaining: 2.81s\n",
      "515:\tlearn: 52247.9157374\ttest: 61345.1706437\tbest: 61326.7302358 (504)\ttotal: 2.99s\tremaining: 2.81s\n",
      "516:\tlearn: 52233.7232217\ttest: 61336.8206188\tbest: 61326.7302358 (504)\ttotal: 3s\tremaining: 2.8s\n",
      "517:\tlearn: 52217.5171872\ttest: 61331.5109942\tbest: 61326.7302358 (504)\ttotal: 3.01s\tremaining: 2.8s\n",
      "518:\tlearn: 52175.9551460\ttest: 61307.5711871\tbest: 61307.5711871 (518)\ttotal: 3.01s\tremaining: 2.79s\n",
      "519:\tlearn: 52162.9822610\ttest: 61300.6673394\tbest: 61300.6673394 (519)\ttotal: 3.02s\tremaining: 2.79s\n",
      "520:\tlearn: 52104.2712370\ttest: 61284.7677173\tbest: 61284.7677173 (520)\ttotal: 3.03s\tremaining: 2.78s\n",
      "521:\tlearn: 52073.7842124\ttest: 61287.9315379\tbest: 61284.7677173 (520)\ttotal: 3.04s\tremaining: 2.79s\n",
      "522:\tlearn: 52020.9418193\ttest: 61265.4785429\tbest: 61265.4785429 (522)\ttotal: 3.05s\tremaining: 2.78s\n",
      "523:\tlearn: 51979.2310222\ttest: 61234.2804207\tbest: 61234.2804207 (523)\ttotal: 3.06s\tremaining: 2.78s\n",
      "524:\tlearn: 51974.7669237\ttest: 61231.4749431\tbest: 61231.4749431 (524)\ttotal: 3.07s\tremaining: 2.77s\n",
      "525:\tlearn: 51958.9751715\ttest: 61222.0163720\tbest: 61222.0163720 (525)\ttotal: 3.07s\tremaining: 2.77s\n",
      "526:\tlearn: 51937.2246373\ttest: 61223.7877704\tbest: 61222.0163720 (525)\ttotal: 3.08s\tremaining: 2.76s\n",
      "527:\tlearn: 51928.5927547\ttest: 61217.0083549\tbest: 61217.0083549 (527)\ttotal: 3.08s\tremaining: 2.76s\n",
      "528:\tlearn: 51867.6217482\ttest: 61196.3246837\tbest: 61196.3246837 (528)\ttotal: 3.09s\tremaining: 2.75s\n",
      "529:\tlearn: 51842.7024176\ttest: 61176.3341359\tbest: 61176.3341359 (529)\ttotal: 3.1s\tremaining: 2.75s\n",
      "530:\tlearn: 51830.4844188\ttest: 61183.0889651\tbest: 61176.3341359 (529)\ttotal: 3.1s\tremaining: 2.74s\n",
      "531:\tlearn: 51818.2299895\ttest: 61178.8083507\tbest: 61176.3341359 (529)\ttotal: 3.11s\tremaining: 2.73s\n",
      "532:\tlearn: 51811.1642691\ttest: 61187.5358293\tbest: 61176.3341359 (529)\ttotal: 3.11s\tremaining: 2.73s\n",
      "533:\tlearn: 51800.0323763\ttest: 61190.3420643\tbest: 61176.3341359 (529)\ttotal: 3.12s\tremaining: 2.72s\n",
      "534:\tlearn: 51722.4864855\ttest: 61145.2290618\tbest: 61145.2290618 (534)\ttotal: 3.13s\tremaining: 2.72s\n",
      "535:\tlearn: 51717.7297007\ttest: 61146.1992687\tbest: 61145.2290618 (534)\ttotal: 3.13s\tremaining: 2.71s\n",
      "536:\tlearn: 51667.5649196\ttest: 61158.4374196\tbest: 61145.2290618 (534)\ttotal: 3.14s\tremaining: 2.7s\n",
      "537:\tlearn: 51613.3954627\ttest: 61123.5377977\tbest: 61123.5377977 (537)\ttotal: 3.14s\tremaining: 2.7s\n",
      "538:\tlearn: 51581.7806931\ttest: 61113.6085669\tbest: 61113.6085669 (538)\ttotal: 3.15s\tremaining: 2.69s\n",
      "539:\tlearn: 51558.3666818\ttest: 61108.9336730\tbest: 61108.9336730 (539)\ttotal: 3.15s\tremaining: 2.69s\n",
      "540:\tlearn: 51546.6327559\ttest: 61110.2095057\tbest: 61108.9336730 (539)\ttotal: 3.16s\tremaining: 2.68s\n",
      "541:\tlearn: 51545.8705582\ttest: 61112.6311589\tbest: 61108.9336730 (539)\ttotal: 3.16s\tremaining: 2.67s\n",
      "542:\tlearn: 51526.0592902\ttest: 61121.6615964\tbest: 61108.9336730 (539)\ttotal: 3.17s\tremaining: 2.67s\n",
      "543:\tlearn: 51485.1023293\ttest: 61098.8938881\tbest: 61098.8938881 (543)\ttotal: 3.18s\tremaining: 2.66s\n",
      "544:\tlearn: 51419.5470449\ttest: 61073.7352191\tbest: 61073.7352191 (544)\ttotal: 3.19s\tremaining: 2.66s\n",
      "545:\tlearn: 51418.5738043\ttest: 61074.0995557\tbest: 61073.7352191 (544)\ttotal: 3.19s\tremaining: 2.65s\n",
      "546:\tlearn: 51385.7208556\ttest: 61077.8504699\tbest: 61073.7352191 (544)\ttotal: 3.2s\tremaining: 2.65s\n",
      "547:\tlearn: 51367.5031210\ttest: 61081.7743964\tbest: 61073.7352191 (544)\ttotal: 3.21s\tremaining: 2.64s\n",
      "548:\tlearn: 51326.7468710\ttest: 61052.7586346\tbest: 61052.7586346 (548)\ttotal: 3.21s\tremaining: 2.64s\n",
      "549:\tlearn: 51326.0313339\ttest: 61054.3446596\tbest: 61052.7586346 (548)\ttotal: 3.22s\tremaining: 2.63s\n",
      "550:\tlearn: 51323.1618128\ttest: 61061.8314208\tbest: 61052.7586346 (548)\ttotal: 3.23s\tremaining: 2.63s\n",
      "551:\tlearn: 51318.1586924\ttest: 61065.6824038\tbest: 61052.7586346 (548)\ttotal: 3.23s\tremaining: 2.62s\n",
      "552:\tlearn: 51284.6731004\ttest: 61076.9211529\tbest: 61052.7586346 (548)\ttotal: 3.24s\tremaining: 2.62s\n",
      "553:\tlearn: 51244.2538500\ttest: 61051.6688492\tbest: 61051.6688492 (553)\ttotal: 3.25s\tremaining: 2.61s\n",
      "554:\tlearn: 51224.2505004\ttest: 61056.8114746\tbest: 61051.6688492 (553)\ttotal: 3.25s\tremaining: 2.61s\n",
      "555:\tlearn: 51185.8811848\ttest: 61067.6348016\tbest: 61051.6688492 (553)\ttotal: 3.26s\tremaining: 2.6s\n",
      "556:\tlearn: 51146.6853314\ttest: 61068.3314923\tbest: 61051.6688492 (553)\ttotal: 3.26s\tremaining: 2.6s\n",
      "557:\tlearn: 51123.7456082\ttest: 61062.3633045\tbest: 61051.6688492 (553)\ttotal: 3.27s\tremaining: 2.59s\n",
      "558:\tlearn: 51115.6306537\ttest: 61050.3841856\tbest: 61050.3841856 (558)\ttotal: 3.27s\tremaining: 2.58s\n",
      "559:\tlearn: 51086.7360154\ttest: 61030.7176714\tbest: 61030.7176714 (559)\ttotal: 3.28s\tremaining: 2.58s\n",
      "560:\tlearn: 51070.3617636\ttest: 61012.2615540\tbest: 61012.2615540 (560)\ttotal: 3.29s\tremaining: 2.57s\n",
      "561:\tlearn: 51061.3099317\ttest: 61015.2289256\tbest: 61012.2615540 (560)\ttotal: 3.29s\tremaining: 2.57s\n",
      "562:\tlearn: 51043.9532294\ttest: 61002.2322225\tbest: 61002.2322225 (562)\ttotal: 3.3s\tremaining: 2.56s\n",
      "563:\tlearn: 51040.1963992\ttest: 61000.6938544\tbest: 61000.6938544 (563)\ttotal: 3.31s\tremaining: 2.55s\n",
      "564:\tlearn: 51025.5512550\ttest: 61002.6206410\tbest: 61000.6938544 (563)\ttotal: 3.31s\tremaining: 2.55s\n",
      "565:\tlearn: 51019.4462640\ttest: 60999.2922408\tbest: 60999.2922408 (565)\ttotal: 3.32s\tremaining: 2.54s\n",
      "566:\tlearn: 50991.6221656\ttest: 60999.9637512\tbest: 60999.2922408 (565)\ttotal: 3.32s\tremaining: 2.54s\n",
      "567:\tlearn: 50988.3397903\ttest: 60992.9885096\tbest: 60992.9885096 (567)\ttotal: 3.33s\tremaining: 2.53s\n",
      "568:\tlearn: 50981.3690144\ttest: 60980.4315551\tbest: 60980.4315551 (568)\ttotal: 3.33s\tremaining: 2.52s\n",
      "569:\tlearn: 50941.8189301\ttest: 60976.9491106\tbest: 60976.9491106 (569)\ttotal: 3.34s\tremaining: 2.52s\n",
      "570:\tlearn: 50899.1452398\ttest: 60953.4945133\tbest: 60953.4945133 (570)\ttotal: 3.34s\tremaining: 2.51s\n",
      "571:\tlearn: 50890.1329660\ttest: 60960.9811531\tbest: 60953.4945133 (570)\ttotal: 3.35s\tremaining: 2.51s\n",
      "572:\tlearn: 50865.8209122\ttest: 60954.3603225\tbest: 60953.4945133 (570)\ttotal: 3.36s\tremaining: 2.5s\n",
      "573:\tlearn: 50845.0606452\ttest: 60943.7773721\tbest: 60943.7773721 (573)\ttotal: 3.36s\tremaining: 2.5s\n",
      "574:\tlearn: 50833.4885782\ttest: 60944.0000764\tbest: 60943.7773721 (573)\ttotal: 3.37s\tremaining: 2.49s\n",
      "575:\tlearn: 50824.7630325\ttest: 60948.0844773\tbest: 60943.7773721 (573)\ttotal: 3.37s\tremaining: 2.48s\n",
      "576:\tlearn: 50816.8785924\ttest: 60941.2514804\tbest: 60941.2514804 (576)\ttotal: 3.38s\tremaining: 2.48s\n",
      "577:\tlearn: 50812.7330135\ttest: 60940.4101285\tbest: 60940.4101285 (577)\ttotal: 3.39s\tremaining: 2.47s\n",
      "578:\tlearn: 50787.5849561\ttest: 60929.4339915\tbest: 60929.4339915 (578)\ttotal: 3.39s\tremaining: 2.47s\n",
      "579:\tlearn: 50771.7530924\ttest: 60940.7494243\tbest: 60929.4339915 (578)\ttotal: 3.4s\tremaining: 2.46s\n",
      "580:\tlearn: 50750.2517998\ttest: 60947.9494910\tbest: 60929.4339915 (578)\ttotal: 3.41s\tremaining: 2.46s\n",
      "581:\tlearn: 50737.3878656\ttest: 60945.4051618\tbest: 60929.4339915 (578)\ttotal: 3.42s\tremaining: 2.45s\n",
      "582:\tlearn: 50714.5724102\ttest: 60957.7474822\tbest: 60929.4339915 (578)\ttotal: 3.42s\tremaining: 2.45s\n",
      "583:\tlearn: 50691.4489647\ttest: 60960.1483446\tbest: 60929.4339915 (578)\ttotal: 3.43s\tremaining: 2.44s\n",
      "584:\tlearn: 50685.2815220\ttest: 60959.3741372\tbest: 60929.4339915 (578)\ttotal: 3.43s\tremaining: 2.44s\n",
      "585:\tlearn: 50659.7759367\ttest: 60969.4344397\tbest: 60929.4339915 (578)\ttotal: 3.44s\tremaining: 2.43s\n",
      "586:\tlearn: 50627.9464907\ttest: 60949.9612724\tbest: 60929.4339915 (578)\ttotal: 3.44s\tremaining: 2.42s\n",
      "587:\tlearn: 50618.2496079\ttest: 60946.1438130\tbest: 60929.4339915 (578)\ttotal: 3.45s\tremaining: 2.42s\n",
      "588:\tlearn: 50608.2238840\ttest: 60948.1763327\tbest: 60929.4339915 (578)\ttotal: 3.46s\tremaining: 2.41s\n",
      "589:\tlearn: 50567.1829196\ttest: 60956.5283379\tbest: 60929.4339915 (578)\ttotal: 3.46s\tremaining: 2.41s\n",
      "590:\tlearn: 50550.3782272\ttest: 60966.5414598\tbest: 60929.4339915 (578)\ttotal: 3.47s\tremaining: 2.4s\n",
      "591:\tlearn: 50548.5758371\ttest: 60966.7001400\tbest: 60929.4339915 (578)\ttotal: 3.48s\tremaining: 2.4s\n",
      "592:\tlearn: 50490.0940393\ttest: 60918.8730481\tbest: 60918.8730481 (592)\ttotal: 3.48s\tremaining: 2.39s\n",
      "593:\tlearn: 50454.2936076\ttest: 60892.2766502\tbest: 60892.2766502 (593)\ttotal: 3.49s\tremaining: 2.38s\n",
      "594:\tlearn: 50442.9726850\ttest: 60894.5621950\tbest: 60892.2766502 (593)\ttotal: 3.5s\tremaining: 2.38s\n",
      "595:\tlearn: 50432.2662718\ttest: 60882.9099563\tbest: 60882.9099563 (595)\ttotal: 3.5s\tremaining: 2.37s\n",
      "596:\tlearn: 50420.0573692\ttest: 60862.7684139\tbest: 60862.7684139 (596)\ttotal: 3.51s\tremaining: 2.37s\n",
      "597:\tlearn: 50378.2025589\ttest: 60856.6290837\tbest: 60856.6290837 (597)\ttotal: 3.51s\tremaining: 2.36s\n",
      "598:\tlearn: 50362.2329458\ttest: 60846.4719492\tbest: 60846.4719492 (598)\ttotal: 3.52s\tremaining: 2.35s\n",
      "599:\tlearn: 50323.9204535\ttest: 60837.5593734\tbest: 60837.5593734 (599)\ttotal: 3.52s\tremaining: 2.35s\n",
      "600:\tlearn: 50282.1927493\ttest: 60854.6899256\tbest: 60837.5593734 (599)\ttotal: 3.53s\tremaining: 2.34s\n",
      "601:\tlearn: 50279.0750854\ttest: 60855.3445612\tbest: 60837.5593734 (599)\ttotal: 3.53s\tremaining: 2.33s\n",
      "602:\tlearn: 50256.2569442\ttest: 60860.1325626\tbest: 60837.5593734 (599)\ttotal: 3.54s\tremaining: 2.33s\n",
      "603:\tlearn: 50254.1869168\ttest: 60861.3321067\tbest: 60837.5593734 (599)\ttotal: 3.54s\tremaining: 2.32s\n",
      "604:\tlearn: 50230.7633373\ttest: 60852.9385846\tbest: 60837.5593734 (599)\ttotal: 3.55s\tremaining: 2.32s\n",
      "605:\tlearn: 50226.0730625\ttest: 60843.1058019\tbest: 60837.5593734 (599)\ttotal: 3.56s\tremaining: 2.31s\n",
      "606:\tlearn: 50172.8692265\ttest: 60835.7486555\tbest: 60835.7486555 (606)\ttotal: 3.56s\tremaining: 2.31s\n",
      "607:\tlearn: 50150.7258657\ttest: 60810.1126788\tbest: 60810.1126788 (607)\ttotal: 3.57s\tremaining: 2.3s\n",
      "608:\tlearn: 50142.6834964\ttest: 60810.1029521\tbest: 60810.1029521 (608)\ttotal: 3.58s\tremaining: 2.29s\n",
      "609:\tlearn: 50050.6151574\ttest: 60798.2789939\tbest: 60798.2789939 (609)\ttotal: 3.58s\tremaining: 2.29s\n",
      "610:\tlearn: 50025.8007958\ttest: 60794.7130976\tbest: 60794.7130976 (610)\ttotal: 3.59s\tremaining: 2.28s\n",
      "611:\tlearn: 50024.4657828\ttest: 60795.2353414\tbest: 60794.7130976 (610)\ttotal: 3.59s\tremaining: 2.28s\n",
      "612:\tlearn: 50020.8997722\ttest: 60792.6076291\tbest: 60792.6076291 (612)\ttotal: 3.6s\tremaining: 2.27s\n",
      "613:\tlearn: 50007.0612721\ttest: 60783.4667559\tbest: 60783.4667559 (613)\ttotal: 3.61s\tremaining: 2.27s\n",
      "614:\tlearn: 49965.0619183\ttest: 60790.2944290\tbest: 60783.4667559 (613)\ttotal: 3.61s\tremaining: 2.26s\n",
      "615:\tlearn: 49964.6711834\ttest: 60791.5097944\tbest: 60783.4667559 (613)\ttotal: 3.62s\tremaining: 2.26s\n",
      "616:\tlearn: 49925.7808342\ttest: 60801.9190408\tbest: 60783.4667559 (613)\ttotal: 3.63s\tremaining: 2.25s\n",
      "617:\tlearn: 49900.2561795\ttest: 60799.1335291\tbest: 60783.4667559 (613)\ttotal: 3.63s\tremaining: 2.25s\n",
      "618:\tlearn: 49898.3073035\ttest: 60796.0814900\tbest: 60783.4667559 (613)\ttotal: 3.64s\tremaining: 2.24s\n",
      "619:\tlearn: 49895.3002295\ttest: 60801.1856021\tbest: 60783.4667559 (613)\ttotal: 3.65s\tremaining: 2.24s\n",
      "620:\tlearn: 49890.1003625\ttest: 60800.9366750\tbest: 60783.4667559 (613)\ttotal: 3.66s\tremaining: 2.23s\n",
      "621:\tlearn: 49853.7177303\ttest: 60796.5696219\tbest: 60783.4667559 (613)\ttotal: 3.67s\tremaining: 2.23s\n",
      "622:\tlearn: 49837.2890979\ttest: 60796.6114049\tbest: 60783.4667559 (613)\ttotal: 3.67s\tremaining: 2.22s\n",
      "623:\tlearn: 49822.1703521\ttest: 60799.8940925\tbest: 60783.4667559 (613)\ttotal: 3.67s\tremaining: 2.21s\n",
      "624:\tlearn: 49820.1568163\ttest: 60796.1827473\tbest: 60783.4667559 (613)\ttotal: 3.68s\tremaining: 2.21s\n",
      "625:\tlearn: 49805.6723603\ttest: 60777.8112501\tbest: 60777.8112501 (625)\ttotal: 3.69s\tremaining: 2.2s\n",
      "626:\tlearn: 49796.4440762\ttest: 60759.9485687\tbest: 60759.9485687 (626)\ttotal: 3.69s\tremaining: 2.2s\n",
      "627:\tlearn: 49781.2355116\ttest: 60752.3334498\tbest: 60752.3334498 (627)\ttotal: 3.7s\tremaining: 2.19s\n",
      "628:\tlearn: 49780.2249258\ttest: 60750.8602492\tbest: 60750.8602492 (628)\ttotal: 3.7s\tremaining: 2.18s\n",
      "629:\tlearn: 49743.1632968\ttest: 60751.9049399\tbest: 60750.8602492 (628)\ttotal: 3.71s\tremaining: 2.18s\n",
      "630:\tlearn: 49731.4896677\ttest: 60754.3742723\tbest: 60750.8602492 (628)\ttotal: 3.71s\tremaining: 2.17s\n",
      "631:\tlearn: 49730.8267986\ttest: 60749.0769163\tbest: 60749.0769163 (631)\ttotal: 3.72s\tremaining: 2.17s\n",
      "632:\tlearn: 49707.2038759\ttest: 60732.7375546\tbest: 60732.7375546 (632)\ttotal: 3.73s\tremaining: 2.16s\n",
      "633:\tlearn: 49664.5663686\ttest: 60708.6537870\tbest: 60708.6537870 (633)\ttotal: 3.73s\tremaining: 2.15s\n",
      "634:\tlearn: 49639.7025450\ttest: 60698.3166903\tbest: 60698.3166903 (634)\ttotal: 3.74s\tremaining: 2.15s\n",
      "635:\tlearn: 49625.1441599\ttest: 60691.4131540\tbest: 60691.4131540 (635)\ttotal: 3.74s\tremaining: 2.14s\n",
      "636:\tlearn: 49606.2088679\ttest: 60707.1440637\tbest: 60691.4131540 (635)\ttotal: 3.75s\tremaining: 2.14s\n",
      "637:\tlearn: 49580.2520245\ttest: 60699.1809391\tbest: 60691.4131540 (635)\ttotal: 3.75s\tremaining: 2.13s\n",
      "638:\tlearn: 49570.5871627\ttest: 60693.7462237\tbest: 60691.4131540 (635)\ttotal: 3.76s\tremaining: 2.12s\n",
      "639:\tlearn: 49544.2285667\ttest: 60694.8162607\tbest: 60691.4131540 (635)\ttotal: 3.77s\tremaining: 2.12s\n",
      "640:\tlearn: 49522.3650776\ttest: 60698.0353841\tbest: 60691.4131540 (635)\ttotal: 3.77s\tremaining: 2.11s\n",
      "641:\tlearn: 49499.8281875\ttest: 60698.4162144\tbest: 60691.4131540 (635)\ttotal: 3.78s\tremaining: 2.11s\n",
      "642:\tlearn: 49489.0309484\ttest: 60694.5168926\tbest: 60691.4131540 (635)\ttotal: 3.79s\tremaining: 2.1s\n",
      "643:\tlearn: 49483.0037218\ttest: 60697.5342351\tbest: 60691.4131540 (635)\ttotal: 3.79s\tremaining: 2.1s\n",
      "644:\tlearn: 49454.4519536\ttest: 60696.9953767\tbest: 60691.4131540 (635)\ttotal: 3.8s\tremaining: 2.09s\n",
      "645:\tlearn: 49433.3103617\ttest: 60698.6194549\tbest: 60691.4131540 (635)\ttotal: 3.8s\tremaining: 2.08s\n",
      "646:\tlearn: 49402.7137684\ttest: 60699.2143656\tbest: 60691.4131540 (635)\ttotal: 3.81s\tremaining: 2.08s\n",
      "647:\tlearn: 49371.5984372\ttest: 60674.6825911\tbest: 60674.6825911 (647)\ttotal: 3.81s\tremaining: 2.07s\n",
      "648:\tlearn: 49366.7616448\ttest: 60674.2880640\tbest: 60674.2880640 (648)\ttotal: 3.82s\tremaining: 2.06s\n",
      "649:\tlearn: 49359.1414438\ttest: 60674.2555475\tbest: 60674.2555475 (649)\ttotal: 3.83s\tremaining: 2.06s\n",
      "650:\tlearn: 49344.6073234\ttest: 60681.2095989\tbest: 60674.2555475 (649)\ttotal: 3.83s\tremaining: 2.06s\n",
      "651:\tlearn: 49335.6625991\ttest: 60677.6157983\tbest: 60674.2555475 (649)\ttotal: 3.84s\tremaining: 2.05s\n",
      "652:\tlearn: 49321.7735440\ttest: 60676.2762988\tbest: 60674.2555475 (649)\ttotal: 3.85s\tremaining: 2.04s\n",
      "653:\tlearn: 49312.6609985\ttest: 60674.5799131\tbest: 60674.2555475 (649)\ttotal: 3.86s\tremaining: 2.04s\n",
      "654:\tlearn: 49310.4782731\ttest: 60674.6156296\tbest: 60674.2555475 (649)\ttotal: 3.86s\tremaining: 2.04s\n",
      "655:\tlearn: 49274.5521353\ttest: 60677.0427801\tbest: 60674.2555475 (649)\ttotal: 3.87s\tremaining: 2.03s\n",
      "656:\tlearn: 49259.5410022\ttest: 60658.9555369\tbest: 60658.9555369 (656)\ttotal: 3.88s\tremaining: 2.02s\n",
      "657:\tlearn: 49259.2008858\ttest: 60659.8620043\tbest: 60658.9555369 (656)\ttotal: 3.88s\tremaining: 2.02s\n",
      "658:\tlearn: 49235.8844714\ttest: 60642.7754955\tbest: 60642.7754955 (658)\ttotal: 3.89s\tremaining: 2.01s\n",
      "659:\tlearn: 49204.2098323\ttest: 60633.0882268\tbest: 60633.0882268 (659)\ttotal: 3.89s\tremaining: 2s\n",
      "660:\tlearn: 49195.3010819\ttest: 60636.7600583\tbest: 60633.0882268 (659)\ttotal: 3.9s\tremaining: 2s\n",
      "661:\tlearn: 49166.1065797\ttest: 60658.8976732\tbest: 60633.0882268 (659)\ttotal: 3.9s\tremaining: 1.99s\n",
      "662:\tlearn: 49134.2576705\ttest: 60650.4512388\tbest: 60633.0882268 (659)\ttotal: 3.91s\tremaining: 1.99s\n",
      "663:\tlearn: 49125.5872266\ttest: 60647.8868495\tbest: 60633.0882268 (659)\ttotal: 3.92s\tremaining: 1.98s\n",
      "664:\tlearn: 49125.3372315\ttest: 60648.5991675\tbest: 60633.0882268 (659)\ttotal: 3.92s\tremaining: 1.98s\n",
      "665:\tlearn: 49117.6970297\ttest: 60650.1519231\tbest: 60633.0882268 (659)\ttotal: 3.93s\tremaining: 1.97s\n",
      "666:\tlearn: 49113.7512897\ttest: 60643.9021917\tbest: 60633.0882268 (659)\ttotal: 3.93s\tremaining: 1.96s\n",
      "667:\tlearn: 49096.9587634\ttest: 60618.8325143\tbest: 60618.8325143 (667)\ttotal: 3.94s\tremaining: 1.96s\n",
      "668:\tlearn: 49009.4993423\ttest: 60575.8389798\tbest: 60575.8389798 (668)\ttotal: 3.94s\tremaining: 1.95s\n",
      "669:\tlearn: 48993.6472285\ttest: 60567.3775962\tbest: 60567.3775962 (669)\ttotal: 3.95s\tremaining: 1.95s\n",
      "670:\tlearn: 48993.6412846\ttest: 60567.3659437\tbest: 60567.3659437 (670)\ttotal: 3.95s\tremaining: 1.94s\n",
      "671:\tlearn: 48989.2030470\ttest: 60567.7009088\tbest: 60567.3659437 (670)\ttotal: 3.96s\tremaining: 1.93s\n",
      "672:\tlearn: 48973.9100027\ttest: 60564.6121462\tbest: 60564.6121462 (672)\ttotal: 3.96s\tremaining: 1.93s\n",
      "673:\tlearn: 48947.4648610\ttest: 60563.1122510\tbest: 60563.1122510 (673)\ttotal: 3.97s\tremaining: 1.92s\n",
      "674:\tlearn: 48946.3679615\ttest: 60568.4079727\tbest: 60563.1122510 (673)\ttotal: 3.98s\tremaining: 1.91s\n",
      "675:\tlearn: 48941.8052939\ttest: 60566.5834365\tbest: 60563.1122510 (673)\ttotal: 3.98s\tremaining: 1.91s\n",
      "676:\tlearn: 48939.9145365\ttest: 60563.0818887\tbest: 60563.0818887 (676)\ttotal: 3.99s\tremaining: 1.9s\n",
      "677:\tlearn: 48932.1062113\ttest: 60568.6041275\tbest: 60563.0818887 (676)\ttotal: 4s\tremaining: 1.9s\n",
      "678:\tlearn: 48924.0007761\ttest: 60568.3552559\tbest: 60563.0818887 (676)\ttotal: 4.01s\tremaining: 1.89s\n",
      "679:\tlearn: 48896.4671311\ttest: 60565.2209729\tbest: 60563.0818887 (676)\ttotal: 4.01s\tremaining: 1.89s\n",
      "680:\tlearn: 48894.8077542\ttest: 60566.2519529\tbest: 60563.0818887 (676)\ttotal: 4.02s\tremaining: 1.88s\n",
      "681:\tlearn: 48891.1568797\ttest: 60570.4856320\tbest: 60563.0818887 (676)\ttotal: 4.03s\tremaining: 1.88s\n",
      "682:\tlearn: 48866.8056706\ttest: 60562.5547663\tbest: 60562.5547663 (682)\ttotal: 4.03s\tremaining: 1.87s\n",
      "683:\tlearn: 48862.4314954\ttest: 60567.6294143\tbest: 60562.5547663 (682)\ttotal: 4.04s\tremaining: 1.86s\n",
      "684:\tlearn: 48832.0649385\ttest: 60545.3615062\tbest: 60545.3615062 (684)\ttotal: 4.04s\tremaining: 1.86s\n",
      "685:\tlearn: 48822.0250663\ttest: 60546.6734567\tbest: 60545.3615062 (684)\ttotal: 4.05s\tremaining: 1.85s\n",
      "686:\tlearn: 48793.5700777\ttest: 60546.0191031\tbest: 60545.3615062 (684)\ttotal: 4.06s\tremaining: 1.85s\n",
      "687:\tlearn: 48789.9630619\ttest: 60545.7256416\tbest: 60545.3615062 (684)\ttotal: 4.06s\tremaining: 1.84s\n",
      "688:\tlearn: 48761.7181856\ttest: 60537.4486816\tbest: 60537.4486816 (688)\ttotal: 4.07s\tremaining: 1.84s\n",
      "689:\tlearn: 48745.1311412\ttest: 60544.0721090\tbest: 60537.4486816 (688)\ttotal: 4.08s\tremaining: 1.83s\n",
      "690:\tlearn: 48732.5329618\ttest: 60539.1506883\tbest: 60537.4486816 (688)\ttotal: 4.08s\tremaining: 1.82s\n",
      "691:\tlearn: 48709.8300204\ttest: 60535.6924260\tbest: 60535.6924260 (691)\ttotal: 4.09s\tremaining: 1.82s\n",
      "692:\tlearn: 48707.0679002\ttest: 60532.2486008\tbest: 60532.2486008 (692)\ttotal: 4.09s\tremaining: 1.81s\n",
      "693:\tlearn: 48705.8728398\ttest: 60532.6141589\tbest: 60532.2486008 (692)\ttotal: 4.1s\tremaining: 1.81s\n",
      "694:\tlearn: 48704.5252068\ttest: 60534.0644190\tbest: 60532.2486008 (692)\ttotal: 4.11s\tremaining: 1.8s\n",
      "695:\tlearn: 48694.0089940\ttest: 60531.9546083\tbest: 60531.9546083 (695)\ttotal: 4.11s\tremaining: 1.8s\n",
      "696:\tlearn: 48691.8835673\ttest: 60530.0655304\tbest: 60530.0655304 (696)\ttotal: 4.12s\tremaining: 1.79s\n",
      "697:\tlearn: 48680.5495041\ttest: 60527.5489551\tbest: 60527.5489551 (697)\ttotal: 4.12s\tremaining: 1.78s\n",
      "698:\tlearn: 48592.4078640\ttest: 60512.9455735\tbest: 60512.9455735 (698)\ttotal: 4.13s\tremaining: 1.78s\n",
      "699:\tlearn: 48579.8393124\ttest: 60514.7913546\tbest: 60512.9455735 (698)\ttotal: 4.13s\tremaining: 1.77s\n",
      "700:\tlearn: 48552.9237435\ttest: 60516.2461600\tbest: 60512.9455735 (698)\ttotal: 4.14s\tremaining: 1.77s\n",
      "701:\tlearn: 48529.8000003\ttest: 60518.5463487\tbest: 60512.9455735 (698)\ttotal: 4.15s\tremaining: 1.76s\n",
      "702:\tlearn: 48528.8492108\ttest: 60521.0136894\tbest: 60512.9455735 (698)\ttotal: 4.15s\tremaining: 1.75s\n",
      "703:\tlearn: 48503.2918576\ttest: 60526.7907211\tbest: 60512.9455735 (698)\ttotal: 4.16s\tremaining: 1.75s\n",
      "704:\tlearn: 48494.6584533\ttest: 60527.5921588\tbest: 60512.9455735 (698)\ttotal: 4.17s\tremaining: 1.74s\n",
      "705:\tlearn: 48486.0450788\ttest: 60516.4609525\tbest: 60512.9455735 (698)\ttotal: 4.17s\tremaining: 1.74s\n",
      "706:\tlearn: 48480.9124323\ttest: 60511.8261868\tbest: 60511.8261868 (706)\ttotal: 4.18s\tremaining: 1.73s\n",
      "707:\tlearn: 48477.1689178\ttest: 60510.6351785\tbest: 60510.6351785 (707)\ttotal: 4.18s\tremaining: 1.72s\n",
      "708:\tlearn: 48443.7186035\ttest: 60515.5809820\tbest: 60510.6351785 (707)\ttotal: 4.19s\tremaining: 1.72s\n",
      "709:\tlearn: 48405.0178615\ttest: 60508.8999191\tbest: 60508.8999191 (709)\ttotal: 4.19s\tremaining: 1.71s\n",
      "710:\tlearn: 48381.2082474\ttest: 60502.2577072\tbest: 60502.2577072 (710)\ttotal: 4.2s\tremaining: 1.71s\n",
      "711:\tlearn: 48366.0726543\ttest: 60504.0722632\tbest: 60502.2577072 (710)\ttotal: 4.2s\tremaining: 1.7s\n",
      "712:\tlearn: 48340.0989150\ttest: 60507.3917510\tbest: 60502.2577072 (710)\ttotal: 4.21s\tremaining: 1.69s\n",
      "713:\tlearn: 48303.6369671\ttest: 60516.7760130\tbest: 60502.2577072 (710)\ttotal: 4.21s\tremaining: 1.69s\n",
      "714:\tlearn: 48301.9587388\ttest: 60516.0689456\tbest: 60502.2577072 (710)\ttotal: 4.22s\tremaining: 1.68s\n",
      "715:\tlearn: 48300.4117955\ttest: 60515.4461082\tbest: 60502.2577072 (710)\ttotal: 4.23s\tremaining: 1.68s\n",
      "716:\tlearn: 48280.6934540\ttest: 60507.9238765\tbest: 60502.2577072 (710)\ttotal: 4.23s\tremaining: 1.67s\n",
      "717:\tlearn: 48208.9651371\ttest: 60492.7890896\tbest: 60492.7890896 (717)\ttotal: 4.24s\tremaining: 1.66s\n",
      "718:\tlearn: 48183.8600726\ttest: 60489.5178957\tbest: 60489.5178957 (718)\ttotal: 4.25s\tremaining: 1.66s\n",
      "719:\tlearn: 48142.2890224\ttest: 60480.9526301\tbest: 60480.9526301 (719)\ttotal: 4.25s\tremaining: 1.65s\n",
      "720:\tlearn: 48122.3784851\ttest: 60485.2710166\tbest: 60480.9526301 (719)\ttotal: 4.26s\tremaining: 1.65s\n",
      "721:\tlearn: 48105.3579313\ttest: 60495.0156381\tbest: 60480.9526301 (719)\ttotal: 4.26s\tremaining: 1.64s\n",
      "722:\tlearn: 48096.2694579\ttest: 60497.2503992\tbest: 60480.9526301 (719)\ttotal: 4.27s\tremaining: 1.64s\n",
      "723:\tlearn: 48067.9363126\ttest: 60487.5453830\tbest: 60480.9526301 (719)\ttotal: 4.28s\tremaining: 1.63s\n",
      "724:\tlearn: 48061.2087050\ttest: 60482.8873139\tbest: 60480.9526301 (719)\ttotal: 4.28s\tremaining: 1.62s\n",
      "725:\tlearn: 48055.4990090\ttest: 60486.5097180\tbest: 60480.9526301 (719)\ttotal: 4.29s\tremaining: 1.62s\n",
      "726:\tlearn: 48051.6852557\ttest: 60485.9122062\tbest: 60480.9526301 (719)\ttotal: 4.29s\tremaining: 1.61s\n",
      "727:\tlearn: 48035.7108836\ttest: 60485.8459725\tbest: 60480.9526301 (719)\ttotal: 4.3s\tremaining: 1.61s\n",
      "728:\tlearn: 48034.9703452\ttest: 60484.0185676\tbest: 60480.9526301 (719)\ttotal: 4.3s\tremaining: 1.6s\n",
      "729:\tlearn: 48028.9583686\ttest: 60482.1524323\tbest: 60480.9526301 (719)\ttotal: 4.31s\tremaining: 1.59s\n",
      "730:\tlearn: 48021.5003920\ttest: 60469.8640486\tbest: 60469.8640486 (730)\ttotal: 4.32s\tremaining: 1.59s\n",
      "731:\tlearn: 47997.7608479\ttest: 60454.5138326\tbest: 60454.5138326 (731)\ttotal: 4.32s\tremaining: 1.58s\n",
      "732:\tlearn: 47960.0524401\ttest: 60446.1617618\tbest: 60446.1617618 (732)\ttotal: 4.33s\tremaining: 1.58s\n",
      "733:\tlearn: 47957.7482025\ttest: 60442.9363712\tbest: 60442.9363712 (733)\ttotal: 4.33s\tremaining: 1.57s\n",
      "734:\tlearn: 47955.7508802\ttest: 60442.2463105\tbest: 60442.2463105 (734)\ttotal: 4.34s\tremaining: 1.56s\n",
      "735:\tlearn: 47937.2854691\ttest: 60451.4551856\tbest: 60442.2463105 (734)\ttotal: 4.35s\tremaining: 1.56s\n",
      "736:\tlearn: 47930.4397200\ttest: 60446.5607959\tbest: 60442.2463105 (734)\ttotal: 4.35s\tremaining: 1.55s\n",
      "737:\tlearn: 47923.6966693\ttest: 60452.1149055\tbest: 60442.2463105 (734)\ttotal: 4.36s\tremaining: 1.55s\n",
      "738:\tlearn: 47923.6789149\ttest: 60453.4707093\tbest: 60442.2463105 (734)\ttotal: 4.36s\tremaining: 1.54s\n",
      "739:\tlearn: 47919.6939667\ttest: 60450.9732541\tbest: 60442.2463105 (734)\ttotal: 4.37s\tremaining: 1.53s\n",
      "740:\tlearn: 47885.8062780\ttest: 60430.3509640\tbest: 60430.3509640 (740)\ttotal: 4.37s\tremaining: 1.53s\n",
      "741:\tlearn: 47865.8029670\ttest: 60423.7999905\tbest: 60423.7999905 (741)\ttotal: 4.38s\tremaining: 1.52s\n",
      "742:\tlearn: 47842.0838096\ttest: 60426.1846792\tbest: 60423.7999905 (741)\ttotal: 4.38s\tremaining: 1.52s\n",
      "743:\tlearn: 47802.7334489\ttest: 60398.9249885\tbest: 60398.9249885 (743)\ttotal: 4.39s\tremaining: 1.51s\n",
      "744:\tlearn: 47770.8980762\ttest: 60401.3808632\tbest: 60398.9249885 (743)\ttotal: 4.39s\tremaining: 1.5s\n",
      "745:\tlearn: 47767.8835362\ttest: 60403.2331250\tbest: 60398.9249885 (743)\ttotal: 4.4s\tremaining: 1.5s\n",
      "746:\tlearn: 47746.0080733\ttest: 60423.7315032\tbest: 60398.9249885 (743)\ttotal: 4.41s\tremaining: 1.49s\n",
      "747:\tlearn: 47743.6123082\ttest: 60422.5230001\tbest: 60398.9249885 (743)\ttotal: 4.41s\tremaining: 1.49s\n",
      "748:\tlearn: 47709.1213051\ttest: 60422.2627809\tbest: 60398.9249885 (743)\ttotal: 4.42s\tremaining: 1.48s\n",
      "749:\tlearn: 47694.2865426\ttest: 60419.7712016\tbest: 60398.9249885 (743)\ttotal: 4.42s\tremaining: 1.47s\n",
      "750:\tlearn: 47679.3484748\ttest: 60421.8477226\tbest: 60398.9249885 (743)\ttotal: 4.43s\tremaining: 1.47s\n",
      "751:\tlearn: 47650.4677499\ttest: 60427.1862884\tbest: 60398.9249885 (743)\ttotal: 4.43s\tremaining: 1.46s\n",
      "752:\tlearn: 47635.4896596\ttest: 60432.3476217\tbest: 60398.9249885 (743)\ttotal: 4.44s\tremaining: 1.46s\n",
      "753:\tlearn: 47615.7470754\ttest: 60439.6039846\tbest: 60398.9249885 (743)\ttotal: 4.45s\tremaining: 1.45s\n",
      "754:\tlearn: 47615.4938945\ttest: 60439.2134544\tbest: 60398.9249885 (743)\ttotal: 4.45s\tremaining: 1.44s\n",
      "755:\tlearn: 47611.4241251\ttest: 60434.0151371\tbest: 60398.9249885 (743)\ttotal: 4.46s\tremaining: 1.44s\n",
      "756:\tlearn: 47602.5695373\ttest: 60430.7976042\tbest: 60398.9249885 (743)\ttotal: 4.46s\tremaining: 1.43s\n",
      "757:\tlearn: 47593.2035096\ttest: 60441.9287420\tbest: 60398.9249885 (743)\ttotal: 4.47s\tremaining: 1.43s\n",
      "758:\tlearn: 47580.7352422\ttest: 60429.4575593\tbest: 60398.9249885 (743)\ttotal: 4.48s\tremaining: 1.42s\n",
      "759:\tlearn: 47570.6632768\ttest: 60424.8909522\tbest: 60398.9249885 (743)\ttotal: 4.48s\tremaining: 1.42s\n",
      "760:\tlearn: 47545.9708901\ttest: 60424.6704656\tbest: 60398.9249885 (743)\ttotal: 4.49s\tremaining: 1.41s\n",
      "761:\tlearn: 47543.6148284\ttest: 60426.0210750\tbest: 60398.9249885 (743)\ttotal: 4.5s\tremaining: 1.4s\n",
      "762:\tlearn: 47532.4735702\ttest: 60423.0615106\tbest: 60398.9249885 (743)\ttotal: 4.5s\tremaining: 1.4s\n",
      "763:\tlearn: 47531.7473709\ttest: 60423.8088056\tbest: 60398.9249885 (743)\ttotal: 4.51s\tremaining: 1.39s\n",
      "764:\tlearn: 47527.9800426\ttest: 60424.6978845\tbest: 60398.9249885 (743)\ttotal: 4.51s\tremaining: 1.39s\n",
      "765:\tlearn: 47509.7747205\ttest: 60424.6755420\tbest: 60398.9249885 (743)\ttotal: 4.52s\tremaining: 1.38s\n",
      "766:\tlearn: 47490.0689513\ttest: 60405.2642302\tbest: 60398.9249885 (743)\ttotal: 4.52s\tremaining: 1.37s\n",
      "767:\tlearn: 47485.2264064\ttest: 60410.5889270\tbest: 60398.9249885 (743)\ttotal: 4.53s\tremaining: 1.37s\n",
      "768:\tlearn: 47463.0824377\ttest: 60404.6437686\tbest: 60398.9249885 (743)\ttotal: 4.53s\tremaining: 1.36s\n",
      "769:\tlearn: 47427.6177446\ttest: 60403.3968456\tbest: 60398.9249885 (743)\ttotal: 4.54s\tremaining: 1.36s\n",
      "770:\tlearn: 47425.6374501\ttest: 60403.3957129\tbest: 60398.9249885 (743)\ttotal: 4.54s\tremaining: 1.35s\n",
      "771:\tlearn: 47424.7893705\ttest: 60403.0982720\tbest: 60398.9249885 (743)\ttotal: 4.55s\tremaining: 1.34s\n",
      "772:\tlearn: 47389.5353480\ttest: 60384.9825861\tbest: 60384.9825861 (772)\ttotal: 4.56s\tremaining: 1.34s\n",
      "773:\tlearn: 47381.3247782\ttest: 60389.6773505\tbest: 60384.9825861 (772)\ttotal: 4.56s\tremaining: 1.33s\n",
      "774:\tlearn: 47378.9138596\ttest: 60395.7681974\tbest: 60384.9825861 (772)\ttotal: 4.57s\tremaining: 1.33s\n",
      "775:\tlearn: 47313.5202953\ttest: 60400.3312537\tbest: 60384.9825861 (772)\ttotal: 4.58s\tremaining: 1.32s\n",
      "776:\tlearn: 47312.5631192\ttest: 60396.9258812\tbest: 60384.9825861 (772)\ttotal: 4.58s\tremaining: 1.31s\n",
      "777:\tlearn: 47308.2938440\ttest: 60397.9293572\tbest: 60384.9825861 (772)\ttotal: 4.59s\tremaining: 1.31s\n",
      "778:\tlearn: 47266.8929182\ttest: 60377.4900557\tbest: 60377.4900557 (778)\ttotal: 4.59s\tremaining: 1.3s\n",
      "779:\tlearn: 47249.4243032\ttest: 60405.6234962\tbest: 60377.4900557 (778)\ttotal: 4.6s\tremaining: 1.3s\n",
      "780:\tlearn: 47246.3139409\ttest: 60405.2466513\tbest: 60377.4900557 (778)\ttotal: 4.61s\tremaining: 1.29s\n",
      "781:\tlearn: 47219.2227440\ttest: 60413.6158549\tbest: 60377.4900557 (778)\ttotal: 4.61s\tremaining: 1.28s\n",
      "782:\tlearn: 47211.5360533\ttest: 60405.6601024\tbest: 60377.4900557 (778)\ttotal: 4.62s\tremaining: 1.28s\n",
      "783:\tlearn: 47203.7781095\ttest: 60411.8827554\tbest: 60377.4900557 (778)\ttotal: 4.62s\tremaining: 1.27s\n",
      "784:\tlearn: 47196.7696307\ttest: 60400.7515208\tbest: 60377.4900557 (778)\ttotal: 4.63s\tremaining: 1.27s\n",
      "785:\tlearn: 47150.2465974\ttest: 60389.4053687\tbest: 60377.4900557 (778)\ttotal: 4.63s\tremaining: 1.26s\n",
      "786:\tlearn: 47148.7900216\ttest: 60388.0109157\tbest: 60377.4900557 (778)\ttotal: 4.64s\tremaining: 1.26s\n",
      "787:\tlearn: 47128.5167224\ttest: 60393.2072104\tbest: 60377.4900557 (778)\ttotal: 4.65s\tremaining: 1.25s\n",
      "788:\tlearn: 47105.8332018\ttest: 60406.7602886\tbest: 60377.4900557 (778)\ttotal: 4.65s\tremaining: 1.24s\n",
      "789:\tlearn: 47096.5857852\ttest: 60412.9782333\tbest: 60377.4900557 (778)\ttotal: 4.66s\tremaining: 1.24s\n",
      "790:\tlearn: 47067.1581435\ttest: 60404.5754937\tbest: 60377.4900557 (778)\ttotal: 4.66s\tremaining: 1.23s\n",
      "791:\tlearn: 47052.9672015\ttest: 60405.8429574\tbest: 60377.4900557 (778)\ttotal: 4.67s\tremaining: 1.23s\n",
      "792:\tlearn: 47041.1796746\ttest: 60406.3649602\tbest: 60377.4900557 (778)\ttotal: 4.67s\tremaining: 1.22s\n",
      "793:\tlearn: 47011.8759006\ttest: 60391.4760518\tbest: 60377.4900557 (778)\ttotal: 4.68s\tremaining: 1.21s\n",
      "794:\tlearn: 47007.8953264\ttest: 60392.4339269\tbest: 60377.4900557 (778)\ttotal: 4.69s\tremaining: 1.21s\n",
      "795:\tlearn: 46999.8031265\ttest: 60379.7800607\tbest: 60377.4900557 (778)\ttotal: 4.69s\tremaining: 1.2s\n",
      "796:\tlearn: 46968.1260308\ttest: 60396.3273496\tbest: 60377.4900557 (778)\ttotal: 4.7s\tremaining: 1.2s\n",
      "797:\tlearn: 46949.5138841\ttest: 60390.0093595\tbest: 60377.4900557 (778)\ttotal: 4.71s\tremaining: 1.19s\n",
      "798:\tlearn: 46896.0595517\ttest: 60382.3368027\tbest: 60377.4900557 (778)\ttotal: 4.71s\tremaining: 1.18s\n",
      "799:\tlearn: 46869.8852902\ttest: 60366.0752534\tbest: 60366.0752534 (799)\ttotal: 4.72s\tremaining: 1.18s\n",
      "800:\tlearn: 46866.9639379\ttest: 60369.4759825\tbest: 60366.0752534 (799)\ttotal: 4.72s\tremaining: 1.17s\n",
      "801:\tlearn: 46857.1786690\ttest: 60362.6742429\tbest: 60362.6742429 (801)\ttotal: 4.73s\tremaining: 1.17s\n",
      "802:\tlearn: 46847.0572843\ttest: 60359.8537069\tbest: 60359.8537069 (802)\ttotal: 4.74s\tremaining: 1.16s\n",
      "803:\tlearn: 46839.0076682\ttest: 60358.5524233\tbest: 60358.5524233 (803)\ttotal: 4.74s\tremaining: 1.16s\n",
      "804:\tlearn: 46816.5632207\ttest: 60355.9061505\tbest: 60355.9061505 (804)\ttotal: 4.75s\tremaining: 1.15s\n",
      "805:\tlearn: 46811.6884676\ttest: 60356.7524840\tbest: 60355.9061505 (804)\ttotal: 4.76s\tremaining: 1.14s\n",
      "806:\tlearn: 46789.6875402\ttest: 60363.1686047\tbest: 60355.9061505 (804)\ttotal: 4.76s\tremaining: 1.14s\n",
      "807:\tlearn: 46750.7238909\ttest: 60355.0110658\tbest: 60355.0110658 (807)\ttotal: 4.77s\tremaining: 1.13s\n",
      "808:\tlearn: 46736.9442917\ttest: 60355.9161572\tbest: 60355.0110658 (807)\ttotal: 4.77s\tremaining: 1.13s\n",
      "809:\tlearn: 46735.6768101\ttest: 60354.9278413\tbest: 60354.9278413 (809)\ttotal: 4.78s\tremaining: 1.12s\n",
      "810:\tlearn: 46705.8260201\ttest: 60329.8911411\tbest: 60329.8911411 (810)\ttotal: 4.78s\tremaining: 1.11s\n",
      "811:\tlearn: 46684.6563639\ttest: 60309.6225582\tbest: 60309.6225582 (811)\ttotal: 4.79s\tremaining: 1.11s\n",
      "812:\tlearn: 46655.4979107\ttest: 60306.6672537\tbest: 60306.6672537 (812)\ttotal: 4.8s\tremaining: 1.1s\n",
      "813:\tlearn: 46655.0302227\ttest: 60307.7894499\tbest: 60306.6672537 (812)\ttotal: 4.8s\tremaining: 1.1s\n",
      "814:\tlearn: 46635.2463103\ttest: 60304.4364598\tbest: 60304.4364598 (814)\ttotal: 4.81s\tremaining: 1.09s\n",
      "815:\tlearn: 46601.1362045\ttest: 60304.8118712\tbest: 60304.4364598 (814)\ttotal: 4.82s\tremaining: 1.08s\n",
      "816:\tlearn: 46600.5266883\ttest: 60303.3175461\tbest: 60303.3175461 (816)\ttotal: 4.82s\tremaining: 1.08s\n",
      "817:\tlearn: 46595.9074850\ttest: 60302.0597810\tbest: 60302.0597810 (817)\ttotal: 4.83s\tremaining: 1.07s\n",
      "818:\tlearn: 46586.0793907\ttest: 60298.6135813\tbest: 60298.6135813 (818)\ttotal: 4.83s\tremaining: 1.07s\n",
      "819:\tlearn: 46557.1007567\ttest: 60288.3335555\tbest: 60288.3335555 (819)\ttotal: 4.84s\tremaining: 1.06s\n",
      "820:\tlearn: 46536.1502700\ttest: 60286.1781504\tbest: 60286.1781504 (820)\ttotal: 4.84s\tremaining: 1.06s\n",
      "821:\tlearn: 46512.7714838\ttest: 60271.2014669\tbest: 60271.2014669 (821)\ttotal: 4.85s\tremaining: 1.05s\n",
      "822:\tlearn: 46510.1647953\ttest: 60272.5075012\tbest: 60271.2014669 (821)\ttotal: 4.86s\tremaining: 1.04s\n",
      "823:\tlearn: 46502.2614315\ttest: 60274.7734315\tbest: 60271.2014669 (821)\ttotal: 4.86s\tremaining: 1.04s\n",
      "824:\tlearn: 46482.1842559\ttest: 60282.2591137\tbest: 60271.2014669 (821)\ttotal: 4.87s\tremaining: 1.03s\n",
      "825:\tlearn: 46462.0224368\ttest: 60269.6082529\tbest: 60269.6082529 (825)\ttotal: 4.87s\tremaining: 1.03s\n",
      "826:\tlearn: 46454.6518180\ttest: 60268.0679661\tbest: 60268.0679661 (826)\ttotal: 4.88s\tremaining: 1.02s\n",
      "827:\tlearn: 46431.7879064\ttest: 60268.7799552\tbest: 60268.0679661 (826)\ttotal: 4.88s\tremaining: 1.01s\n",
      "828:\tlearn: 46416.3150083\ttest: 60252.4068266\tbest: 60252.4068266 (828)\ttotal: 4.89s\tremaining: 1.01s\n",
      "829:\tlearn: 46399.8892101\ttest: 60248.6077717\tbest: 60248.6077717 (829)\ttotal: 4.89s\tremaining: 1s\n",
      "830:\tlearn: 46378.4867383\ttest: 60249.4678581\tbest: 60248.6077717 (829)\ttotal: 4.9s\tremaining: 997ms\n",
      "831:\tlearn: 46327.6234166\ttest: 60247.6560742\tbest: 60247.6560742 (831)\ttotal: 4.91s\tremaining: 991ms\n",
      "832:\tlearn: 46314.2286947\ttest: 60239.0436042\tbest: 60239.0436042 (832)\ttotal: 4.91s\tremaining: 985ms\n",
      "833:\tlearn: 46296.2732268\ttest: 60243.2681775\tbest: 60239.0436042 (832)\ttotal: 4.92s\tremaining: 979ms\n",
      "834:\tlearn: 46290.5525647\ttest: 60238.8332793\tbest: 60238.8332793 (834)\ttotal: 4.92s\tremaining: 973ms\n",
      "835:\tlearn: 46245.5919318\ttest: 60205.8721846\tbest: 60205.8721846 (835)\ttotal: 4.93s\tremaining: 967ms\n",
      "836:\tlearn: 46233.9019300\ttest: 60198.0836314\tbest: 60198.0836314 (836)\ttotal: 4.93s\tremaining: 961ms\n",
      "837:\tlearn: 46229.8519998\ttest: 60198.5763508\tbest: 60198.0836314 (836)\ttotal: 4.94s\tremaining: 955ms\n",
      "838:\tlearn: 46201.4258655\ttest: 60206.7012334\tbest: 60198.0836314 (836)\ttotal: 4.95s\tremaining: 949ms\n",
      "839:\tlearn: 46181.3815792\ttest: 60210.7739202\tbest: 60198.0836314 (836)\ttotal: 4.96s\tremaining: 944ms\n",
      "840:\tlearn: 46173.9245834\ttest: 60206.3996127\tbest: 60198.0836314 (836)\ttotal: 4.97s\tremaining: 940ms\n",
      "841:\tlearn: 46125.7376202\ttest: 60196.8718202\tbest: 60196.8718202 (841)\ttotal: 4.98s\tremaining: 934ms\n",
      "842:\tlearn: 46102.7040722\ttest: 60183.5138406\tbest: 60183.5138406 (842)\ttotal: 4.98s\tremaining: 928ms\n",
      "843:\tlearn: 46083.8268730\ttest: 60190.0225291\tbest: 60183.5138406 (842)\ttotal: 4.99s\tremaining: 922ms\n",
      "844:\tlearn: 46070.0590327\ttest: 60194.6541269\tbest: 60183.5138406 (842)\ttotal: 5s\tremaining: 916ms\n",
      "845:\tlearn: 46037.8794581\ttest: 60187.3219667\tbest: 60183.5138406 (842)\ttotal: 5s\tremaining: 911ms\n",
      "846:\tlearn: 46003.0098030\ttest: 60176.5485882\tbest: 60176.5485882 (846)\ttotal: 5.01s\tremaining: 905ms\n",
      "847:\tlearn: 46001.2340784\ttest: 60175.4945893\tbest: 60175.4945893 (847)\ttotal: 5.02s\tremaining: 899ms\n",
      "848:\tlearn: 45996.5458767\ttest: 60183.6931362\tbest: 60175.4945893 (847)\ttotal: 5.03s\tremaining: 894ms\n",
      "849:\tlearn: 45995.9501617\ttest: 60183.3516491\tbest: 60175.4945893 (847)\ttotal: 5.03s\tremaining: 888ms\n",
      "850:\tlearn: 45970.7676679\ttest: 60173.3520312\tbest: 60173.3520312 (850)\ttotal: 5.04s\tremaining: 882ms\n",
      "851:\tlearn: 45959.7681918\ttest: 60174.6572951\tbest: 60173.3520312 (850)\ttotal: 5.04s\tremaining: 876ms\n",
      "852:\tlearn: 45917.2486943\ttest: 60170.6426825\tbest: 60170.6426825 (852)\ttotal: 5.05s\tremaining: 870ms\n",
      "853:\tlearn: 45909.1954195\ttest: 60163.5361965\tbest: 60163.5361965 (853)\ttotal: 5.06s\tremaining: 864ms\n",
      "854:\tlearn: 45889.9189716\ttest: 60162.9776311\tbest: 60162.9776311 (854)\ttotal: 5.06s\tremaining: 859ms\n",
      "855:\tlearn: 45878.0088905\ttest: 60166.4503428\tbest: 60162.9776311 (854)\ttotal: 5.07s\tremaining: 853ms\n",
      "856:\tlearn: 45858.2856685\ttest: 60158.3201686\tbest: 60158.3201686 (856)\ttotal: 5.07s\tremaining: 847ms\n",
      "857:\tlearn: 45842.5669566\ttest: 60158.0213968\tbest: 60158.0213968 (857)\ttotal: 5.08s\tremaining: 841ms\n",
      "858:\tlearn: 45821.9175981\ttest: 60150.0886405\tbest: 60150.0886405 (858)\ttotal: 5.08s\tremaining: 835ms\n",
      "859:\tlearn: 45802.4158948\ttest: 60151.9304002\tbest: 60150.0886405 (858)\ttotal: 5.09s\tremaining: 829ms\n",
      "860:\tlearn: 45786.7113897\ttest: 60172.9851111\tbest: 60150.0886405 (858)\ttotal: 5.09s\tremaining: 823ms\n",
      "861:\tlearn: 45780.1853143\ttest: 60171.6991222\tbest: 60150.0886405 (858)\ttotal: 5.1s\tremaining: 817ms\n",
      "862:\tlearn: 45767.3058912\ttest: 60167.5969524\tbest: 60150.0886405 (858)\ttotal: 5.11s\tremaining: 811ms\n",
      "863:\tlearn: 45760.1347055\ttest: 60174.5859057\tbest: 60150.0886405 (858)\ttotal: 5.11s\tremaining: 805ms\n",
      "864:\tlearn: 45744.4658153\ttest: 60164.0828783\tbest: 60150.0886405 (858)\ttotal: 5.12s\tremaining: 799ms\n",
      "865:\tlearn: 45731.0553756\ttest: 60165.0882741\tbest: 60150.0886405 (858)\ttotal: 5.13s\tremaining: 793ms\n",
      "866:\tlearn: 45705.1778707\ttest: 60167.1555776\tbest: 60150.0886405 (858)\ttotal: 5.13s\tremaining: 787ms\n",
      "867:\tlearn: 45634.4773943\ttest: 60147.1542261\tbest: 60147.1542261 (867)\ttotal: 5.14s\tremaining: 781ms\n",
      "868:\tlearn: 45617.4966580\ttest: 60143.0140059\tbest: 60143.0140059 (868)\ttotal: 5.14s\tremaining: 775ms\n",
      "869:\tlearn: 45588.7391226\ttest: 60146.3207586\tbest: 60143.0140059 (868)\ttotal: 5.15s\tremaining: 769ms\n",
      "870:\tlearn: 45578.0040623\ttest: 60151.4274129\tbest: 60143.0140059 (868)\ttotal: 5.16s\tremaining: 764ms\n",
      "871:\tlearn: 45550.0222008\ttest: 60166.3130127\tbest: 60143.0140059 (868)\ttotal: 5.16s\tremaining: 757ms\n",
      "872:\tlearn: 45531.9263277\ttest: 60178.0323918\tbest: 60143.0140059 (868)\ttotal: 5.17s\tremaining: 752ms\n",
      "873:\tlearn: 45527.0147686\ttest: 60177.2418540\tbest: 60143.0140059 (868)\ttotal: 5.17s\tremaining: 746ms\n",
      "874:\tlearn: 45523.2024333\ttest: 60179.4391473\tbest: 60143.0140059 (868)\ttotal: 5.18s\tremaining: 740ms\n",
      "875:\tlearn: 45489.7696243\ttest: 60178.3956601\tbest: 60143.0140059 (868)\ttotal: 5.18s\tremaining: 734ms\n",
      "876:\tlearn: 45482.6155390\ttest: 60180.3675694\tbest: 60143.0140059 (868)\ttotal: 5.19s\tremaining: 728ms\n",
      "877:\tlearn: 45476.9250533\ttest: 60181.1011448\tbest: 60143.0140059 (868)\ttotal: 5.2s\tremaining: 722ms\n",
      "878:\tlearn: 45444.3905968\ttest: 60187.6395846\tbest: 60143.0140059 (868)\ttotal: 5.2s\tremaining: 716ms\n",
      "879:\tlearn: 45428.5159385\ttest: 60214.6718373\tbest: 60143.0140059 (868)\ttotal: 5.21s\tremaining: 710ms\n",
      "880:\tlearn: 45412.7277968\ttest: 60218.9568366\tbest: 60143.0140059 (868)\ttotal: 5.21s\tremaining: 704ms\n",
      "881:\tlearn: 45386.7222203\ttest: 60216.4108596\tbest: 60143.0140059 (868)\ttotal: 5.22s\tremaining: 698ms\n",
      "882:\tlearn: 45379.9320276\ttest: 60231.0699164\tbest: 60143.0140059 (868)\ttotal: 5.22s\tremaining: 692ms\n",
      "883:\tlearn: 45378.4546379\ttest: 60232.6721994\tbest: 60143.0140059 (868)\ttotal: 5.23s\tremaining: 686ms\n",
      "884:\tlearn: 45363.4458142\ttest: 60237.9379513\tbest: 60143.0140059 (868)\ttotal: 5.24s\tremaining: 680ms\n",
      "885:\tlearn: 45359.4976823\ttest: 60232.8047913\tbest: 60143.0140059 (868)\ttotal: 5.24s\tremaining: 674ms\n",
      "886:\tlearn: 45347.4323191\ttest: 60219.5964447\tbest: 60143.0140059 (868)\ttotal: 5.25s\tremaining: 669ms\n",
      "887:\tlearn: 45345.9357761\ttest: 60221.4529008\tbest: 60143.0140059 (868)\ttotal: 5.25s\tremaining: 663ms\n",
      "888:\tlearn: 45342.8328678\ttest: 60220.5367629\tbest: 60143.0140059 (868)\ttotal: 5.26s\tremaining: 657ms\n",
      "889:\tlearn: 45339.9950832\ttest: 60219.0847369\tbest: 60143.0140059 (868)\ttotal: 5.26s\tremaining: 651ms\n",
      "890:\tlearn: 45318.3429046\ttest: 60226.8359127\tbest: 60143.0140059 (868)\ttotal: 5.27s\tremaining: 645ms\n",
      "891:\tlearn: 45311.4828113\ttest: 60222.1064115\tbest: 60143.0140059 (868)\ttotal: 5.28s\tremaining: 639ms\n",
      "892:\tlearn: 45297.9799497\ttest: 60223.7403833\tbest: 60143.0140059 (868)\ttotal: 5.28s\tremaining: 633ms\n",
      "893:\tlearn: 45291.0737870\ttest: 60224.6065235\tbest: 60143.0140059 (868)\ttotal: 5.29s\tremaining: 627ms\n",
      "894:\tlearn: 45270.8478215\ttest: 60223.5044498\tbest: 60143.0140059 (868)\ttotal: 5.29s\tremaining: 621ms\n",
      "895:\tlearn: 45266.0956818\ttest: 60226.6250018\tbest: 60143.0140059 (868)\ttotal: 5.3s\tremaining: 615ms\n",
      "896:\tlearn: 45258.1397590\ttest: 60232.7555946\tbest: 60143.0140059 (868)\ttotal: 5.3s\tremaining: 609ms\n",
      "897:\tlearn: 45222.8668572\ttest: 60227.3937028\tbest: 60143.0140059 (868)\ttotal: 5.31s\tremaining: 603ms\n",
      "898:\tlearn: 45208.8357116\ttest: 60234.0479156\tbest: 60143.0140059 (868)\ttotal: 5.32s\tremaining: 597ms\n",
      "899:\tlearn: 45201.2074864\ttest: 60234.2655214\tbest: 60143.0140059 (868)\ttotal: 5.32s\tremaining: 591ms\n",
      "900:\tlearn: 45196.2629453\ttest: 60237.2055150\tbest: 60143.0140059 (868)\ttotal: 5.33s\tremaining: 585ms\n",
      "901:\tlearn: 45180.1217452\ttest: 60231.3635047\tbest: 60143.0140059 (868)\ttotal: 5.33s\tremaining: 579ms\n",
      "902:\tlearn: 45175.4880762\ttest: 60236.0790521\tbest: 60143.0140059 (868)\ttotal: 5.34s\tremaining: 574ms\n",
      "903:\tlearn: 45154.5768151\ttest: 60223.9596562\tbest: 60143.0140059 (868)\ttotal: 5.35s\tremaining: 568ms\n",
      "904:\tlearn: 45153.6496085\ttest: 60228.7563947\tbest: 60143.0140059 (868)\ttotal: 5.35s\tremaining: 562ms\n",
      "905:\tlearn: 45152.2718128\ttest: 60229.7083788\tbest: 60143.0140059 (868)\ttotal: 5.36s\tremaining: 556ms\n",
      "906:\tlearn: 45141.4622902\ttest: 60233.3669080\tbest: 60143.0140059 (868)\ttotal: 5.36s\tremaining: 550ms\n",
      "907:\tlearn: 45115.1900423\ttest: 60225.1410786\tbest: 60143.0140059 (868)\ttotal: 5.37s\tremaining: 544ms\n",
      "908:\tlearn: 45096.0986133\ttest: 60225.6098050\tbest: 60143.0140059 (868)\ttotal: 5.37s\tremaining: 538ms\n",
      "909:\tlearn: 45090.0321399\ttest: 60230.2470007\tbest: 60143.0140059 (868)\ttotal: 5.38s\tremaining: 532ms\n",
      "910:\tlearn: 45071.8448547\ttest: 60235.7631830\tbest: 60143.0140059 (868)\ttotal: 5.38s\tremaining: 526ms\n",
      "911:\tlearn: 45071.0718265\ttest: 60239.5248880\tbest: 60143.0140059 (868)\ttotal: 5.39s\tremaining: 520ms\n",
      "912:\tlearn: 45060.0127893\ttest: 60239.3058720\tbest: 60143.0140059 (868)\ttotal: 5.4s\tremaining: 514ms\n",
      "913:\tlearn: 45051.4373398\ttest: 60236.6524521\tbest: 60143.0140059 (868)\ttotal: 5.4s\tremaining: 508ms\n",
      "914:\tlearn: 45015.0344585\ttest: 60236.0039653\tbest: 60143.0140059 (868)\ttotal: 5.41s\tremaining: 503ms\n",
      "915:\tlearn: 44996.7590817\ttest: 60216.4912787\tbest: 60143.0140059 (868)\ttotal: 5.42s\tremaining: 497ms\n",
      "916:\tlearn: 44981.8691072\ttest: 60212.8692468\tbest: 60143.0140059 (868)\ttotal: 5.42s\tremaining: 491ms\n",
      "917:\tlearn: 44949.1649296\ttest: 60192.7729748\tbest: 60143.0140059 (868)\ttotal: 5.43s\tremaining: 485ms\n",
      "918:\tlearn: 44932.6321717\ttest: 60190.6344047\tbest: 60143.0140059 (868)\ttotal: 5.43s\tremaining: 479ms\n",
      "919:\tlearn: 44921.8496841\ttest: 60192.7160798\tbest: 60143.0140059 (868)\ttotal: 5.44s\tremaining: 473ms\n",
      "920:\tlearn: 44910.4814370\ttest: 60192.5369390\tbest: 60143.0140059 (868)\ttotal: 5.44s\tremaining: 467ms\n",
      "921:\tlearn: 44872.8673099\ttest: 60177.9555751\tbest: 60143.0140059 (868)\ttotal: 5.45s\tremaining: 461ms\n",
      "922:\tlearn: 44843.0701925\ttest: 60169.4080335\tbest: 60143.0140059 (868)\ttotal: 5.45s\tremaining: 455ms\n",
      "923:\tlearn: 44836.5921594\ttest: 60157.8020510\tbest: 60143.0140059 (868)\ttotal: 5.46s\tremaining: 449ms\n",
      "924:\tlearn: 44834.9970841\ttest: 60157.3531026\tbest: 60143.0140059 (868)\ttotal: 5.46s\tremaining: 443ms\n",
      "925:\tlearn: 44825.5084662\ttest: 60156.2298776\tbest: 60143.0140059 (868)\ttotal: 5.47s\tremaining: 437ms\n",
      "926:\tlearn: 44822.6319646\ttest: 60163.5647449\tbest: 60143.0140059 (868)\ttotal: 5.48s\tremaining: 431ms\n",
      "927:\tlearn: 44816.1981494\ttest: 60163.7288770\tbest: 60143.0140059 (868)\ttotal: 5.48s\tremaining: 425ms\n",
      "928:\tlearn: 44803.9158525\ttest: 60152.4598821\tbest: 60143.0140059 (868)\ttotal: 5.49s\tremaining: 419ms\n",
      "929:\tlearn: 44794.0942566\ttest: 60146.5984453\tbest: 60143.0140059 (868)\ttotal: 5.49s\tremaining: 413ms\n",
      "930:\tlearn: 44784.0591162\ttest: 60140.2342043\tbest: 60140.2342043 (930)\ttotal: 5.5s\tremaining: 407ms\n",
      "931:\tlearn: 44763.3812737\ttest: 60128.8728210\tbest: 60128.8728210 (931)\ttotal: 5.5s\tremaining: 402ms\n",
      "932:\tlearn: 44745.9778523\ttest: 60112.8603495\tbest: 60112.8603495 (932)\ttotal: 5.51s\tremaining: 396ms\n",
      "933:\tlearn: 44743.6078397\ttest: 60109.0904747\tbest: 60109.0904747 (933)\ttotal: 5.52s\tremaining: 390ms\n",
      "934:\tlearn: 44729.0763040\ttest: 60113.2986745\tbest: 60109.0904747 (933)\ttotal: 5.52s\tremaining: 384ms\n",
      "935:\tlearn: 44718.6975281\ttest: 60122.0277618\tbest: 60109.0904747 (933)\ttotal: 5.53s\tremaining: 378ms\n",
      "936:\tlearn: 44687.4051043\ttest: 60135.7280019\tbest: 60109.0904747 (933)\ttotal: 5.54s\tremaining: 372ms\n",
      "937:\tlearn: 44671.7178310\ttest: 60130.2429715\tbest: 60109.0904747 (933)\ttotal: 5.54s\tremaining: 366ms\n",
      "938:\tlearn: 44664.7400300\ttest: 60142.4245599\tbest: 60109.0904747 (933)\ttotal: 5.55s\tremaining: 360ms\n",
      "939:\tlearn: 44661.9124151\ttest: 60134.3718689\tbest: 60109.0904747 (933)\ttotal: 5.55s\tremaining: 354ms\n",
      "940:\tlearn: 44646.8412793\ttest: 60134.2457322\tbest: 60109.0904747 (933)\ttotal: 5.56s\tremaining: 349ms\n",
      "941:\tlearn: 44645.4064681\ttest: 60132.7327169\tbest: 60109.0904747 (933)\ttotal: 5.57s\tremaining: 343ms\n",
      "942:\tlearn: 44636.6880613\ttest: 60134.7160914\tbest: 60109.0904747 (933)\ttotal: 5.57s\tremaining: 337ms\n",
      "943:\tlearn: 44621.9735860\ttest: 60136.4659549\tbest: 60109.0904747 (933)\ttotal: 5.58s\tremaining: 331ms\n",
      "944:\tlearn: 44619.7349469\ttest: 60138.6541506\tbest: 60109.0904747 (933)\ttotal: 5.58s\tremaining: 325ms\n",
      "945:\tlearn: 44595.7032948\ttest: 60139.5156102\tbest: 60109.0904747 (933)\ttotal: 5.59s\tremaining: 319ms\n",
      "946:\tlearn: 44592.7957510\ttest: 60141.6770819\tbest: 60109.0904747 (933)\ttotal: 5.59s\tremaining: 313ms\n",
      "947:\tlearn: 44580.6703507\ttest: 60141.0487787\tbest: 60109.0904747 (933)\ttotal: 5.6s\tremaining: 307ms\n",
      "948:\tlearn: 44571.3530843\ttest: 60147.1486658\tbest: 60109.0904747 (933)\ttotal: 5.61s\tremaining: 301ms\n",
      "949:\tlearn: 44557.8710069\ttest: 60155.8953512\tbest: 60109.0904747 (933)\ttotal: 5.61s\tremaining: 295ms\n",
      "950:\tlearn: 44546.7348157\ttest: 60152.2711327\tbest: 60109.0904747 (933)\ttotal: 5.62s\tremaining: 290ms\n",
      "951:\tlearn: 44539.9210749\ttest: 60147.4444650\tbest: 60109.0904747 (933)\ttotal: 5.62s\tremaining: 284ms\n",
      "952:\tlearn: 44511.9476877\ttest: 60144.0595182\tbest: 60109.0904747 (933)\ttotal: 5.63s\tremaining: 278ms\n",
      "953:\tlearn: 44501.3535709\ttest: 60145.2089096\tbest: 60109.0904747 (933)\ttotal: 5.64s\tremaining: 272ms\n",
      "954:\tlearn: 44497.8910099\ttest: 60144.7588359\tbest: 60109.0904747 (933)\ttotal: 5.64s\tremaining: 266ms\n",
      "955:\tlearn: 44492.8140461\ttest: 60143.7689929\tbest: 60109.0904747 (933)\ttotal: 5.65s\tremaining: 260ms\n",
      "956:\tlearn: 44482.0423933\ttest: 60137.8023886\tbest: 60109.0904747 (933)\ttotal: 5.65s\tremaining: 254ms\n",
      "957:\tlearn: 44429.3743971\ttest: 60136.8057388\tbest: 60109.0904747 (933)\ttotal: 5.66s\tremaining: 248ms\n",
      "958:\tlearn: 44394.2508448\ttest: 60132.8790921\tbest: 60109.0904747 (933)\ttotal: 5.67s\tremaining: 242ms\n",
      "959:\tlearn: 44383.3347304\ttest: 60133.9704529\tbest: 60109.0904747 (933)\ttotal: 5.67s\tremaining: 236ms\n",
      "960:\tlearn: 44374.9013357\ttest: 60136.2107936\tbest: 60109.0904747 (933)\ttotal: 5.68s\tremaining: 230ms\n",
      "961:\tlearn: 44367.0415801\ttest: 60130.3522962\tbest: 60109.0904747 (933)\ttotal: 5.68s\tremaining: 225ms\n",
      "962:\tlearn: 44354.8549714\ttest: 60120.4371050\tbest: 60109.0904747 (933)\ttotal: 5.69s\tremaining: 219ms\n",
      "963:\tlearn: 44340.0319990\ttest: 60109.0461203\tbest: 60109.0461203 (963)\ttotal: 5.7s\tremaining: 213ms\n",
      "964:\tlearn: 44335.3648639\ttest: 60115.0793416\tbest: 60109.0461203 (963)\ttotal: 5.7s\tremaining: 207ms\n",
      "965:\tlearn: 44332.7688248\ttest: 60117.5392347\tbest: 60109.0461203 (963)\ttotal: 5.71s\tremaining: 201ms\n",
      "966:\tlearn: 44321.3185393\ttest: 60114.3277225\tbest: 60109.0461203 (963)\ttotal: 5.71s\tremaining: 195ms\n",
      "967:\tlearn: 44309.6919706\ttest: 60102.1847882\tbest: 60102.1847882 (967)\ttotal: 5.72s\tremaining: 189ms\n",
      "968:\tlearn: 44292.0197647\ttest: 60106.1224216\tbest: 60102.1847882 (967)\ttotal: 5.72s\tremaining: 183ms\n",
      "969:\tlearn: 44268.1047133\ttest: 60096.7895991\tbest: 60096.7895991 (969)\ttotal: 5.73s\tremaining: 177ms\n",
      "970:\tlearn: 44262.4219087\ttest: 60093.3394702\tbest: 60093.3394702 (970)\ttotal: 5.74s\tremaining: 171ms\n",
      "971:\tlearn: 44256.2240877\ttest: 60093.3214342\tbest: 60093.3214342 (971)\ttotal: 5.74s\tremaining: 165ms\n",
      "972:\tlearn: 44240.6189954\ttest: 60098.0667534\tbest: 60093.3214342 (971)\ttotal: 5.75s\tremaining: 160ms\n",
      "973:\tlearn: 44217.5493483\ttest: 60101.9559666\tbest: 60093.3214342 (971)\ttotal: 5.75s\tremaining: 154ms\n",
      "974:\tlearn: 44209.3348689\ttest: 60105.5893005\tbest: 60093.3214342 (971)\ttotal: 5.76s\tremaining: 148ms\n",
      "975:\tlearn: 44205.1566913\ttest: 60106.6990072\tbest: 60093.3214342 (971)\ttotal: 5.77s\tremaining: 142ms\n",
      "976:\tlearn: 44199.6390603\ttest: 60109.5471590\tbest: 60093.3214342 (971)\ttotal: 5.78s\tremaining: 136ms\n",
      "977:\tlearn: 44188.6201643\ttest: 60123.5332428\tbest: 60093.3214342 (971)\ttotal: 5.78s\tremaining: 130ms\n",
      "978:\tlearn: 44169.7404578\ttest: 60131.4826468\tbest: 60093.3214342 (971)\ttotal: 5.79s\tremaining: 124ms\n",
      "979:\tlearn: 44163.3127234\ttest: 60135.3605839\tbest: 60093.3214342 (971)\ttotal: 5.8s\tremaining: 118ms\n",
      "980:\tlearn: 44153.5745556\ttest: 60139.0445823\tbest: 60093.3214342 (971)\ttotal: 5.8s\tremaining: 112ms\n",
      "981:\tlearn: 44150.6148673\ttest: 60137.6865559\tbest: 60093.3214342 (971)\ttotal: 5.81s\tremaining: 106ms\n",
      "982:\tlearn: 44149.0360996\ttest: 60136.3031004\tbest: 60093.3214342 (971)\ttotal: 5.81s\tremaining: 101ms\n",
      "983:\tlearn: 44144.4073287\ttest: 60140.8952587\tbest: 60093.3214342 (971)\ttotal: 5.82s\tremaining: 94.6ms\n",
      "984:\tlearn: 44131.1293800\ttest: 60146.7767164\tbest: 60093.3214342 (971)\ttotal: 5.83s\tremaining: 88.7ms\n",
      "985:\tlearn: 44117.9372100\ttest: 60146.2650790\tbest: 60093.3214342 (971)\ttotal: 5.83s\tremaining: 82.8ms\n",
      "986:\tlearn: 44099.5957803\ttest: 60130.7196586\tbest: 60093.3214342 (971)\ttotal: 5.84s\tremaining: 76.9ms\n",
      "987:\tlearn: 44094.7842995\ttest: 60140.3477445\tbest: 60093.3214342 (971)\ttotal: 5.84s\tremaining: 71ms\n",
      "988:\tlearn: 44093.8316261\ttest: 60141.9583459\tbest: 60093.3214342 (971)\ttotal: 5.85s\tremaining: 65ms\n",
      "989:\tlearn: 44093.1081877\ttest: 60141.7057412\tbest: 60093.3214342 (971)\ttotal: 5.85s\tremaining: 59.1ms\n",
      "990:\tlearn: 44077.6756191\ttest: 60121.0801271\tbest: 60093.3214342 (971)\ttotal: 5.86s\tremaining: 53.2ms\n",
      "991:\tlearn: 44044.1508893\ttest: 60109.1948553\tbest: 60093.3214342 (971)\ttotal: 5.86s\tremaining: 47.3ms\n",
      "992:\tlearn: 44030.4103494\ttest: 60110.8906009\tbest: 60093.3214342 (971)\ttotal: 5.87s\tremaining: 41.4ms\n",
      "993:\tlearn: 44028.8291446\ttest: 60109.5189721\tbest: 60093.3214342 (971)\ttotal: 5.87s\tremaining: 35.5ms\n",
      "994:\tlearn: 44005.0557370\ttest: 60115.6861610\tbest: 60093.3214342 (971)\ttotal: 5.88s\tremaining: 29.6ms\n",
      "995:\tlearn: 44004.0257203\ttest: 60116.0061436\tbest: 60093.3214342 (971)\ttotal: 5.89s\tremaining: 23.6ms\n",
      "996:\tlearn: 43997.9251474\ttest: 60114.8303964\tbest: 60093.3214342 (971)\ttotal: 5.89s\tremaining: 17.7ms\n",
      "997:\tlearn: 43995.3305896\ttest: 60112.8794851\tbest: 60093.3214342 (971)\ttotal: 5.9s\tremaining: 11.8ms\n",
      "998:\tlearn: 43993.7631259\ttest: 60113.3032544\tbest: 60093.3214342 (971)\ttotal: 5.9s\tremaining: 5.91ms\n",
      "999:\tlearn: 43983.4532610\ttest: 60118.4906217\tbest: 60093.3214342 (971)\ttotal: 5.91s\tremaining: 0us\n",
      "\n",
      "bestTest = 60093.32143\n",
      "bestIteration = 971\n",
      "\n",
      "Shrink model to first 972 iterations.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<catboost.core.CatBoostRegressor at 0x1a1ff71ba8>"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "catr.fit(train_X, train_y, eval_set=(test_X, test_y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "yhat_catr = catr.predict(test_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "43525.695649187161"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_absolute_error(test_y, yhat_catr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hola! Catboost's default setting is already giving the lowest MAE amongst other models!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for y,x in sorted(zip(catr.feature_importances_, train_X.columns), reverse=False):\n",
    "    #print (x,y)\n",
    "    plt.barh(x,y, color='b')\n",
    "    plt.title('Feature Importances - CatBoost')\n",
    "    plt.xlabel('Relative Importance')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.Results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4 different tree-based model are tested with MAE:\n",
    "* Random Forest (default): 46052\n",
    "* Xgboost (default): 45025\n",
    "* Lightgbm (tuned): 48051\n",
    "* Catboost (default): 43525\n",
    "\n",
    "As such, this is significantly better then the $70,000 MAE by industrial appraisers! *(Are property appraisers going to be replaced by machine learning!?)*\n",
    "\n",
    "Next Step: Tuning parameters of each model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For project background and documentation, please refer to README file."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}