{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {
    "collapsed": true,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import make_regression\n",
    "from sklearn.multioutput import MultiOutputRegressor\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "sns.set(color_codes=True)\n",
    "pal = sns.color_palette(\"viridis\", 10)\n",
    "sns.set_palette('muted')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "导入数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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>Time</th>\n",
       "      <th>index_A</th>\n",
       "      <th>index_B</th>\n",
       "      <th>index_C</th>\n",
       "      <th>index_D</th>\n",
       "      <th>Temperature_of_system1</th>\n",
       "      <th>Temperature_of_system2</th>\n",
       "      <th>Mineral_parameter1</th>\n",
       "      <th>Mineral_parameter2</th>\n",
       "      <th>Mineral_parameter3</th>\n",
       "      <th>Mineral_parameter4</th>\n",
       "      <th>Process_parameter1</th>\n",
       "      <th>Process_parameter2</th>\n",
       "      <th>Process_parameter3</th>\n",
       "      <th>Process_parameter4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2022-01-25 00:50:00</td>\n",
       "      <td>78.31</td>\n",
       "      <td>23.66</td>\n",
       "      <td>12.24</td>\n",
       "      <td>17.81</td>\n",
       "      <td>1343.879322</td>\n",
       "      <td>949.674915</td>\n",
       "      <td>55.26</td>\n",
       "      <td>108.03</td>\n",
       "      <td>43.29</td>\n",
       "      <td>20.92</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>226.16</td>\n",
       "      <td>181.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2022-01-25 01:50:00</td>\n",
       "      <td>78.46</td>\n",
       "      <td>23.88</td>\n",
       "      <td>12.41</td>\n",
       "      <td>17.99</td>\n",
       "      <td>1273.313000</td>\n",
       "      <td>937.436333</td>\n",
       "      <td>55.26</td>\n",
       "      <td>108.03</td>\n",
       "      <td>43.29</td>\n",
       "      <td>20.92</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>226.16</td>\n",
       "      <td>181.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2022-01-25 02:50:00</td>\n",
       "      <td>79.08</td>\n",
       "      <td>23.52</td>\n",
       "      <td>12.41</td>\n",
       "      <td>17.86</td>\n",
       "      <td>1273.320333</td>\n",
       "      <td>937.405000</td>\n",
       "      <td>55.26</td>\n",
       "      <td>108.03</td>\n",
       "      <td>43.29</td>\n",
       "      <td>20.92</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>226.16</td>\n",
       "      <td>181.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2022-01-25 03:50:00</td>\n",
       "      <td>79.29</td>\n",
       "      <td>22.94</td>\n",
       "      <td>11.72</td>\n",
       "      <td>17.86</td>\n",
       "      <td>1273.317500</td>\n",
       "      <td>937.401333</td>\n",
       "      <td>55.26</td>\n",
       "      <td>108.03</td>\n",
       "      <td>43.29</td>\n",
       "      <td>20.92</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>242.44</td>\n",
       "      <td>164.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2022-01-25 04:50:00</td>\n",
       "      <td>79.95</td>\n",
       "      <td>21.42</td>\n",
       "      <td>10.68</td>\n",
       "      <td>17.63</td>\n",
       "      <td>1273.148333</td>\n",
       "      <td>937.256667</td>\n",
       "      <td>55.26</td>\n",
       "      <td>108.03</td>\n",
       "      <td>43.29</td>\n",
       "      <td>20.92</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>242.44</td>\n",
       "      <td>164.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1648</th>\n",
       "      <td>2022-04-07 19:50:00</td>\n",
       "      <td>79.82</td>\n",
       "      <td>23.84</td>\n",
       "      <td>11.03</td>\n",
       "      <td>13.52</td>\n",
       "      <td>495.109833</td>\n",
       "      <td>557.570833</td>\n",
       "      <td>54.40</td>\n",
       "      <td>105.14</td>\n",
       "      <td>49.03</td>\n",
       "      <td>20.82</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>268.69</td>\n",
       "      <td>145.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1649</th>\n",
       "      <td>2022-04-07 20:50:00</td>\n",
       "      <td>78.98</td>\n",
       "      <td>25.36</td>\n",
       "      <td>11.37</td>\n",
       "      <td>12.85</td>\n",
       "      <td>495.076000</td>\n",
       "      <td>571.479667</td>\n",
       "      <td>54.40</td>\n",
       "      <td>105.14</td>\n",
       "      <td>49.03</td>\n",
       "      <td>20.82</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>268.69</td>\n",
       "      <td>145.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1650</th>\n",
       "      <td>2022-04-07 21:50:00</td>\n",
       "      <td>78.86</td>\n",
       "      <td>25.40</td>\n",
       "      <td>11.37</td>\n",
       "      <td>11.42</td>\n",
       "      <td>494.801333</td>\n",
       "      <td>571.781000</td>\n",
       "      <td>54.40</td>\n",
       "      <td>105.14</td>\n",
       "      <td>49.03</td>\n",
       "      <td>20.82</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>303.85</td>\n",
       "      <td>144.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1651</th>\n",
       "      <td>2022-04-07 22:50:00</td>\n",
       "      <td>79.10</td>\n",
       "      <td>25.58</td>\n",
       "      <td>11.37</td>\n",
       "      <td>11.55</td>\n",
       "      <td>495.090167</td>\n",
       "      <td>571.744333</td>\n",
       "      <td>54.40</td>\n",
       "      <td>105.14</td>\n",
       "      <td>49.03</td>\n",
       "      <td>20.82</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>303.85</td>\n",
       "      <td>144.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1652</th>\n",
       "      <td>2022-04-07 23:50:00</td>\n",
       "      <td>79.32</td>\n",
       "      <td>24.82</td>\n",
       "      <td>11.03</td>\n",
       "      <td>11.55</td>\n",
       "      <td>434.989333</td>\n",
       "      <td>540.514333</td>\n",
       "      <td>54.40</td>\n",
       "      <td>105.14</td>\n",
       "      <td>49.03</td>\n",
       "      <td>20.82</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>303.85</td>\n",
       "      <td>144.41</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1653 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Time  index_A  index_B  index_C  index_D  \\\n",
       "0    2022-01-25 00:50:00    78.31    23.66    12.24    17.81   \n",
       "1    2022-01-25 01:50:00    78.46    23.88    12.41    17.99   \n",
       "2    2022-01-25 02:50:00    79.08    23.52    12.41    17.86   \n",
       "3    2022-01-25 03:50:00    79.29    22.94    11.72    17.86   \n",
       "4    2022-01-25 04:50:00    79.95    21.42    10.68    17.63   \n",
       "...                  ...      ...      ...      ...      ...   \n",
       "1648 2022-04-07 19:50:00    79.82    23.84    11.03    13.52   \n",
       "1649 2022-04-07 20:50:00    78.98    25.36    11.37    12.85   \n",
       "1650 2022-04-07 21:50:00    78.86    25.40    11.37    11.42   \n",
       "1651 2022-04-07 22:50:00    79.10    25.58    11.37    11.55   \n",
       "1652 2022-04-07 23:50:00    79.32    24.82    11.03    11.55   \n",
       "\n",
       "      Temperature_of_system1  Temperature_of_system2  Mineral_parameter1  \\\n",
       "0                1343.879322              949.674915               55.26   \n",
       "1                1273.313000              937.436333               55.26   \n",
       "2                1273.320333              937.405000               55.26   \n",
       "3                1273.317500              937.401333               55.26   \n",
       "4                1273.148333              937.256667               55.26   \n",
       "...                      ...                     ...                 ...   \n",
       "1648              495.109833              557.570833               54.40   \n",
       "1649              495.076000              571.479667               54.40   \n",
       "1650              494.801333              571.781000               54.40   \n",
       "1651              495.090167              571.744333               54.40   \n",
       "1652              434.989333              540.514333               54.40   \n",
       "\n",
       "      Mineral_parameter2  Mineral_parameter3  Mineral_parameter4  \\\n",
       "0                 108.03               43.29               20.92   \n",
       "1                 108.03               43.29               20.92   \n",
       "2                 108.03               43.29               20.92   \n",
       "3                 108.03               43.29               20.92   \n",
       "4                 108.03               43.29               20.92   \n",
       "...                  ...                 ...                 ...   \n",
       "1648              105.14               49.03               20.82   \n",
       "1649              105.14               49.03               20.82   \n",
       "1650              105.14               49.03               20.82   \n",
       "1651              105.14               49.03               20.82   \n",
       "1652              105.14               49.03               20.82   \n",
       "\n",
       "      Process_parameter1  Process_parameter2  Process_parameter3  \\\n",
       "0                   1.25                3.09              226.16   \n",
       "1                   1.25                3.09              226.16   \n",
       "2                   1.25                3.09              226.16   \n",
       "3                   1.25                3.09              242.44   \n",
       "4                   1.25                3.09              242.44   \n",
       "...                  ...                 ...                 ...   \n",
       "1648                1.25                3.09              268.69   \n",
       "1649                1.25                3.09              268.69   \n",
       "1650                1.25                3.09              303.85   \n",
       "1651                1.25                3.09              303.85   \n",
       "1652                1.25                3.09              303.85   \n",
       "\n",
       "      Process_parameter4  \n",
       "0                 181.23  \n",
       "1                 181.23  \n",
       "2                 181.23  \n",
       "3                 164.45  \n",
       "4                 164.45  \n",
       "...                  ...  \n",
       "1648              145.94  \n",
       "1649              145.94  \n",
       "1650              144.41  \n",
       "1651              144.41  \n",
       "1652              144.41  \n",
       "\n",
       "[1653 rows x 15 columns]"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 导入数据\n",
    "\n",
    "data = pd.read_excel(\"data.xlsx\")\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1653 entries, 0 to 1652\n",
      "Data columns (total 15 columns):\n",
      " #   Column                  Non-Null Count  Dtype         \n",
      "---  ------                  --------------  -----         \n",
      " 0   Time                    1653 non-null   datetime64[ns]\n",
      " 1   index_A                 1653 non-null   float64       \n",
      " 2   index_B                 1653 non-null   float64       \n",
      " 3   index_C                 1653 non-null   float64       \n",
      " 4   index_D                 1653 non-null   float64       \n",
      " 5   Temperature_of_system1  1653 non-null   float64       \n",
      " 6   Temperature_of_system2  1653 non-null   float64       \n",
      " 7   Mineral_parameter1      1653 non-null   float64       \n",
      " 8   Mineral_parameter2      1653 non-null   float64       \n",
      " 9   Mineral_parameter3      1653 non-null   float64       \n",
      " 10  Mineral_parameter4      1653 non-null   float64       \n",
      " 11  Process_parameter1      1653 non-null   float64       \n",
      " 12  Process_parameter2      1653 non-null   float64       \n",
      " 13  Process_parameter3      1653 non-null   float64       \n",
      " 14  Process_parameter4      1653 non-null   float64       \n",
      "dtypes: datetime64[ns](1), float64(14)\n",
      "memory usage: 193.8 KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "# sns.pairplot(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "name = [\"index_A\",\"index_B\",\"index_C\",\"index_D\",\"Temperature_of_system1\",\t\"Temperature_of_system2\",\t\"Mineral_parameter1\",\t\"Mineral_parameter2\",\t\"Mineral_parameter3\",\t\"Mineral_parameter4\",\"Process_parameter3\",\"Process_parameter4\"]\n",
    "# sns.heatmap(name)\n",
    "date = data[name]\n",
    "# sns.pairplot(date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x864 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "df = pd.DataFrame(date)\n",
    "corr = df.corr()\n",
    "plt.figure(figsize=(16,12))\n",
    "sns.set_context('paper',font_scale=1.4)\n",
    "sns.heatmap(corr, cmap='Blues', annot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "```\n",
    "from sklearn.datasets import make_regression\n",
    "from sklearn.multioutput import MultiOutputRegressor\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "\n",
    "# Generate dataset\n",
    "X, y = make_regression(n_samples=25000, n_features=3, n_targets=2, random_state=33)\n",
    "\n",
    "# Train/test split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=33)\n",
    "\n",
    "# Create the SVR regressor\n",
    "svr = SVR(epsilon=0.2)\n",
    "\n",
    "# Create the Multioutput Regressor\n",
    "mor = MultiOutputRegressor(svr)\n",
    "\n",
    "# Train the regressor\n",
    "mor = mor.fit(X_train, y_train)\n",
    "\n",
    "# Generate predictions for testing data\n",
    "y_pred = mor.predict(X_test)\n",
    "\n",
    "# Evaluate the regressor\n",
    "mse_one = mean_squared_error(y_test[:,0], y_pred[:,0])\n",
    "mse_two = mean_squared_error(y_test[:,1], y_pred[:,1])\n",
    "print(f'MSE for first regressor: {mse_one} - second regressor: {mse_two}')\n",
    "mae_one = mean_absolute_error(y_test[:,0], y_pred[:,0])\n",
    "mae_two = mean_absolute_error(y_test[:,1], y_pred[:,1])\n",
    "print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "```\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 0.9027724\ttotal: 15.4ms\tremaining: 4.6s\n",
      "1:\tlearn: 0.8857041\ttotal: 30.5ms\tremaining: 4.54s\n",
      "2:\tlearn: 0.8707392\ttotal: 46.3ms\tremaining: 4.58s\n",
      "3:\tlearn: 0.8559111\ttotal: 61.1ms\tremaining: 4.52s\n",
      "4:\tlearn: 0.8394876\ttotal: 75.8ms\tremaining: 4.47s\n",
      "5:\tlearn: 0.8263451\ttotal: 90.4ms\tremaining: 4.43s\n",
      "6:\tlearn: 0.8156749\ttotal: 106ms\tremaining: 4.42s\n",
      "7:\tlearn: 0.8066221\ttotal: 120ms\tremaining: 4.39s\n",
      "8:\tlearn: 0.7957522\ttotal: 138ms\tremaining: 4.45s\n",
      "9:\tlearn: 0.7794513\ttotal: 157ms\tremaining: 4.54s\n",
      "10:\tlearn: 0.7689725\ttotal: 190ms\tremaining: 4.98s\n",
      "11:\tlearn: 0.7583217\ttotal: 224ms\tremaining: 5.38s\n",
      "12:\tlearn: 0.7469247\ttotal: 260ms\tremaining: 5.74s\n",
      "13:\tlearn: 0.7363856\ttotal: 297ms\tremaining: 6.06s\n",
      "14:\tlearn: 0.7295647\ttotal: 333ms\tremaining: 6.32s\n",
      "15:\tlearn: 0.7178457\ttotal: 370ms\tremaining: 6.57s\n",
      "16:\tlearn: 0.7097848\ttotal: 404ms\tremaining: 6.73s\n",
      "17:\tlearn: 0.7018915\ttotal: 439ms\tremaining: 6.87s\n",
      "18:\tlearn: 0.6946985\ttotal: 474ms\tremaining: 7.01s\n",
      "19:\tlearn: 0.6868309\ttotal: 509ms\tremaining: 7.12s\n",
      "20:\tlearn: 0.6799213\ttotal: 545ms\tremaining: 7.25s\n",
      "21:\tlearn: 0.6748257\ttotal: 579ms\tremaining: 7.32s\n",
      "22:\tlearn: 0.6685523\ttotal: 611ms\tremaining: 7.35s\n",
      "23:\tlearn: 0.6591215\ttotal: 626ms\tremaining: 7.19s\n",
      "24:\tlearn: 0.6527215\ttotal: 641ms\tremaining: 7.05s\n",
      "25:\tlearn: 0.6444758\ttotal: 656ms\tremaining: 6.91s\n",
      "26:\tlearn: 0.6379541\ttotal: 672ms\tremaining: 6.79s\n",
      "27:\tlearn: 0.6326622\ttotal: 687ms\tremaining: 6.67s\n",
      "28:\tlearn: 0.6267595\ttotal: 703ms\tremaining: 6.57s\n",
      "29:\tlearn: 0.6234496\ttotal: 719ms\tremaining: 6.47s\n",
      "30:\tlearn: 0.6178652\ttotal: 735ms\tremaining: 6.37s\n",
      "31:\tlearn: 0.6125573\ttotal: 751ms\tremaining: 6.29s\n",
      "32:\tlearn: 0.6082044\ttotal: 768ms\tremaining: 6.21s\n",
      "33:\tlearn: 0.6016197\ttotal: 788ms\tremaining: 6.17s\n",
      "34:\tlearn: 0.5966852\ttotal: 816ms\tremaining: 6.17s\n",
      "35:\tlearn: 0.5919425\ttotal: 843ms\tremaining: 6.18s\n",
      "36:\tlearn: 0.5873068\ttotal: 870ms\tremaining: 6.18s\n",
      "37:\tlearn: 0.5857368\ttotal: 881ms\tremaining: 6.07s\n",
      "38:\tlearn: 0.5815310\ttotal: 918ms\tremaining: 6.14s\n",
      "39:\tlearn: 0.5786851\ttotal: 957ms\tremaining: 6.22s\n",
      "40:\tlearn: 0.5747431\ttotal: 993ms\tremaining: 6.27s\n",
      "41:\tlearn: 0.5713537\ttotal: 1.03s\tremaining: 6.32s\n",
      "42:\tlearn: 0.5682074\ttotal: 1.06s\tremaining: 6.35s\n",
      "43:\tlearn: 0.5650709\ttotal: 1.08s\tremaining: 6.3s\n",
      "44:\tlearn: 0.5617470\ttotal: 1.1s\tremaining: 6.25s\n",
      "45:\tlearn: 0.5572216\ttotal: 1.12s\tremaining: 6.18s\n",
      "46:\tlearn: 0.5532515\ttotal: 1.14s\tremaining: 6.11s\n",
      "47:\tlearn: 0.5505857\ttotal: 1.15s\tremaining: 6.05s\n",
      "48:\tlearn: 0.5486121\ttotal: 1.17s\tremaining: 5.98s\n",
      "49:\tlearn: 0.5473484\ttotal: 1.17s\tremaining: 5.86s\n",
      "50:\tlearn: 0.5441234\ttotal: 1.19s\tremaining: 5.81s\n",
      "51:\tlearn: 0.5399666\ttotal: 1.21s\tremaining: 5.75s\n",
      "52:\tlearn: 0.5372123\ttotal: 1.22s\tremaining: 5.7s\n",
      "53:\tlearn: 0.5338176\ttotal: 1.24s\tremaining: 5.64s\n",
      "54:\tlearn: 0.5315230\ttotal: 1.25s\tremaining: 5.59s\n",
      "55:\tlearn: 0.5289176\ttotal: 1.27s\tremaining: 5.55s\n",
      "56:\tlearn: 0.5257941\ttotal: 1.29s\tremaining: 5.52s\n",
      "57:\tlearn: 0.5237948\ttotal: 1.31s\tremaining: 5.48s\n",
      "58:\tlearn: 0.5212155\ttotal: 1.33s\tremaining: 5.44s\n",
      "59:\tlearn: 0.5208975\ttotal: 1.33s\tremaining: 5.33s\n",
      "60:\tlearn: 0.5186644\ttotal: 1.35s\tremaining: 5.28s\n",
      "61:\tlearn: 0.5170268\ttotal: 1.36s\tremaining: 5.23s\n",
      "62:\tlearn: 0.5143009\ttotal: 1.38s\tremaining: 5.18s\n",
      "63:\tlearn: 0.5114821\ttotal: 1.39s\tremaining: 5.14s\n",
      "64:\tlearn: 0.5087542\ttotal: 1.41s\tremaining: 5.09s\n",
      "65:\tlearn: 0.5042334\ttotal: 1.43s\tremaining: 5.05s\n",
      "66:\tlearn: 0.5021766\ttotal: 1.44s\tremaining: 5.02s\n",
      "67:\tlearn: 0.5003120\ttotal: 1.46s\tremaining: 4.98s\n",
      "68:\tlearn: 0.4974088\ttotal: 1.48s\tremaining: 4.95s\n",
      "69:\tlearn: 0.4950380\ttotal: 1.5s\tremaining: 4.93s\n",
      "70:\tlearn: 0.4923716\ttotal: 1.52s\tremaining: 4.9s\n",
      "71:\tlearn: 0.4903903\ttotal: 1.54s\tremaining: 4.88s\n",
      "72:\tlearn: 0.4871405\ttotal: 1.56s\tremaining: 4.85s\n",
      "73:\tlearn: 0.4854978\ttotal: 1.58s\tremaining: 4.82s\n",
      "74:\tlearn: 0.4838173\ttotal: 1.6s\tremaining: 4.8s\n",
      "75:\tlearn: 0.4808359\ttotal: 1.62s\tremaining: 4.78s\n",
      "76:\tlearn: 0.4791120\ttotal: 1.64s\tremaining: 4.76s\n",
      "77:\tlearn: 0.4776614\ttotal: 1.67s\tremaining: 4.74s\n",
      "78:\tlearn: 0.4760998\ttotal: 1.69s\tremaining: 4.73s\n",
      "79:\tlearn: 0.4731619\ttotal: 1.71s\tremaining: 4.71s\n",
      "80:\tlearn: 0.4705886\ttotal: 1.73s\tremaining: 4.68s\n",
      "81:\tlearn: 0.4695408\ttotal: 1.74s\tremaining: 4.61s\n",
      "82:\tlearn: 0.4674592\ttotal: 1.75s\tremaining: 4.58s\n",
      "83:\tlearn: 0.4658844\ttotal: 1.77s\tremaining: 4.54s\n",
      "84:\tlearn: 0.4629767\ttotal: 1.78s\tremaining: 4.51s\n",
      "85:\tlearn: 0.4612918\ttotal: 1.8s\tremaining: 4.48s\n",
      "86:\tlearn: 0.4589271\ttotal: 1.81s\tremaining: 4.44s\n",
      "87:\tlearn: 0.4574091\ttotal: 1.83s\tremaining: 4.41s\n",
      "88:\tlearn: 0.4551623\ttotal: 1.84s\tremaining: 4.38s\n",
      "89:\tlearn: 0.4528067\ttotal: 1.86s\tremaining: 4.34s\n",
      "90:\tlearn: 0.4511643\ttotal: 1.88s\tremaining: 4.31s\n",
      "91:\tlearn: 0.4487880\ttotal: 1.89s\tremaining: 4.28s\n",
      "92:\tlearn: 0.4468585\ttotal: 1.91s\tremaining: 4.26s\n",
      "93:\tlearn: 0.4455784\ttotal: 1.93s\tremaining: 4.23s\n",
      "94:\tlearn: 0.4444577\ttotal: 1.95s\tremaining: 4.21s\n",
      "95:\tlearn: 0.4428975\ttotal: 1.97s\tremaining: 4.18s\n",
      "96:\tlearn: 0.4409071\ttotal: 1.99s\tremaining: 4.17s\n",
      "97:\tlearn: 0.4393545\ttotal: 2.02s\tremaining: 4.17s\n",
      "98:\tlearn: 0.4384534\ttotal: 2.07s\tremaining: 4.21s\n",
      "99:\tlearn: 0.4372094\ttotal: 2.12s\tremaining: 4.24s\n",
      "100:\tlearn: 0.4360251\ttotal: 2.16s\tremaining: 4.26s\n",
      "101:\tlearn: 0.4348828\ttotal: 2.21s\tremaining: 4.28s\n",
      "102:\tlearn: 0.4325371\ttotal: 2.25s\tremaining: 4.3s\n",
      "103:\tlearn: 0.4307909\ttotal: 2.29s\tremaining: 4.32s\n",
      "104:\tlearn: 0.4283043\ttotal: 2.34s\tremaining: 4.34s\n",
      "105:\tlearn: 0.4273596\ttotal: 2.38s\tremaining: 4.36s\n",
      "106:\tlearn: 0.4257999\ttotal: 2.41s\tremaining: 4.35s\n",
      "107:\tlearn: 0.4244279\ttotal: 2.43s\tremaining: 4.32s\n",
      "108:\tlearn: 0.4222873\ttotal: 2.45s\tremaining: 4.29s\n",
      "109:\tlearn: 0.4205941\ttotal: 2.46s\tremaining: 4.25s\n",
      "110:\tlearn: 0.4190632\ttotal: 2.48s\tremaining: 4.22s\n",
      "111:\tlearn: 0.4176217\ttotal: 2.5s\tremaining: 4.19s\n",
      "112:\tlearn: 0.4162377\ttotal: 2.52s\tremaining: 4.16s\n",
      "113:\tlearn: 0.4139198\ttotal: 2.54s\tremaining: 4.14s\n",
      "114:\tlearn: 0.4120281\ttotal: 2.56s\tremaining: 4.12s\n",
      "115:\tlearn: 0.4100696\ttotal: 2.59s\tremaining: 4.11s\n",
      "116:\tlearn: 0.4082171\ttotal: 2.63s\tremaining: 4.12s\n",
      "117:\tlearn: 0.4080598\ttotal: 2.64s\tremaining: 4.07s\n",
      "118:\tlearn: 0.4069743\ttotal: 2.67s\tremaining: 4.07s\n",
      "119:\tlearn: 0.4058691\ttotal: 2.71s\tremaining: 4.07s\n",
      "120:\tlearn: 0.4037229\ttotal: 2.75s\tremaining: 4.07s\n",
      "121:\tlearn: 0.4021740\ttotal: 2.79s\tremaining: 4.07s\n",
      "122:\tlearn: 0.4007067\ttotal: 2.83s\tremaining: 4.07s\n",
      "123:\tlearn: 0.3994914\ttotal: 2.87s\tremaining: 4.08s\n",
      "124:\tlearn: 0.3974348\ttotal: 2.91s\tremaining: 4.08s\n",
      "125:\tlearn: 0.3957182\ttotal: 2.95s\tremaining: 4.07s\n",
      "126:\tlearn: 0.3942443\ttotal: 2.99s\tremaining: 4.07s\n",
      "127:\tlearn: 0.3930287\ttotal: 3.03s\tremaining: 4.07s\n",
      "128:\tlearn: 0.3918003\ttotal: 3.07s\tremaining: 4.08s\n",
      "129:\tlearn: 0.3899529\ttotal: 3.12s\tremaining: 4.07s\n",
      "130:\tlearn: 0.3886634\ttotal: 3.16s\tremaining: 4.07s\n",
      "131:\tlearn: 0.3864876\ttotal: 3.2s\tremaining: 4.08s\n",
      "132:\tlearn: 0.3851861\ttotal: 3.25s\tremaining: 4.08s\n",
      "133:\tlearn: 0.3837239\ttotal: 3.29s\tremaining: 4.08s\n",
      "134:\tlearn: 0.3819281\ttotal: 3.32s\tremaining: 4.06s\n",
      "135:\tlearn: 0.3812554\ttotal: 3.34s\tremaining: 4.03s\n",
      "136:\tlearn: 0.3799933\ttotal: 3.36s\tremaining: 4s\n",
      "137:\tlearn: 0.3783320\ttotal: 3.38s\tremaining: 3.97s\n",
      "138:\tlearn: 0.3769432\ttotal: 3.4s\tremaining: 3.94s\n",
      "139:\tlearn: 0.3761918\ttotal: 3.41s\tremaining: 3.9s\n",
      "140:\tlearn: 0.3751741\ttotal: 3.43s\tremaining: 3.86s\n",
      "141:\tlearn: 0.3736062\ttotal: 3.44s\tremaining: 3.83s\n",
      "142:\tlearn: 0.3721760\ttotal: 3.46s\tremaining: 3.8s\n",
      "143:\tlearn: 0.3715466\ttotal: 3.48s\tremaining: 3.77s\n",
      "144:\tlearn: 0.3711303\ttotal: 3.5s\tremaining: 3.74s\n",
      "145:\tlearn: 0.3700866\ttotal: 3.52s\tremaining: 3.71s\n",
      "146:\tlearn: 0.3692756\ttotal: 3.54s\tremaining: 3.68s\n",
      "147:\tlearn: 0.3683109\ttotal: 3.56s\tremaining: 3.65s\n",
      "148:\tlearn: 0.3672390\ttotal: 3.58s\tremaining: 3.62s\n",
      "149:\tlearn: 0.3663716\ttotal: 3.59s\tremaining: 3.59s\n",
      "150:\tlearn: 0.3651917\ttotal: 3.62s\tremaining: 3.57s\n",
      "151:\tlearn: 0.3642049\ttotal: 3.63s\tremaining: 3.54s\n",
      "152:\tlearn: 0.3635178\ttotal: 3.65s\tremaining: 3.51s\n",
      "153:\tlearn: 0.3634101\ttotal: 3.66s\tremaining: 3.47s\n",
      "154:\tlearn: 0.3630363\ttotal: 3.68s\tremaining: 3.44s\n",
      "155:\tlearn: 0.3616072\ttotal: 3.69s\tremaining: 3.41s\n",
      "156:\tlearn: 0.3606319\ttotal: 3.71s\tremaining: 3.38s\n",
      "157:\tlearn: 0.3597550\ttotal: 3.73s\tremaining: 3.35s\n",
      "158:\tlearn: 0.3583221\ttotal: 3.75s\tremaining: 3.32s\n",
      "159:\tlearn: 0.3572454\ttotal: 3.76s\tremaining: 3.29s\n",
      "160:\tlearn: 0.3564388\ttotal: 3.78s\tremaining: 3.26s\n",
      "161:\tlearn: 0.3553883\ttotal: 3.8s\tremaining: 3.23s\n",
      "162:\tlearn: 0.3540354\ttotal: 3.81s\tremaining: 3.2s\n",
      "163:\tlearn: 0.3528781\ttotal: 3.83s\tremaining: 3.17s\n",
      "164:\tlearn: 0.3520968\ttotal: 3.85s\tremaining: 3.15s\n",
      "165:\tlearn: 0.3505433\ttotal: 3.86s\tremaining: 3.12s\n",
      "166:\tlearn: 0.3491952\ttotal: 3.88s\tremaining: 3.09s\n",
      "167:\tlearn: 0.3482282\ttotal: 3.9s\tremaining: 3.06s\n",
      "168:\tlearn: 0.3471578\ttotal: 3.92s\tremaining: 3.04s\n",
      "169:\tlearn: 0.3463204\ttotal: 3.93s\tremaining: 3.01s\n",
      "170:\tlearn: 0.3458090\ttotal: 3.95s\tremaining: 2.98s\n",
      "171:\tlearn: 0.3454359\ttotal: 3.97s\tremaining: 2.95s\n",
      "172:\tlearn: 0.3443743\ttotal: 3.98s\tremaining: 2.92s\n",
      "173:\tlearn: 0.3437467\ttotal: 4s\tremaining: 2.9s\n",
      "174:\tlearn: 0.3427764\ttotal: 4.02s\tremaining: 2.87s\n",
      "175:\tlearn: 0.3411939\ttotal: 4.04s\tremaining: 2.84s\n",
      "176:\tlearn: 0.3404943\ttotal: 4.05s\tremaining: 2.82s\n",
      "177:\tlearn: 0.3395144\ttotal: 4.07s\tremaining: 2.79s\n",
      "178:\tlearn: 0.3385349\ttotal: 4.09s\tremaining: 2.77s\n",
      "179:\tlearn: 0.3378181\ttotal: 4.11s\tremaining: 2.74s\n",
      "180:\tlearn: 0.3373249\ttotal: 4.13s\tremaining: 2.71s\n",
      "181:\tlearn: 0.3365379\ttotal: 4.14s\tremaining: 2.69s\n",
      "182:\tlearn: 0.3355158\ttotal: 4.16s\tremaining: 2.66s\n",
      "183:\tlearn: 0.3348711\ttotal: 4.18s\tremaining: 2.63s\n",
      "184:\tlearn: 0.3338303\ttotal: 4.19s\tremaining: 2.61s\n",
      "185:\tlearn: 0.3327377\ttotal: 4.21s\tremaining: 2.58s\n",
      "186:\tlearn: 0.3324992\ttotal: 4.22s\tremaining: 2.55s\n",
      "187:\tlearn: 0.3313289\ttotal: 4.24s\tremaining: 2.52s\n",
      "188:\tlearn: 0.3306640\ttotal: 4.26s\tremaining: 2.5s\n",
      "189:\tlearn: 0.3297856\ttotal: 4.29s\tremaining: 2.48s\n",
      "190:\tlearn: 0.3291039\ttotal: 4.31s\tremaining: 2.46s\n",
      "191:\tlearn: 0.3285639\ttotal: 4.32s\tremaining: 2.43s\n",
      "192:\tlearn: 0.3281625\ttotal: 4.34s\tremaining: 2.4s\n",
      "193:\tlearn: 0.3274615\ttotal: 4.36s\tremaining: 2.38s\n",
      "194:\tlearn: 0.3260047\ttotal: 4.37s\tremaining: 2.35s\n",
      "195:\tlearn: 0.3258525\ttotal: 4.38s\tremaining: 2.32s\n",
      "196:\tlearn: 0.3249230\ttotal: 4.39s\tremaining: 2.29s\n",
      "197:\tlearn: 0.3242044\ttotal: 4.41s\tremaining: 2.27s\n",
      "198:\tlearn: 0.3235310\ttotal: 4.42s\tremaining: 2.25s\n",
      "199:\tlearn: 0.3225984\ttotal: 4.44s\tremaining: 2.22s\n",
      "200:\tlearn: 0.3219617\ttotal: 4.46s\tremaining: 2.2s\n",
      "201:\tlearn: 0.3211982\ttotal: 4.48s\tremaining: 2.17s\n",
      "202:\tlearn: 0.3205516\ttotal: 4.51s\tremaining: 2.15s\n",
      "203:\tlearn: 0.3198434\ttotal: 4.53s\tremaining: 2.13s\n",
      "204:\tlearn: 0.3192790\ttotal: 4.55s\tremaining: 2.11s\n",
      "205:\tlearn: 0.3187266\ttotal: 4.57s\tremaining: 2.09s\n",
      "206:\tlearn: 0.3185153\ttotal: 4.6s\tremaining: 2.06s\n",
      "207:\tlearn: 0.3173375\ttotal: 4.62s\tremaining: 2.04s\n",
      "208:\tlearn: 0.3167788\ttotal: 4.65s\tremaining: 2.02s\n",
      "209:\tlearn: 0.3161590\ttotal: 4.68s\tremaining: 2s\n",
      "210:\tlearn: 0.3154829\ttotal: 4.71s\tremaining: 1.98s\n",
      "211:\tlearn: 0.3147803\ttotal: 4.73s\tremaining: 1.96s\n",
      "212:\tlearn: 0.3137334\ttotal: 4.76s\tremaining: 1.94s\n",
      "213:\tlearn: 0.3129189\ttotal: 4.79s\tremaining: 1.92s\n",
      "214:\tlearn: 0.3124724\ttotal: 4.82s\tremaining: 1.9s\n",
      "215:\tlearn: 0.3120558\ttotal: 4.84s\tremaining: 1.88s\n",
      "216:\tlearn: 0.3116885\ttotal: 4.87s\tremaining: 1.86s\n",
      "217:\tlearn: 0.3109412\ttotal: 4.91s\tremaining: 1.84s\n",
      "218:\tlearn: 0.3099780\ttotal: 4.93s\tremaining: 1.82s\n",
      "219:\tlearn: 0.3092318\ttotal: 4.96s\tremaining: 1.8s\n",
      "220:\tlearn: 0.3084630\ttotal: 4.99s\tremaining: 1.78s\n",
      "221:\tlearn: 0.3082392\ttotal: 5.01s\tremaining: 1.76s\n",
      "222:\tlearn: 0.3075669\ttotal: 5.04s\tremaining: 1.74s\n",
      "223:\tlearn: 0.3069930\ttotal: 5.07s\tremaining: 1.72s\n",
      "224:\tlearn: 0.3062942\ttotal: 5.1s\tremaining: 1.7s\n",
      "225:\tlearn: 0.3058720\ttotal: 5.13s\tremaining: 1.68s\n",
      "226:\tlearn: 0.3049241\ttotal: 5.15s\tremaining: 1.66s\n",
      "227:\tlearn: 0.3042129\ttotal: 5.17s\tremaining: 1.63s\n",
      "228:\tlearn: 0.3038060\ttotal: 5.18s\tremaining: 1.61s\n",
      "229:\tlearn: 0.3030035\ttotal: 5.2s\tremaining: 1.58s\n",
      "230:\tlearn: 0.3022542\ttotal: 5.22s\tremaining: 1.56s\n",
      "231:\tlearn: 0.3018630\ttotal: 5.24s\tremaining: 1.54s\n",
      "232:\tlearn: 0.3011805\ttotal: 5.26s\tremaining: 1.51s\n",
      "233:\tlearn: 0.3008401\ttotal: 5.28s\tremaining: 1.49s\n",
      "234:\tlearn: 0.3000377\ttotal: 5.3s\tremaining: 1.47s\n",
      "235:\tlearn: 0.2995787\ttotal: 5.32s\tremaining: 1.44s\n",
      "236:\tlearn: 0.2990686\ttotal: 5.34s\tremaining: 1.42s\n",
      "237:\tlearn: 0.2986950\ttotal: 5.36s\tremaining: 1.4s\n",
      "238:\tlearn: 0.2979496\ttotal: 5.38s\tremaining: 1.37s\n",
      "239:\tlearn: 0.2974020\ttotal: 5.4s\tremaining: 1.35s\n",
      "240:\tlearn: 0.2968514\ttotal: 5.42s\tremaining: 1.33s\n",
      "241:\tlearn: 0.2964699\ttotal: 5.45s\tremaining: 1.3s\n",
      "242:\tlearn: 0.2957398\ttotal: 5.47s\tremaining: 1.28s\n",
      "243:\tlearn: 0.2951971\ttotal: 5.49s\tremaining: 1.26s\n",
      "244:\tlearn: 0.2945958\ttotal: 5.51s\tremaining: 1.24s\n",
      "245:\tlearn: 0.2939149\ttotal: 5.54s\tremaining: 1.22s\n",
      "246:\tlearn: 0.2933055\ttotal: 5.56s\tremaining: 1.19s\n",
      "247:\tlearn: 0.2928785\ttotal: 5.59s\tremaining: 1.17s\n",
      "248:\tlearn: 0.2922725\ttotal: 5.61s\tremaining: 1.15s\n",
      "249:\tlearn: 0.2917317\ttotal: 5.64s\tremaining: 1.13s\n",
      "250:\tlearn: 0.2913579\ttotal: 5.67s\tremaining: 1.11s\n",
      "251:\tlearn: 0.2907833\ttotal: 5.69s\tremaining: 1.08s\n",
      "252:\tlearn: 0.2903579\ttotal: 5.71s\tremaining: 1.06s\n",
      "253:\tlearn: 0.2897200\ttotal: 5.74s\tremaining: 1.04s\n",
      "254:\tlearn: 0.2894511\ttotal: 5.76s\tremaining: 1.02s\n",
      "255:\tlearn: 0.2890175\ttotal: 5.78s\tremaining: 994ms\n",
      "256:\tlearn: 0.2886175\ttotal: 5.82s\tremaining: 973ms\n",
      "257:\tlearn: 0.2878008\ttotal: 5.85s\tremaining: 953ms\n",
      "258:\tlearn: 0.2872126\ttotal: 5.89s\tremaining: 932ms\n",
      "259:\tlearn: 0.2864566\ttotal: 5.92s\tremaining: 911ms\n",
      "260:\tlearn: 0.2858082\ttotal: 5.96s\tremaining: 891ms\n",
      "261:\tlearn: 0.2854165\ttotal: 6s\tremaining: 870ms\n",
      "262:\tlearn: 0.2849273\ttotal: 6.03s\tremaining: 849ms\n",
      "263:\tlearn: 0.2845772\ttotal: 6.07s\tremaining: 828ms\n",
      "264:\tlearn: 0.2841200\ttotal: 6.11s\tremaining: 806ms\n",
      "265:\tlearn: 0.2834014\ttotal: 6.14s\tremaining: 785ms\n",
      "266:\tlearn: 0.2831080\ttotal: 6.18s\tremaining: 764ms\n",
      "267:\tlearn: 0.2826561\ttotal: 6.22s\tremaining: 742ms\n",
      "268:\tlearn: 0.2821802\ttotal: 6.25s\tremaining: 721ms\n",
      "269:\tlearn: 0.2815142\ttotal: 6.29s\tremaining: 698ms\n",
      "270:\tlearn: 0.2808675\ttotal: 6.31s\tremaining: 675ms\n",
      "271:\tlearn: 0.2804088\ttotal: 6.33s\tremaining: 652ms\n",
      "272:\tlearn: 0.2799232\ttotal: 6.36s\tremaining: 629ms\n",
      "273:\tlearn: 0.2796013\ttotal: 6.39s\tremaining: 606ms\n",
      "274:\tlearn: 0.2790405\ttotal: 6.41s\tremaining: 583ms\n",
      "275:\tlearn: 0.2787007\ttotal: 6.43s\tremaining: 560ms\n",
      "276:\tlearn: 0.2784335\ttotal: 6.46s\tremaining: 536ms\n",
      "277:\tlearn: 0.2777802\ttotal: 6.47s\tremaining: 512ms\n",
      "278:\tlearn: 0.2774006\ttotal: 6.49s\tremaining: 489ms\n",
      "279:\tlearn: 0.2769195\ttotal: 6.51s\tremaining: 465ms\n",
      "280:\tlearn: 0.2763484\ttotal: 6.53s\tremaining: 441ms\n",
      "281:\tlearn: 0.2759783\ttotal: 6.54s\tremaining: 418ms\n",
      "282:\tlearn: 0.2755944\ttotal: 6.57s\tremaining: 395ms\n",
      "283:\tlearn: 0.2750944\ttotal: 6.59s\tremaining: 371ms\n",
      "284:\tlearn: 0.2742052\ttotal: 6.61s\tremaining: 348ms\n",
      "285:\tlearn: 0.2733286\ttotal: 6.63s\tremaining: 325ms\n",
      "286:\tlearn: 0.2729633\ttotal: 6.65s\tremaining: 301ms\n",
      "287:\tlearn: 0.2724323\ttotal: 6.67s\tremaining: 278ms\n",
      "288:\tlearn: 0.2721791\ttotal: 6.69s\tremaining: 255ms\n",
      "289:\tlearn: 0.2717790\ttotal: 6.71s\tremaining: 232ms\n",
      "290:\tlearn: 0.2713203\ttotal: 6.73s\tremaining: 208ms\n",
      "291:\tlearn: 0.2709690\ttotal: 6.75s\tremaining: 185ms\n",
      "292:\tlearn: 0.2704877\ttotal: 6.78s\tremaining: 162ms\n",
      "293:\tlearn: 0.2698658\ttotal: 6.8s\tremaining: 139ms\n",
      "294:\tlearn: 0.2693008\ttotal: 6.82s\tremaining: 116ms\n",
      "295:\tlearn: 0.2690063\ttotal: 6.83s\tremaining: 92.4ms\n",
      "296:\tlearn: 0.2686288\ttotal: 6.85s\tremaining: 69.2ms\n",
      "297:\tlearn: 0.2682407\ttotal: 6.87s\tremaining: 46.1ms\n",
      "298:\tlearn: 0.2675884\ttotal: 6.88s\tremaining: 23ms\n",
      "299:\tlearn: 0.2672923\ttotal: 6.9s\tremaining: 0us\n",
      "0:\tlearn: 1.2716733\ttotal: 15ms\tremaining: 4.48s\n",
      "1:\tlearn: 1.2410093\ttotal: 30.2ms\tremaining: 4.5s\n",
      "2:\tlearn: 1.2147488\ttotal: 45.2ms\tremaining: 4.48s\n",
      "3:\tlearn: 1.1848309\ttotal: 60.8ms\tremaining: 4.5s\n",
      "4:\tlearn: 1.1606342\ttotal: 76.1ms\tremaining: 4.49s\n",
      "5:\tlearn: 1.1376161\ttotal: 94.1ms\tremaining: 4.61s\n",
      "6:\tlearn: 1.1150122\ttotal: 111ms\tremaining: 4.64s\n",
      "7:\tlearn: 1.0980865\ttotal: 131ms\tremaining: 4.77s\n",
      "8:\tlearn: 1.0843103\ttotal: 150ms\tremaining: 4.83s\n",
      "9:\tlearn: 1.0692257\ttotal: 168ms\tremaining: 4.88s\n",
      "10:\tlearn: 1.0535986\ttotal: 188ms\tremaining: 4.93s\n",
      "11:\tlearn: 1.0376544\ttotal: 207ms\tremaining: 4.97s\n",
      "12:\tlearn: 1.0270013\ttotal: 227ms\tremaining: 5.02s\n",
      "13:\tlearn: 1.0172938\ttotal: 247ms\tremaining: 5.04s\n",
      "14:\tlearn: 1.0044287\ttotal: 266ms\tremaining: 5.05s\n",
      "15:\tlearn: 0.9908467\ttotal: 285ms\tremaining: 5.05s\n",
      "16:\tlearn: 0.9785829\ttotal: 305ms\tremaining: 5.08s\n",
      "17:\tlearn: 0.9706868\ttotal: 325ms\tremaining: 5.09s\n",
      "18:\tlearn: 0.9611772\ttotal: 345ms\tremaining: 5.11s\n",
      "19:\tlearn: 0.9527227\ttotal: 364ms\tremaining: 5.1s\n",
      "20:\tlearn: 0.9431204\ttotal: 384ms\tremaining: 5.09s\n",
      "21:\tlearn: 0.9374886\ttotal: 401ms\tremaining: 5.07s\n",
      "22:\tlearn: 0.9288238\ttotal: 419ms\tremaining: 5.04s\n",
      "23:\tlearn: 0.9210807\ttotal: 437ms\tremaining: 5.02s\n",
      "24:\tlearn: 0.9126490\ttotal: 455ms\tremaining: 5.01s\n",
      "25:\tlearn: 0.9031412\ttotal: 474ms\tremaining: 5s\n",
      "26:\tlearn: 0.8947925\ttotal: 491ms\tremaining: 4.97s\n",
      "27:\tlearn: 0.8893184\ttotal: 509ms\tremaining: 4.94s\n",
      "28:\tlearn: 0.8817980\ttotal: 525ms\tremaining: 4.91s\n",
      "29:\tlearn: 0.8768413\ttotal: 542ms\tremaining: 4.88s\n",
      "30:\tlearn: 0.8702700\ttotal: 559ms\tremaining: 4.85s\n",
      "31:\tlearn: 0.8655479\ttotal: 576ms\tremaining: 4.82s\n",
      "32:\tlearn: 0.8617074\ttotal: 593ms\tremaining: 4.8s\n",
      "33:\tlearn: 0.8531272\ttotal: 611ms\tremaining: 4.78s\n",
      "34:\tlearn: 0.8438615\ttotal: 628ms\tremaining: 4.76s\n",
      "35:\tlearn: 0.8379335\ttotal: 647ms\tremaining: 4.74s\n",
      "36:\tlearn: 0.8317166\ttotal: 665ms\tremaining: 4.73s\n",
      "37:\tlearn: 0.8247974\ttotal: 684ms\tremaining: 4.71s\n",
      "38:\tlearn: 0.8194211\ttotal: 700ms\tremaining: 4.69s\n",
      "39:\tlearn: 0.8168784\ttotal: 717ms\tremaining: 4.66s\n",
      "40:\tlearn: 0.8098259\ttotal: 734ms\tremaining: 4.63s\n",
      "41:\tlearn: 0.8031636\ttotal: 750ms\tremaining: 4.61s\n",
      "42:\tlearn: 0.7987634\ttotal: 767ms\tremaining: 4.58s\n",
      "43:\tlearn: 0.7940497\ttotal: 784ms\tremaining: 4.56s\n",
      "44:\tlearn: 0.7889450\ttotal: 801ms\tremaining: 4.54s\n",
      "45:\tlearn: 0.7863625\ttotal: 820ms\tremaining: 4.53s\n",
      "46:\tlearn: 0.7831962\ttotal: 845ms\tremaining: 4.55s\n",
      "47:\tlearn: 0.7802244\ttotal: 873ms\tremaining: 4.58s\n",
      "48:\tlearn: 0.7770239\ttotal: 900ms\tremaining: 4.61s\n",
      "49:\tlearn: 0.7729939\ttotal: 928ms\tremaining: 4.64s\n",
      "50:\tlearn: 0.7664801\ttotal: 955ms\tremaining: 4.66s\n",
      "51:\tlearn: 0.7623963\ttotal: 981ms\tremaining: 4.68s\n",
      "52:\tlearn: 0.7590460\ttotal: 1.01s\tremaining: 4.7s\n",
      "53:\tlearn: 0.7565750\ttotal: 1.03s\tremaining: 4.71s\n",
      "54:\tlearn: 0.7550371\ttotal: 1.06s\tremaining: 4.73s\n",
      "55:\tlearn: 0.7498586\ttotal: 1.09s\tremaining: 4.75s\n",
      "56:\tlearn: 0.7452792\ttotal: 1.12s\tremaining: 4.78s\n",
      "57:\tlearn: 0.7397208\ttotal: 1.15s\tremaining: 4.79s\n",
      "58:\tlearn: 0.7364394\ttotal: 1.17s\tremaining: 4.8s\n",
      "59:\tlearn: 0.7323213\ttotal: 1.21s\tremaining: 4.83s\n",
      "60:\tlearn: 0.7296014\ttotal: 1.23s\tremaining: 4.83s\n",
      "61:\tlearn: 0.7259497\ttotal: 1.26s\tremaining: 4.84s\n",
      "62:\tlearn: 0.7223159\ttotal: 1.29s\tremaining: 4.85s\n",
      "63:\tlearn: 0.7189823\ttotal: 1.32s\tremaining: 4.86s\n",
      "64:\tlearn: 0.7151184\ttotal: 1.34s\tremaining: 4.86s\n",
      "65:\tlearn: 0.7123385\ttotal: 1.37s\tremaining: 4.87s\n",
      "66:\tlearn: 0.7080324\ttotal: 1.4s\tremaining: 4.87s\n",
      "67:\tlearn: 0.7049729\ttotal: 1.43s\tremaining: 4.87s\n",
      "68:\tlearn: 0.7033115\ttotal: 1.45s\tremaining: 4.87s\n",
      "69:\tlearn: 0.7027406\ttotal: 1.46s\tremaining: 4.79s\n",
      "70:\tlearn: 0.6997366\ttotal: 1.48s\tremaining: 4.79s\n",
      "71:\tlearn: 0.6951989\ttotal: 1.51s\tremaining: 4.79s\n",
      "72:\tlearn: 0.6924279\ttotal: 1.54s\tremaining: 4.79s\n",
      "73:\tlearn: 0.6893967\ttotal: 1.57s\tremaining: 4.79s\n",
      "74:\tlearn: 0.6873748\ttotal: 1.59s\tremaining: 4.77s\n",
      "75:\tlearn: 0.6854350\ttotal: 1.61s\tremaining: 4.74s\n",
      "76:\tlearn: 0.6828649\ttotal: 1.63s\tremaining: 4.71s\n",
      "77:\tlearn: 0.6794728\ttotal: 1.64s\tremaining: 4.68s\n",
      "78:\tlearn: 0.6777862\ttotal: 1.66s\tremaining: 4.64s\n",
      "79:\tlearn: 0.6746937\ttotal: 1.67s\tremaining: 4.6s\n",
      "80:\tlearn: 0.6713247\ttotal: 1.69s\tremaining: 4.57s\n",
      "81:\tlearn: 0.6701479\ttotal: 1.7s\tremaining: 4.53s\n",
      "82:\tlearn: 0.6677038\ttotal: 1.72s\tremaining: 4.5s\n",
      "83:\tlearn: 0.6651931\ttotal: 1.74s\tremaining: 4.47s\n",
      "84:\tlearn: 0.6626857\ttotal: 1.76s\tremaining: 4.45s\n",
      "85:\tlearn: 0.6625677\ttotal: 1.76s\tremaining: 4.38s\n",
      "86:\tlearn: 0.6606673\ttotal: 1.78s\tremaining: 4.35s\n",
      "87:\tlearn: 0.6596336\ttotal: 1.79s\tremaining: 4.33s\n",
      "88:\tlearn: 0.6588569\ttotal: 1.8s\tremaining: 4.26s\n",
      "89:\tlearn: 0.6585857\ttotal: 1.8s\tremaining: 4.2s\n",
      "90:\tlearn: 0.6569656\ttotal: 1.81s\tremaining: 4.16s\n",
      "91:\tlearn: 0.6557074\ttotal: 1.83s\tremaining: 4.14s\n",
      "92:\tlearn: 0.6554914\ttotal: 1.83s\tremaining: 4.08s\n",
      "93:\tlearn: 0.6533227\ttotal: 1.85s\tremaining: 4.06s\n",
      "94:\tlearn: 0.6505941\ttotal: 1.87s\tremaining: 4.03s\n",
      "95:\tlearn: 0.6480271\ttotal: 1.9s\tremaining: 4.05s\n",
      "96:\tlearn: 0.6469751\ttotal: 1.93s\tremaining: 4.05s\n",
      "97:\tlearn: 0.6461141\ttotal: 1.94s\tremaining: 4s\n",
      "98:\tlearn: 0.6446595\ttotal: 1.96s\tremaining: 3.98s\n",
      "99:\tlearn: 0.6426704\ttotal: 1.97s\tremaining: 3.94s\n",
      "100:\tlearn: 0.6397374\ttotal: 1.99s\tremaining: 3.92s\n",
      "101:\tlearn: 0.6388510\ttotal: 2.01s\tremaining: 3.9s\n",
      "102:\tlearn: 0.6378057\ttotal: 2.03s\tremaining: 3.89s\n",
      "103:\tlearn: 0.6355011\ttotal: 2.06s\tremaining: 3.87s\n",
      "104:\tlearn: 0.6346428\ttotal: 2.08s\tremaining: 3.85s\n",
      "105:\tlearn: 0.6324607\ttotal: 2.1s\tremaining: 3.83s\n",
      "106:\tlearn: 0.6307602\ttotal: 2.11s\tremaining: 3.81s\n",
      "107:\tlearn: 0.6282415\ttotal: 2.13s\tremaining: 3.79s\n",
      "108:\tlearn: 0.6261082\ttotal: 2.16s\tremaining: 3.78s\n",
      "109:\tlearn: 0.6230551\ttotal: 2.18s\tremaining: 3.77s\n",
      "110:\tlearn: 0.6209138\ttotal: 2.21s\tremaining: 3.76s\n",
      "111:\tlearn: 0.6174260\ttotal: 2.24s\tremaining: 3.76s\n",
      "112:\tlearn: 0.6149122\ttotal: 2.27s\tremaining: 3.75s\n",
      "113:\tlearn: 0.6136733\ttotal: 2.29s\tremaining: 3.75s\n",
      "114:\tlearn: 0.6122829\ttotal: 2.32s\tremaining: 3.74s\n",
      "115:\tlearn: 0.6099511\ttotal: 2.35s\tremaining: 3.73s\n",
      "116:\tlearn: 0.6069926\ttotal: 2.38s\tremaining: 3.72s\n",
      "117:\tlearn: 0.6058601\ttotal: 2.41s\tremaining: 3.71s\n",
      "118:\tlearn: 0.6039671\ttotal: 2.44s\tremaining: 3.7s\n",
      "119:\tlearn: 0.6007293\ttotal: 2.46s\tremaining: 3.69s\n",
      "120:\tlearn: 0.5988075\ttotal: 2.49s\tremaining: 3.68s\n",
      "121:\tlearn: 0.5959783\ttotal: 2.52s\tremaining: 3.67s\n",
      "122:\tlearn: 0.5939403\ttotal: 2.54s\tremaining: 3.66s\n",
      "123:\tlearn: 0.5931743\ttotal: 2.57s\tremaining: 3.65s\n",
      "124:\tlearn: 0.5913174\ttotal: 2.6s\tremaining: 3.64s\n",
      "125:\tlearn: 0.5897431\ttotal: 2.63s\tremaining: 3.63s\n",
      "126:\tlearn: 0.5854938\ttotal: 2.66s\tremaining: 3.62s\n",
      "127:\tlearn: 0.5840750\ttotal: 2.69s\tremaining: 3.61s\n",
      "128:\tlearn: 0.5817819\ttotal: 2.71s\tremaining: 3.6s\n",
      "129:\tlearn: 0.5791307\ttotal: 2.74s\tremaining: 3.58s\n",
      "130:\tlearn: 0.5773855\ttotal: 2.77s\tremaining: 3.57s\n",
      "131:\tlearn: 0.5744268\ttotal: 2.79s\tremaining: 3.56s\n",
      "132:\tlearn: 0.5717606\ttotal: 2.82s\tremaining: 3.54s\n",
      "133:\tlearn: 0.5692147\ttotal: 2.85s\tremaining: 3.53s\n",
      "134:\tlearn: 0.5671857\ttotal: 2.88s\tremaining: 3.51s\n",
      "135:\tlearn: 0.5650819\ttotal: 2.9s\tremaining: 3.5s\n",
      "136:\tlearn: 0.5637813\ttotal: 2.93s\tremaining: 3.49s\n",
      "137:\tlearn: 0.5616308\ttotal: 2.96s\tremaining: 3.47s\n",
      "138:\tlearn: 0.5603332\ttotal: 2.99s\tremaining: 3.46s\n",
      "139:\tlearn: 0.5594397\ttotal: 3s\tremaining: 3.43s\n",
      "140:\tlearn: 0.5575417\ttotal: 3.03s\tremaining: 3.42s\n",
      "141:\tlearn: 0.5556111\ttotal: 3.06s\tremaining: 3.4s\n",
      "142:\tlearn: 0.5549449\ttotal: 3.09s\tremaining: 3.39s\n",
      "143:\tlearn: 0.5536292\ttotal: 3.11s\tremaining: 3.37s\n",
      "144:\tlearn: 0.5514548\ttotal: 3.14s\tremaining: 3.36s\n",
      "145:\tlearn: 0.5504784\ttotal: 3.18s\tremaining: 3.36s\n",
      "146:\tlearn: 0.5495914\ttotal: 3.23s\tremaining: 3.36s\n",
      "147:\tlearn: 0.5478397\ttotal: 3.28s\tremaining: 3.37s\n",
      "148:\tlearn: 0.5458961\ttotal: 3.32s\tremaining: 3.37s\n",
      "149:\tlearn: 0.5443932\ttotal: 3.37s\tremaining: 3.37s\n",
      "150:\tlearn: 0.5437248\ttotal: 3.42s\tremaining: 3.37s\n",
      "151:\tlearn: 0.5416697\ttotal: 3.46s\tremaining: 3.37s\n",
      "152:\tlearn: 0.5403799\ttotal: 3.51s\tremaining: 3.37s\n",
      "153:\tlearn: 0.5391929\ttotal: 3.56s\tremaining: 3.37s\n",
      "154:\tlearn: 0.5365854\ttotal: 3.6s\tremaining: 3.37s\n",
      "155:\tlearn: 0.5342608\ttotal: 3.65s\tremaining: 3.37s\n",
      "156:\tlearn: 0.5329512\ttotal: 3.69s\tremaining: 3.36s\n",
      "157:\tlearn: 0.5313104\ttotal: 3.7s\tremaining: 3.33s\n",
      "158:\tlearn: 0.5294627\ttotal: 3.72s\tremaining: 3.3s\n",
      "159:\tlearn: 0.5278591\ttotal: 3.74s\tremaining: 3.27s\n",
      "160:\tlearn: 0.5258985\ttotal: 3.75s\tremaining: 3.24s\n",
      "161:\tlearn: 0.5246683\ttotal: 3.77s\tremaining: 3.21s\n",
      "162:\tlearn: 0.5227652\ttotal: 3.79s\tremaining: 3.18s\n",
      "163:\tlearn: 0.5212695\ttotal: 3.8s\tremaining: 3.15s\n",
      "164:\tlearn: 0.5195954\ttotal: 3.82s\tremaining: 3.13s\n",
      "165:\tlearn: 0.5174194\ttotal: 3.84s\tremaining: 3.1s\n",
      "166:\tlearn: 0.5165119\ttotal: 3.86s\tremaining: 3.07s\n",
      "167:\tlearn: 0.5159141\ttotal: 3.88s\tremaining: 3.04s\n",
      "168:\tlearn: 0.5142798\ttotal: 3.9s\tremaining: 3.02s\n",
      "169:\tlearn: 0.5134133\ttotal: 3.92s\tremaining: 2.99s\n",
      "170:\tlearn: 0.5119878\ttotal: 3.94s\tremaining: 2.97s\n",
      "171:\tlearn: 0.5111829\ttotal: 3.97s\tremaining: 2.95s\n",
      "172:\tlearn: 0.5092819\ttotal: 3.99s\tremaining: 2.93s\n",
      "173:\tlearn: 0.5078459\ttotal: 4.02s\tremaining: 2.91s\n",
      "174:\tlearn: 0.5068259\ttotal: 4.05s\tremaining: 2.89s\n",
      "175:\tlearn: 0.5052893\ttotal: 4.08s\tremaining: 2.87s\n",
      "176:\tlearn: 0.5040140\ttotal: 4.11s\tremaining: 2.85s\n",
      "177:\tlearn: 0.5029975\ttotal: 4.13s\tremaining: 2.83s\n",
      "178:\tlearn: 0.5019407\ttotal: 4.16s\tremaining: 2.81s\n",
      "179:\tlearn: 0.5003895\ttotal: 4.19s\tremaining: 2.79s\n",
      "180:\tlearn: 0.5000524\ttotal: 4.22s\tremaining: 2.77s\n",
      "181:\tlearn: 0.4988829\ttotal: 4.24s\tremaining: 2.75s\n",
      "182:\tlearn: 0.4972757\ttotal: 4.27s\tremaining: 2.73s\n",
      "183:\tlearn: 0.4963687\ttotal: 4.3s\tremaining: 2.71s\n",
      "184:\tlearn: 0.4960904\ttotal: 4.33s\tremaining: 2.69s\n",
      "185:\tlearn: 0.4942732\ttotal: 4.35s\tremaining: 2.67s\n",
      "186:\tlearn: 0.4934366\ttotal: 4.38s\tremaining: 2.65s\n",
      "187:\tlearn: 0.4926491\ttotal: 4.41s\tremaining: 2.63s\n",
      "188:\tlearn: 0.4917773\ttotal: 4.44s\tremaining: 2.6s\n",
      "189:\tlearn: 0.4909371\ttotal: 4.46s\tremaining: 2.58s\n",
      "190:\tlearn: 0.4903802\ttotal: 4.49s\tremaining: 2.56s\n",
      "191:\tlearn: 0.4892378\ttotal: 4.52s\tremaining: 2.54s\n",
      "192:\tlearn: 0.4885299\ttotal: 4.55s\tremaining: 2.52s\n",
      "193:\tlearn: 0.4877387\ttotal: 4.58s\tremaining: 2.5s\n",
      "194:\tlearn: 0.4871193\ttotal: 4.6s\tremaining: 2.48s\n",
      "195:\tlearn: 0.4855954\ttotal: 4.63s\tremaining: 2.46s\n",
      "196:\tlearn: 0.4837057\ttotal: 4.66s\tremaining: 2.44s\n",
      "197:\tlearn: 0.4825941\ttotal: 4.68s\tremaining: 2.41s\n",
      "198:\tlearn: 0.4812569\ttotal: 4.71s\tremaining: 2.39s\n",
      "199:\tlearn: 0.4806191\ttotal: 4.74s\tremaining: 2.37s\n",
      "200:\tlearn: 0.4800712\ttotal: 4.76s\tremaining: 2.35s\n",
      "201:\tlearn: 0.4784668\ttotal: 4.79s\tremaining: 2.32s\n",
      "202:\tlearn: 0.4777960\ttotal: 4.82s\tremaining: 2.3s\n",
      "203:\tlearn: 0.4766551\ttotal: 4.84s\tremaining: 2.28s\n",
      "204:\tlearn: 0.4757117\ttotal: 4.87s\tremaining: 2.26s\n",
      "205:\tlearn: 0.4742840\ttotal: 4.9s\tremaining: 2.24s\n",
      "206:\tlearn: 0.4739188\ttotal: 4.93s\tremaining: 2.21s\n",
      "207:\tlearn: 0.4734173\ttotal: 4.96s\tremaining: 2.19s\n",
      "208:\tlearn: 0.4725025\ttotal: 4.98s\tremaining: 2.17s\n",
      "209:\tlearn: 0.4717951\ttotal: 5.01s\tremaining: 2.15s\n",
      "210:\tlearn: 0.4710548\ttotal: 5.04s\tremaining: 2.13s\n",
      "211:\tlearn: 0.4696235\ttotal: 5.07s\tremaining: 2.1s\n",
      "212:\tlearn: 0.4682614\ttotal: 5.09s\tremaining: 2.08s\n",
      "213:\tlearn: 0.4667118\ttotal: 5.12s\tremaining: 2.06s\n",
      "214:\tlearn: 0.4652851\ttotal: 5.15s\tremaining: 2.04s\n",
      "215:\tlearn: 0.4643429\ttotal: 5.18s\tremaining: 2.01s\n",
      "216:\tlearn: 0.4632689\ttotal: 5.2s\tremaining: 1.99s\n",
      "217:\tlearn: 0.4623682\ttotal: 5.23s\tremaining: 1.97s\n",
      "218:\tlearn: 0.4616871\ttotal: 5.26s\tremaining: 1.95s\n",
      "219:\tlearn: 0.4606171\ttotal: 5.29s\tremaining: 1.92s\n",
      "220:\tlearn: 0.4600212\ttotal: 5.31s\tremaining: 1.9s\n",
      "221:\tlearn: 0.4591722\ttotal: 5.33s\tremaining: 1.87s\n",
      "222:\tlearn: 0.4588371\ttotal: 5.35s\tremaining: 1.85s\n",
      "223:\tlearn: 0.4571441\ttotal: 5.37s\tremaining: 1.82s\n",
      "224:\tlearn: 0.4557445\ttotal: 5.38s\tremaining: 1.79s\n",
      "225:\tlearn: 0.4542682\ttotal: 5.4s\tremaining: 1.77s\n",
      "226:\tlearn: 0.4535410\ttotal: 5.42s\tremaining: 1.74s\n",
      "227:\tlearn: 0.4524622\ttotal: 5.43s\tremaining: 1.72s\n",
      "228:\tlearn: 0.4514538\ttotal: 5.45s\tremaining: 1.69s\n",
      "229:\tlearn: 0.4508005\ttotal: 5.47s\tremaining: 1.66s\n",
      "230:\tlearn: 0.4502926\ttotal: 5.49s\tremaining: 1.64s\n",
      "231:\tlearn: 0.4495453\ttotal: 5.5s\tremaining: 1.61s\n",
      "232:\tlearn: 0.4489633\ttotal: 5.52s\tremaining: 1.59s\n",
      "233:\tlearn: 0.4478684\ttotal: 5.54s\tremaining: 1.56s\n",
      "234:\tlearn: 0.4474861\ttotal: 5.56s\tremaining: 1.54s\n",
      "235:\tlearn: 0.4465008\ttotal: 5.59s\tremaining: 1.51s\n",
      "236:\tlearn: 0.4457375\ttotal: 5.61s\tremaining: 1.49s\n",
      "237:\tlearn: 0.4453715\ttotal: 5.62s\tremaining: 1.47s\n",
      "238:\tlearn: 0.4444897\ttotal: 5.64s\tremaining: 1.44s\n",
      "239:\tlearn: 0.4438954\ttotal: 5.66s\tremaining: 1.42s\n",
      "240:\tlearn: 0.4428634\ttotal: 5.68s\tremaining: 1.39s\n",
      "241:\tlearn: 0.4420349\ttotal: 5.7s\tremaining: 1.37s\n",
      "242:\tlearn: 0.4410712\ttotal: 5.72s\tremaining: 1.34s\n",
      "243:\tlearn: 0.4397501\ttotal: 5.74s\tremaining: 1.32s\n",
      "244:\tlearn: 0.4392573\ttotal: 5.76s\tremaining: 1.29s\n",
      "245:\tlearn: 0.4386529\ttotal: 5.78s\tremaining: 1.27s\n",
      "246:\tlearn: 0.4377137\ttotal: 5.79s\tremaining: 1.24s\n",
      "247:\tlearn: 0.4366597\ttotal: 5.82s\tremaining: 1.22s\n",
      "248:\tlearn: 0.4358801\ttotal: 5.84s\tremaining: 1.2s\n",
      "249:\tlearn: 0.4350351\ttotal: 5.86s\tremaining: 1.17s\n",
      "250:\tlearn: 0.4341439\ttotal: 5.88s\tremaining: 1.15s\n",
      "251:\tlearn: 0.4337722\ttotal: 5.91s\tremaining: 1.12s\n",
      "252:\tlearn: 0.4324101\ttotal: 5.93s\tremaining: 1.1s\n",
      "253:\tlearn: 0.4311022\ttotal: 5.95s\tremaining: 1.08s\n",
      "254:\tlearn: 0.4303978\ttotal: 5.98s\tremaining: 1.05s\n",
      "255:\tlearn: 0.4294163\ttotal: 6s\tremaining: 1.03s\n",
      "256:\tlearn: 0.4288670\ttotal: 6.03s\tremaining: 1.01s\n",
      "257:\tlearn: 0.4286597\ttotal: 6.06s\tremaining: 987ms\n",
      "258:\tlearn: 0.4274646\ttotal: 6.09s\tremaining: 964ms\n",
      "259:\tlearn: 0.4269329\ttotal: 6.12s\tremaining: 941ms\n",
      "260:\tlearn: 0.4262040\ttotal: 6.15s\tremaining: 918ms\n",
      "261:\tlearn: 0.4249686\ttotal: 6.17s\tremaining: 895ms\n",
      "262:\tlearn: 0.4240198\ttotal: 6.2s\tremaining: 872ms\n",
      "263:\tlearn: 0.4234071\ttotal: 6.23s\tremaining: 849ms\n",
      "264:\tlearn: 0.4226018\ttotal: 6.25s\tremaining: 826ms\n",
      "265:\tlearn: 0.4216316\ttotal: 6.28s\tremaining: 803ms\n",
      "266:\tlearn: 0.4207519\ttotal: 6.31s\tremaining: 780ms\n",
      "267:\tlearn: 0.4199660\ttotal: 6.34s\tremaining: 757ms\n",
      "268:\tlearn: 0.4195807\ttotal: 6.37s\tremaining: 734ms\n",
      "269:\tlearn: 0.4184653\ttotal: 6.39s\tremaining: 710ms\n",
      "270:\tlearn: 0.4175683\ttotal: 6.42s\tremaining: 687ms\n",
      "271:\tlearn: 0.4171281\ttotal: 6.45s\tremaining: 664ms\n",
      "272:\tlearn: 0.4162671\ttotal: 6.48s\tremaining: 641ms\n",
      "273:\tlearn: 0.4156868\ttotal: 6.5s\tremaining: 617ms\n",
      "274:\tlearn: 0.4152749\ttotal: 6.54s\tremaining: 594ms\n",
      "275:\tlearn: 0.4150028\ttotal: 6.57s\tremaining: 571ms\n",
      "276:\tlearn: 0.4141032\ttotal: 6.59s\tremaining: 547ms\n",
      "277:\tlearn: 0.4134032\ttotal: 6.62s\tremaining: 524ms\n",
      "278:\tlearn: 0.4129248\ttotal: 6.63s\tremaining: 499ms\n",
      "279:\tlearn: 0.4119810\ttotal: 6.65s\tremaining: 475ms\n",
      "280:\tlearn: 0.4115535\ttotal: 6.66s\tremaining: 451ms\n",
      "281:\tlearn: 0.4109756\ttotal: 6.68s\tremaining: 426ms\n",
      "282:\tlearn: 0.4105487\ttotal: 6.69s\tremaining: 402ms\n",
      "283:\tlearn: 0.4099223\ttotal: 6.71s\tremaining: 378ms\n",
      "284:\tlearn: 0.4090377\ttotal: 6.72s\tremaining: 354ms\n",
      "285:\tlearn: 0.4084476\ttotal: 6.74s\tremaining: 330ms\n",
      "286:\tlearn: 0.4076974\ttotal: 6.76s\tremaining: 306ms\n",
      "287:\tlearn: 0.4068690\ttotal: 6.78s\tremaining: 282ms\n",
      "288:\tlearn: 0.4060669\ttotal: 6.79s\tremaining: 259ms\n",
      "289:\tlearn: 0.4055336\ttotal: 6.81s\tremaining: 235ms\n",
      "290:\tlearn: 0.4046186\ttotal: 6.82s\tremaining: 211ms\n",
      "291:\tlearn: 0.4042265\ttotal: 6.84s\tremaining: 187ms\n",
      "292:\tlearn: 0.4037020\ttotal: 6.86s\tremaining: 164ms\n",
      "293:\tlearn: 0.4028816\ttotal: 6.87s\tremaining: 140ms\n",
      "294:\tlearn: 0.4026049\ttotal: 6.89s\tremaining: 117ms\n",
      "295:\tlearn: 0.4018111\ttotal: 6.9s\tremaining: 93.3ms\n",
      "296:\tlearn: 0.4014364\ttotal: 6.92s\tremaining: 69.9ms\n",
      "297:\tlearn: 0.4005799\ttotal: 6.94s\tremaining: 46.6ms\n",
      "298:\tlearn: 0.3999280\ttotal: 6.96s\tremaining: 23.3ms\n",
      "299:\tlearn: 0.3993396\ttotal: 6.98s\tremaining: 0us\n",
      "0:\tlearn: 1.2257700\ttotal: 15.3ms\tremaining: 4.57s\n",
      "1:\tlearn: 1.1818465\ttotal: 32.5ms\tremaining: 4.84s\n",
      "2:\tlearn: 1.1366061\ttotal: 50.4ms\tremaining: 4.99s\n",
      "3:\tlearn: 1.0903609\ttotal: 69ms\tremaining: 5.11s\n",
      "4:\tlearn: 1.0545867\ttotal: 86.1ms\tremaining: 5.08s\n",
      "5:\tlearn: 1.0214132\ttotal: 102ms\tremaining: 5.02s\n",
      "6:\tlearn: 0.9926711\ttotal: 121ms\tremaining: 5.06s\n",
      "7:\tlearn: 0.9830558\ttotal: 123ms\tremaining: 4.47s\n",
      "8:\tlearn: 0.9584544\ttotal: 143ms\tremaining: 4.62s\n",
      "9:\tlearn: 0.9286442\ttotal: 165ms\tremaining: 4.79s\n",
      "10:\tlearn: 0.9051199\ttotal: 189ms\tremaining: 4.97s\n",
      "11:\tlearn: 0.8877729\ttotal: 214ms\tremaining: 5.14s\n",
      "12:\tlearn: 0.8681942\ttotal: 241ms\tremaining: 5.33s\n",
      "13:\tlearn: 0.8466564\ttotal: 268ms\tremaining: 5.48s\n",
      "14:\tlearn: 0.8312373\ttotal: 295ms\tremaining: 5.6s\n",
      "15:\tlearn: 0.8151748\ttotal: 310ms\tremaining: 5.5s\n",
      "16:\tlearn: 0.8034440\ttotal: 327ms\tremaining: 5.44s\n",
      "17:\tlearn: 0.7857559\ttotal: 343ms\tremaining: 5.38s\n",
      "18:\tlearn: 0.7730143\ttotal: 360ms\tremaining: 5.33s\n",
      "19:\tlearn: 0.7623655\ttotal: 378ms\tremaining: 5.29s\n",
      "20:\tlearn: 0.7490928\ttotal: 396ms\tremaining: 5.25s\n",
      "21:\tlearn: 0.7390769\ttotal: 413ms\tremaining: 5.22s\n",
      "22:\tlearn: 0.7280904\ttotal: 432ms\tremaining: 5.2s\n",
      "23:\tlearn: 0.7197305\ttotal: 449ms\tremaining: 5.16s\n",
      "24:\tlearn: 0.7147100\ttotal: 467ms\tremaining: 5.13s\n",
      "25:\tlearn: 0.7051670\ttotal: 486ms\tremaining: 5.12s\n",
      "26:\tlearn: 0.6951725\ttotal: 504ms\tremaining: 5.1s\n",
      "27:\tlearn: 0.6871575\ttotal: 523ms\tremaining: 5.08s\n",
      "28:\tlearn: 0.6775026\ttotal: 542ms\tremaining: 5.07s\n",
      "29:\tlearn: 0.6690867\ttotal: 562ms\tremaining: 5.05s\n",
      "30:\tlearn: 0.6629808\ttotal: 584ms\tremaining: 5.07s\n",
      "31:\tlearn: 0.6560104\ttotal: 613ms\tremaining: 5.13s\n",
      "32:\tlearn: 0.6495602\ttotal: 644ms\tremaining: 5.21s\n",
      "33:\tlearn: 0.6417743\ttotal: 674ms\tremaining: 5.28s\n",
      "34:\tlearn: 0.6374108\ttotal: 704ms\tremaining: 5.33s\n",
      "35:\tlearn: 0.6324756\ttotal: 734ms\tremaining: 5.38s\n",
      "36:\tlearn: 0.6265660\ttotal: 763ms\tremaining: 5.42s\n",
      "37:\tlearn: 0.6215047\ttotal: 792ms\tremaining: 5.46s\n",
      "38:\tlearn: 0.6162842\ttotal: 820ms\tremaining: 5.49s\n",
      "39:\tlearn: 0.6112501\ttotal: 852ms\tremaining: 5.54s\n",
      "40:\tlearn: 0.6063102\ttotal: 882ms\tremaining: 5.57s\n",
      "41:\tlearn: 0.6023954\ttotal: 911ms\tremaining: 5.6s\n",
      "42:\tlearn: 0.5986239\ttotal: 941ms\tremaining: 5.62s\n",
      "43:\tlearn: 0.5940628\ttotal: 971ms\tremaining: 5.65s\n",
      "44:\tlearn: 0.5889279\ttotal: 1s\tremaining: 5.67s\n",
      "45:\tlearn: 0.5862471\ttotal: 1.03s\tremaining: 5.68s\n",
      "46:\tlearn: 0.5815208\ttotal: 1.06s\tremaining: 5.7s\n",
      "47:\tlearn: 0.5764766\ttotal: 1.09s\tremaining: 5.72s\n",
      "48:\tlearn: 0.5714066\ttotal: 1.12s\tremaining: 5.73s\n",
      "49:\tlearn: 0.5682893\ttotal: 1.15s\tremaining: 5.74s\n",
      "50:\tlearn: 0.5655104\ttotal: 1.18s\tremaining: 5.74s\n",
      "51:\tlearn: 0.5612795\ttotal: 1.2s\tremaining: 5.74s\n",
      "52:\tlearn: 0.5573486\ttotal: 1.23s\tremaining: 5.75s\n",
      "53:\tlearn: 0.5538202\ttotal: 1.26s\tremaining: 5.75s\n",
      "54:\tlearn: 0.5497826\ttotal: 1.29s\tremaining: 5.75s\n",
      "55:\tlearn: 0.5463401\ttotal: 1.32s\tremaining: 5.75s\n",
      "56:\tlearn: 0.5426100\ttotal: 1.35s\tremaining: 5.76s\n",
      "57:\tlearn: 0.5394160\ttotal: 1.38s\tremaining: 5.76s\n",
      "58:\tlearn: 0.5365221\ttotal: 1.41s\tremaining: 5.76s\n",
      "59:\tlearn: 0.5319332\ttotal: 1.44s\tremaining: 5.76s\n",
      "60:\tlearn: 0.5295666\ttotal: 1.47s\tremaining: 5.76s\n",
      "61:\tlearn: 0.5257020\ttotal: 1.5s\tremaining: 5.75s\n",
      "62:\tlearn: 0.5225211\ttotal: 1.53s\tremaining: 5.75s\n",
      "63:\tlearn: 0.5192729\ttotal: 1.55s\tremaining: 5.7s\n",
      "64:\tlearn: 0.5168883\ttotal: 1.56s\tremaining: 5.66s\n",
      "65:\tlearn: 0.5136674\ttotal: 1.58s\tremaining: 5.61s\n",
      "66:\tlearn: 0.5103213\ttotal: 1.6s\tremaining: 5.56s\n",
      "67:\tlearn: 0.5087326\ttotal: 1.61s\tremaining: 5.51s\n",
      "68:\tlearn: 0.5071617\ttotal: 1.63s\tremaining: 5.47s\n",
      "69:\tlearn: 0.5048834\ttotal: 1.65s\tremaining: 5.42s\n",
      "70:\tlearn: 0.5025594\ttotal: 1.67s\tremaining: 5.38s\n",
      "71:\tlearn: 0.4988714\ttotal: 1.68s\tremaining: 5.33s\n",
      "72:\tlearn: 0.4960450\ttotal: 1.7s\tremaining: 5.29s\n",
      "73:\tlearn: 0.4927273\ttotal: 1.72s\tremaining: 5.25s\n",
      "74:\tlearn: 0.4900551\ttotal: 1.74s\tremaining: 5.21s\n",
      "75:\tlearn: 0.4866641\ttotal: 1.76s\tremaining: 5.18s\n",
      "76:\tlearn: 0.4833585\ttotal: 1.78s\tremaining: 5.15s\n",
      "77:\tlearn: 0.4804517\ttotal: 1.8s\tremaining: 5.13s\n",
      "78:\tlearn: 0.4779465\ttotal: 1.82s\tremaining: 5.1s\n",
      "79:\tlearn: 0.4753937\ttotal: 1.84s\tremaining: 5.07s\n",
      "80:\tlearn: 0.4731388\ttotal: 1.87s\tremaining: 5.05s\n",
      "81:\tlearn: 0.4700917\ttotal: 1.89s\tremaining: 5.02s\n",
      "82:\tlearn: 0.4673372\ttotal: 1.91s\tremaining: 5s\n",
      "83:\tlearn: 0.4655339\ttotal: 1.94s\tremaining: 4.99s\n",
      "84:\tlearn: 0.4636682\ttotal: 1.97s\tremaining: 4.97s\n",
      "85:\tlearn: 0.4621383\ttotal: 1.99s\tremaining: 4.96s\n",
      "86:\tlearn: 0.4601667\ttotal: 2.02s\tremaining: 4.95s\n",
      "87:\tlearn: 0.4589240\ttotal: 2.05s\tremaining: 4.93s\n",
      "88:\tlearn: 0.4570993\ttotal: 2.08s\tremaining: 4.92s\n",
      "89:\tlearn: 0.4536825\ttotal: 2.1s\tremaining: 4.91s\n",
      "90:\tlearn: 0.4518176\ttotal: 2.13s\tremaining: 4.88s\n",
      "91:\tlearn: 0.4504914\ttotal: 2.15s\tremaining: 4.86s\n",
      "92:\tlearn: 0.4487218\ttotal: 2.17s\tremaining: 4.83s\n",
      "93:\tlearn: 0.4469995\ttotal: 2.19s\tremaining: 4.8s\n",
      "94:\tlearn: 0.4451481\ttotal: 2.21s\tremaining: 4.77s\n",
      "95:\tlearn: 0.4437565\ttotal: 2.23s\tremaining: 4.74s\n",
      "96:\tlearn: 0.4417733\ttotal: 2.25s\tremaining: 4.72s\n",
      "97:\tlearn: 0.4405008\ttotal: 2.28s\tremaining: 4.71s\n",
      "98:\tlearn: 0.4388370\ttotal: 2.31s\tremaining: 4.7s\n",
      "99:\tlearn: 0.4367408\ttotal: 2.35s\tremaining: 4.69s\n",
      "100:\tlearn: 0.4337635\ttotal: 2.38s\tremaining: 4.69s\n",
      "101:\tlearn: 0.4313024\ttotal: 2.41s\tremaining: 4.68s\n",
      "102:\tlearn: 0.4295943\ttotal: 2.44s\tremaining: 4.67s\n",
      "103:\tlearn: 0.4283154\ttotal: 2.48s\tremaining: 4.67s\n",
      "104:\tlearn: 0.4259669\ttotal: 2.51s\tremaining: 4.66s\n",
      "105:\tlearn: 0.4239722\ttotal: 2.54s\tremaining: 4.65s\n",
      "106:\tlearn: 0.4225989\ttotal: 2.58s\tremaining: 4.64s\n",
      "107:\tlearn: 0.4216207\ttotal: 2.61s\tremaining: 4.63s\n",
      "108:\tlearn: 0.4192525\ttotal: 2.64s\tremaining: 4.63s\n",
      "109:\tlearn: 0.4176424\ttotal: 2.67s\tremaining: 4.62s\n",
      "110:\tlearn: 0.4161521\ttotal: 2.71s\tremaining: 4.61s\n",
      "111:\tlearn: 0.4147906\ttotal: 2.74s\tremaining: 4.6s\n",
      "112:\tlearn: 0.4126549\ttotal: 2.77s\tremaining: 4.59s\n",
      "113:\tlearn: 0.4114165\ttotal: 2.8s\tremaining: 4.58s\n",
      "114:\tlearn: 0.4103243\ttotal: 2.84s\tremaining: 4.57s\n",
      "115:\tlearn: 0.4083185\ttotal: 2.87s\tremaining: 4.55s\n",
      "116:\tlearn: 0.4066662\ttotal: 2.9s\tremaining: 4.54s\n",
      "117:\tlearn: 0.4050372\ttotal: 2.93s\tremaining: 4.53s\n",
      "118:\tlearn: 0.4039171\ttotal: 2.97s\tremaining: 4.51s\n",
      "119:\tlearn: 0.4031624\ttotal: 3s\tremaining: 4.5s\n",
      "120:\tlearn: 0.4017373\ttotal: 3.03s\tremaining: 4.49s\n",
      "121:\tlearn: 0.4005056\ttotal: 3.07s\tremaining: 4.47s\n",
      "122:\tlearn: 0.3996263\ttotal: 3.1s\tremaining: 4.46s\n",
      "123:\tlearn: 0.3983090\ttotal: 3.13s\tremaining: 4.44s\n",
      "124:\tlearn: 0.3973643\ttotal: 3.13s\tremaining: 4.39s\n",
      "125:\tlearn: 0.3950691\ttotal: 3.15s\tremaining: 4.35s\n",
      "126:\tlearn: 0.3944159\ttotal: 3.17s\tremaining: 4.32s\n",
      "127:\tlearn: 0.3926229\ttotal: 3.19s\tremaining: 4.29s\n",
      "128:\tlearn: 0.3910932\ttotal: 3.21s\tremaining: 4.25s\n",
      "129:\tlearn: 0.3899908\ttotal: 3.23s\tremaining: 4.22s\n",
      "130:\tlearn: 0.3887893\ttotal: 3.25s\tremaining: 4.2s\n",
      "131:\tlearn: 0.3878779\ttotal: 3.28s\tremaining: 4.17s\n",
      "132:\tlearn: 0.3854324\ttotal: 3.3s\tremaining: 4.15s\n",
      "133:\tlearn: 0.3848655\ttotal: 3.33s\tremaining: 4.12s\n",
      "134:\tlearn: 0.3829461\ttotal: 3.35s\tremaining: 4.09s\n",
      "135:\tlearn: 0.3812624\ttotal: 3.38s\tremaining: 4.07s\n",
      "136:\tlearn: 0.3807075\ttotal: 3.4s\tremaining: 4.05s\n",
      "137:\tlearn: 0.3793296\ttotal: 3.43s\tremaining: 4.03s\n",
      "138:\tlearn: 0.3775384\ttotal: 3.46s\tremaining: 4.01s\n",
      "139:\tlearn: 0.3770394\ttotal: 3.49s\tremaining: 3.98s\n",
      "140:\tlearn: 0.3760403\ttotal: 3.5s\tremaining: 3.95s\n",
      "141:\tlearn: 0.3754060\ttotal: 3.52s\tremaining: 3.91s\n",
      "142:\tlearn: 0.3742802\ttotal: 3.53s\tremaining: 3.88s\n",
      "143:\tlearn: 0.3729668\ttotal: 3.55s\tremaining: 3.85s\n",
      "144:\tlearn: 0.3718032\ttotal: 3.57s\tremaining: 3.82s\n",
      "145:\tlearn: 0.3705672\ttotal: 3.59s\tremaining: 3.78s\n",
      "146:\tlearn: 0.3691455\ttotal: 3.6s\tremaining: 3.75s\n",
      "147:\tlearn: 0.3686972\ttotal: 3.61s\tremaining: 3.71s\n",
      "148:\tlearn: 0.3679673\ttotal: 3.63s\tremaining: 3.68s\n",
      "149:\tlearn: 0.3673694\ttotal: 3.65s\tremaining: 3.65s\n",
      "150:\tlearn: 0.3662388\ttotal: 3.66s\tremaining: 3.62s\n",
      "151:\tlearn: 0.3649266\ttotal: 3.68s\tremaining: 3.58s\n",
      "152:\tlearn: 0.3635472\ttotal: 3.7s\tremaining: 3.55s\n",
      "153:\tlearn: 0.3630888\ttotal: 3.71s\tremaining: 3.52s\n",
      "154:\tlearn: 0.3621698\ttotal: 3.73s\tremaining: 3.49s\n",
      "155:\tlearn: 0.3615773\ttotal: 3.75s\tremaining: 3.46s\n",
      "156:\tlearn: 0.3606084\ttotal: 3.77s\tremaining: 3.43s\n",
      "157:\tlearn: 0.3591553\ttotal: 3.79s\tremaining: 3.4s\n",
      "158:\tlearn: 0.3582056\ttotal: 3.8s\tremaining: 3.37s\n",
      "159:\tlearn: 0.3576034\ttotal: 3.82s\tremaining: 3.34s\n",
      "160:\tlearn: 0.3566517\ttotal: 3.84s\tremaining: 3.32s\n",
      "161:\tlearn: 0.3563953\ttotal: 3.87s\tremaining: 3.29s\n",
      "162:\tlearn: 0.3555426\ttotal: 3.89s\tremaining: 3.27s\n",
      "163:\tlearn: 0.3538127\ttotal: 3.92s\tremaining: 3.25s\n",
      "164:\tlearn: 0.3523386\ttotal: 3.94s\tremaining: 3.22s\n",
      "165:\tlearn: 0.3509496\ttotal: 3.96s\tremaining: 3.2s\n",
      "166:\tlearn: 0.3499814\ttotal: 3.99s\tremaining: 3.17s\n",
      "167:\tlearn: 0.3494707\ttotal: 4.01s\tremaining: 3.15s\n",
      "168:\tlearn: 0.3479599\ttotal: 4.04s\tremaining: 3.13s\n",
      "169:\tlearn: 0.3468880\ttotal: 4.06s\tremaining: 3.1s\n",
      "170:\tlearn: 0.3457600\ttotal: 4.07s\tremaining: 3.07s\n",
      "171:\tlearn: 0.3452815\ttotal: 4.09s\tremaining: 3.04s\n",
      "172:\tlearn: 0.3446983\ttotal: 4.11s\tremaining: 3.01s\n",
      "173:\tlearn: 0.3443792\ttotal: 4.12s\tremaining: 2.98s\n",
      "174:\tlearn: 0.3432527\ttotal: 4.13s\tremaining: 2.95s\n",
      "175:\tlearn: 0.3421646\ttotal: 4.15s\tremaining: 2.92s\n",
      "176:\tlearn: 0.3409661\ttotal: 4.16s\tremaining: 2.89s\n",
      "177:\tlearn: 0.3404454\ttotal: 4.18s\tremaining: 2.86s\n",
      "178:\tlearn: 0.3402090\ttotal: 4.19s\tremaining: 2.83s\n",
      "179:\tlearn: 0.3389590\ttotal: 4.2s\tremaining: 2.8s\n",
      "180:\tlearn: 0.3377824\ttotal: 4.22s\tremaining: 2.77s\n",
      "181:\tlearn: 0.3366980\ttotal: 4.24s\tremaining: 2.75s\n",
      "182:\tlearn: 0.3363215\ttotal: 4.25s\tremaining: 2.72s\n",
      "183:\tlearn: 0.3351214\ttotal: 4.27s\tremaining: 2.69s\n",
      "184:\tlearn: 0.3337771\ttotal: 4.29s\tremaining: 2.67s\n",
      "185:\tlearn: 0.3335901\ttotal: 4.31s\tremaining: 2.64s\n",
      "186:\tlearn: 0.3332062\ttotal: 4.33s\tremaining: 2.61s\n",
      "187:\tlearn: 0.3318233\ttotal: 4.35s\tremaining: 2.59s\n",
      "188:\tlearn: 0.3306851\ttotal: 4.37s\tremaining: 2.56s\n",
      "189:\tlearn: 0.3305706\ttotal: 4.38s\tremaining: 2.54s\n",
      "190:\tlearn: 0.3295583\ttotal: 4.4s\tremaining: 2.51s\n",
      "191:\tlearn: 0.3284863\ttotal: 4.42s\tremaining: 2.49s\n",
      "192:\tlearn: 0.3281150\ttotal: 4.44s\tremaining: 2.46s\n",
      "193:\tlearn: 0.3275913\ttotal: 4.46s\tremaining: 2.44s\n",
      "194:\tlearn: 0.3263121\ttotal: 4.49s\tremaining: 2.42s\n",
      "195:\tlearn: 0.3250736\ttotal: 4.51s\tremaining: 2.39s\n",
      "196:\tlearn: 0.3244863\ttotal: 4.54s\tremaining: 2.37s\n",
      "197:\tlearn: 0.3237127\ttotal: 4.57s\tremaining: 2.35s\n",
      "198:\tlearn: 0.3227509\ttotal: 4.59s\tremaining: 2.33s\n",
      "199:\tlearn: 0.3225427\ttotal: 4.62s\tremaining: 2.31s\n",
      "200:\tlearn: 0.3216052\ttotal: 4.64s\tremaining: 2.29s\n",
      "201:\tlearn: 0.3210849\ttotal: 4.67s\tremaining: 2.27s\n",
      "202:\tlearn: 0.3200747\ttotal: 4.71s\tremaining: 2.25s\n",
      "203:\tlearn: 0.3197091\ttotal: 4.75s\tremaining: 2.23s\n",
      "204:\tlearn: 0.3188926\ttotal: 4.78s\tremaining: 2.21s\n",
      "205:\tlearn: 0.3184885\ttotal: 4.82s\tremaining: 2.2s\n",
      "206:\tlearn: 0.3178186\ttotal: 4.85s\tremaining: 2.18s\n",
      "207:\tlearn: 0.3169427\ttotal: 4.89s\tremaining: 2.16s\n",
      "208:\tlearn: 0.3163718\ttotal: 4.92s\tremaining: 2.14s\n",
      "209:\tlearn: 0.3158060\ttotal: 4.96s\tremaining: 2.13s\n",
      "210:\tlearn: 0.3149162\ttotal: 5s\tremaining: 2.11s\n",
      "211:\tlearn: 0.3148685\ttotal: 5s\tremaining: 2.08s\n",
      "212:\tlearn: 0.3139562\ttotal: 5.04s\tremaining: 2.06s\n",
      "213:\tlearn: 0.3134186\ttotal: 5.07s\tremaining: 2.04s\n",
      "214:\tlearn: 0.3127172\ttotal: 5.1s\tremaining: 2.02s\n",
      "215:\tlearn: 0.3121172\ttotal: 5.12s\tremaining: 1.99s\n",
      "216:\tlearn: 0.3116492\ttotal: 5.15s\tremaining: 1.97s\n",
      "217:\tlearn: 0.3111142\ttotal: 5.17s\tremaining: 1.95s\n",
      "218:\tlearn: 0.3097195\ttotal: 5.2s\tremaining: 1.92s\n",
      "219:\tlearn: 0.3093234\ttotal: 5.23s\tremaining: 1.9s\n",
      "220:\tlearn: 0.3083128\ttotal: 5.25s\tremaining: 1.88s\n",
      "221:\tlearn: 0.3077216\ttotal: 5.27s\tremaining: 1.85s\n",
      "222:\tlearn: 0.3067435\ttotal: 5.3s\tremaining: 1.83s\n",
      "223:\tlearn: 0.3062165\ttotal: 5.32s\tremaining: 1.8s\n",
      "224:\tlearn: 0.3054927\ttotal: 5.35s\tremaining: 1.78s\n",
      "225:\tlearn: 0.3048022\ttotal: 5.37s\tremaining: 1.76s\n",
      "226:\tlearn: 0.3043235\ttotal: 5.39s\tremaining: 1.73s\n",
      "227:\tlearn: 0.3037640\ttotal: 5.41s\tremaining: 1.71s\n",
      "228:\tlearn: 0.3030603\ttotal: 5.43s\tremaining: 1.68s\n",
      "229:\tlearn: 0.3023518\ttotal: 5.45s\tremaining: 1.66s\n",
      "230:\tlearn: 0.3020708\ttotal: 5.47s\tremaining: 1.63s\n",
      "231:\tlearn: 0.3015116\ttotal: 5.49s\tremaining: 1.61s\n",
      "232:\tlearn: 0.3007897\ttotal: 5.52s\tremaining: 1.59s\n",
      "233:\tlearn: 0.2997428\ttotal: 5.54s\tremaining: 1.56s\n",
      "234:\tlearn: 0.2991210\ttotal: 5.57s\tremaining: 1.54s\n",
      "235:\tlearn: 0.2987425\ttotal: 5.6s\tremaining: 1.52s\n",
      "236:\tlearn: 0.2980339\ttotal: 5.63s\tremaining: 1.5s\n",
      "237:\tlearn: 0.2978281\ttotal: 5.64s\tremaining: 1.47s\n",
      "238:\tlearn: 0.2973135\ttotal: 5.66s\tremaining: 1.45s\n",
      "239:\tlearn: 0.2966227\ttotal: 5.69s\tremaining: 1.42s\n",
      "240:\tlearn: 0.2959717\ttotal: 5.71s\tremaining: 1.4s\n",
      "241:\tlearn: 0.2956204\ttotal: 5.73s\tremaining: 1.37s\n",
      "242:\tlearn: 0.2952998\ttotal: 5.76s\tremaining: 1.35s\n",
      "243:\tlearn: 0.2951033\ttotal: 5.81s\tremaining: 1.33s\n",
      "244:\tlearn: 0.2944837\ttotal: 5.86s\tremaining: 1.31s\n",
      "245:\tlearn: 0.2942383\ttotal: 5.9s\tremaining: 1.3s\n",
      "246:\tlearn: 0.2932377\ttotal: 5.95s\tremaining: 1.28s\n",
      "247:\tlearn: 0.2926755\ttotal: 6s\tremaining: 1.26s\n",
      "248:\tlearn: 0.2924892\ttotal: 6.04s\tremaining: 1.24s\n",
      "249:\tlearn: 0.2921062\ttotal: 6.08s\tremaining: 1.22s\n",
      "250:\tlearn: 0.2918480\ttotal: 6.1s\tremaining: 1.19s\n",
      "251:\tlearn: 0.2912666\ttotal: 6.12s\tremaining: 1.17s\n",
      "252:\tlearn: 0.2910736\ttotal: 6.13s\tremaining: 1.14s\n",
      "253:\tlearn: 0.2905123\ttotal: 6.15s\tremaining: 1.11s\n",
      "254:\tlearn: 0.2899805\ttotal: 6.17s\tremaining: 1.09s\n",
      "255:\tlearn: 0.2894331\ttotal: 6.19s\tremaining: 1.06s\n",
      "256:\tlearn: 0.2885934\ttotal: 6.21s\tremaining: 1.04s\n",
      "257:\tlearn: 0.2880241\ttotal: 6.24s\tremaining: 1.01s\n",
      "258:\tlearn: 0.2877345\ttotal: 6.26s\tremaining: 991ms\n",
      "259:\tlearn: 0.2871031\ttotal: 6.29s\tremaining: 968ms\n",
      "260:\tlearn: 0.2865257\ttotal: 6.31s\tremaining: 943ms\n",
      "261:\tlearn: 0.2857696\ttotal: 6.33s\tremaining: 919ms\n",
      "262:\tlearn: 0.2851386\ttotal: 6.36s\tremaining: 894ms\n",
      "263:\tlearn: 0.2846715\ttotal: 6.38s\tremaining: 869ms\n",
      "264:\tlearn: 0.2843212\ttotal: 6.39s\tremaining: 845ms\n",
      "265:\tlearn: 0.2837335\ttotal: 6.41s\tremaining: 820ms\n",
      "266:\tlearn: 0.2829457\ttotal: 6.43s\tremaining: 795ms\n",
      "267:\tlearn: 0.2825119\ttotal: 6.45s\tremaining: 770ms\n",
      "268:\tlearn: 0.2817245\ttotal: 6.46s\tremaining: 745ms\n",
      "269:\tlearn: 0.2813144\ttotal: 6.48s\tremaining: 720ms\n",
      "270:\tlearn: 0.2807644\ttotal: 6.5s\tremaining: 695ms\n",
      "271:\tlearn: 0.2802664\ttotal: 6.51s\tremaining: 671ms\n",
      "272:\tlearn: 0.2795455\ttotal: 6.53s\tremaining: 646ms\n",
      "273:\tlearn: 0.2788351\ttotal: 6.55s\tremaining: 621ms\n",
      "274:\tlearn: 0.2782771\ttotal: 6.56s\tremaining: 597ms\n",
      "275:\tlearn: 0.2779257\ttotal: 6.58s\tremaining: 572ms\n",
      "276:\tlearn: 0.2774382\ttotal: 6.59s\tremaining: 547ms\n",
      "277:\tlearn: 0.2771612\ttotal: 6.61s\tremaining: 523ms\n",
      "278:\tlearn: 0.2769162\ttotal: 6.63s\tremaining: 499ms\n",
      "279:\tlearn: 0.2766319\ttotal: 6.64s\tremaining: 475ms\n",
      "280:\tlearn: 0.2761934\ttotal: 6.66s\tremaining: 451ms\n",
      "281:\tlearn: 0.2757988\ttotal: 6.68s\tremaining: 427ms\n",
      "282:\tlearn: 0.2752298\ttotal: 6.71s\tremaining: 403ms\n",
      "283:\tlearn: 0.2747692\ttotal: 6.74s\tremaining: 379ms\n",
      "284:\tlearn: 0.2742907\ttotal: 6.76s\tremaining: 356ms\n",
      "285:\tlearn: 0.2737693\ttotal: 6.79s\tremaining: 332ms\n",
      "286:\tlearn: 0.2732739\ttotal: 6.81s\tremaining: 309ms\n",
      "287:\tlearn: 0.2727215\ttotal: 6.83s\tremaining: 285ms\n",
      "288:\tlearn: 0.2723342\ttotal: 6.85s\tremaining: 261ms\n",
      "289:\tlearn: 0.2720700\ttotal: 6.87s\tremaining: 237ms\n",
      "290:\tlearn: 0.2719063\ttotal: 6.9s\tremaining: 213ms\n",
      "291:\tlearn: 0.2714158\ttotal: 6.92s\tremaining: 190ms\n",
      "292:\tlearn: 0.2712886\ttotal: 6.94s\tremaining: 166ms\n",
      "293:\tlearn: 0.2711350\ttotal: 6.97s\tremaining: 142ms\n",
      "294:\tlearn: 0.2709294\ttotal: 6.99s\tremaining: 119ms\n",
      "295:\tlearn: 0.2704545\ttotal: 7.02s\tremaining: 94.9ms\n",
      "296:\tlearn: 0.2700182\ttotal: 7.05s\tremaining: 71.2ms\n",
      "297:\tlearn: 0.2693281\ttotal: 7.08s\tremaining: 47.5ms\n",
      "298:\tlearn: 0.2689187\ttotal: 7.1s\tremaining: 23.8ms\n",
      "299:\tlearn: 0.2682112\ttotal: 7.13s\tremaining: 0us\n",
      "0:\tlearn: 3.1220953\ttotal: 16ms\tremaining: 4.8s\n",
      "1:\tlearn: 3.0279566\ttotal: 32.8ms\tremaining: 4.89s\n",
      "2:\tlearn: 2.9619169\ttotal: 49.5ms\tremaining: 4.9s\n",
      "3:\tlearn: 2.8847731\ttotal: 66.4ms\tremaining: 4.91s\n",
      "4:\tlearn: 2.8078074\ttotal: 83.8ms\tremaining: 4.94s\n",
      "5:\tlearn: 2.7414490\ttotal: 101ms\tremaining: 4.94s\n",
      "6:\tlearn: 2.6695528\ttotal: 119ms\tremaining: 5s\n",
      "7:\tlearn: 2.6174181\ttotal: 136ms\tremaining: 4.97s\n",
      "8:\tlearn: 2.5576099\ttotal: 153ms\tremaining: 4.95s\n",
      "9:\tlearn: 2.5039134\ttotal: 172ms\tremaining: 4.98s\n",
      "10:\tlearn: 2.4503224\ttotal: 191ms\tremaining: 5.03s\n",
      "11:\tlearn: 2.4043972\ttotal: 215ms\tremaining: 5.16s\n",
      "12:\tlearn: 2.3634579\ttotal: 240ms\tremaining: 5.3s\n",
      "13:\tlearn: 2.3161349\ttotal: 266ms\tremaining: 5.44s\n",
      "14:\tlearn: 2.2672801\ttotal: 293ms\tremaining: 5.56s\n",
      "15:\tlearn: 2.2284433\ttotal: 319ms\tremaining: 5.66s\n",
      "16:\tlearn: 2.1923848\ttotal: 346ms\tremaining: 5.76s\n",
      "17:\tlearn: 2.1632076\ttotal: 372ms\tremaining: 5.83s\n",
      "18:\tlearn: 2.1326013\ttotal: 400ms\tremaining: 5.91s\n",
      "19:\tlearn: 2.0963571\ttotal: 427ms\tremaining: 5.98s\n",
      "20:\tlearn: 2.0667126\ttotal: 455ms\tremaining: 6.04s\n",
      "21:\tlearn: 2.0467455\ttotal: 486ms\tremaining: 6.14s\n",
      "22:\tlearn: 2.0210141\ttotal: 513ms\tremaining: 6.17s\n",
      "23:\tlearn: 1.9904106\ttotal: 539ms\tremaining: 6.2s\n",
      "24:\tlearn: 1.9619748\ttotal: 567ms\tremaining: 6.23s\n",
      "25:\tlearn: 1.9361171\ttotal: 593ms\tremaining: 6.25s\n",
      "26:\tlearn: 1.9147064\ttotal: 619ms\tremaining: 6.26s\n",
      "27:\tlearn: 1.8917572\ttotal: 647ms\tremaining: 6.28s\n",
      "28:\tlearn: 1.8779286\ttotal: 674ms\tremaining: 6.3s\n",
      "29:\tlearn: 1.8519131\ttotal: 702ms\tremaining: 6.32s\n",
      "30:\tlearn: 1.8283043\ttotal: 729ms\tremaining: 6.33s\n",
      "31:\tlearn: 1.8099532\ttotal: 757ms\tremaining: 6.34s\n",
      "32:\tlearn: 1.7936736\ttotal: 784ms\tremaining: 6.34s\n",
      "33:\tlearn: 1.7847067\ttotal: 811ms\tremaining: 6.35s\n",
      "34:\tlearn: 1.7626917\ttotal: 838ms\tremaining: 6.34s\n",
      "35:\tlearn: 1.7411332\ttotal: 867ms\tremaining: 6.35s\n",
      "36:\tlearn: 1.7271528\ttotal: 894ms\tremaining: 6.35s\n",
      "37:\tlearn: 1.7043367\ttotal: 921ms\tremaining: 6.35s\n",
      "38:\tlearn: 1.6911512\ttotal: 948ms\tremaining: 6.34s\n",
      "39:\tlearn: 1.6825454\ttotal: 975ms\tremaining: 6.34s\n",
      "40:\tlearn: 1.6647777\ttotal: 1s\tremaining: 6.33s\n",
      "41:\tlearn: 1.6454367\ttotal: 1.03s\tremaining: 6.33s\n",
      "42:\tlearn: 1.6387950\ttotal: 1.05s\tremaining: 6.3s\n",
      "43:\tlearn: 1.6192610\ttotal: 1.07s\tremaining: 6.23s\n",
      "44:\tlearn: 1.6084176\ttotal: 1.09s\tremaining: 6.17s\n",
      "45:\tlearn: 1.5975700\ttotal: 1.1s\tremaining: 6.1s\n",
      "46:\tlearn: 1.5807874\ttotal: 1.12s\tremaining: 6.04s\n",
      "47:\tlearn: 1.5719989\ttotal: 1.14s\tremaining: 5.98s\n",
      "48:\tlearn: 1.5627490\ttotal: 1.16s\tremaining: 5.92s\n",
      "49:\tlearn: 1.5530045\ttotal: 1.17s\tremaining: 5.86s\n",
      "50:\tlearn: 1.5401255\ttotal: 1.19s\tremaining: 5.8s\n",
      "51:\tlearn: 1.5322666\ttotal: 1.2s\tremaining: 5.74s\n",
      "52:\tlearn: 1.5237779\ttotal: 1.22s\tremaining: 5.68s\n",
      "53:\tlearn: 1.5149322\ttotal: 1.23s\tremaining: 5.62s\n",
      "54:\tlearn: 1.5081602\ttotal: 1.25s\tremaining: 5.57s\n",
      "55:\tlearn: 1.4958389\ttotal: 1.27s\tremaining: 5.52s\n",
      "56:\tlearn: 1.4846483\ttotal: 1.28s\tremaining: 5.47s\n",
      "57:\tlearn: 1.4753205\ttotal: 1.3s\tremaining: 5.43s\n",
      "58:\tlearn: 1.4653666\ttotal: 1.32s\tremaining: 5.39s\n",
      "59:\tlearn: 1.4576744\ttotal: 1.34s\tremaining: 5.35s\n",
      "60:\tlearn: 1.4436830\ttotal: 1.35s\tremaining: 5.31s\n",
      "61:\tlearn: 1.4339756\ttotal: 1.38s\tremaining: 5.28s\n",
      "62:\tlearn: 1.4237399\ttotal: 1.4s\tremaining: 5.25s\n",
      "63:\tlearn: 1.4165287\ttotal: 1.42s\tremaining: 5.23s\n",
      "64:\tlearn: 1.4083537\ttotal: 1.44s\tremaining: 5.21s\n",
      "65:\tlearn: 1.4015401\ttotal: 1.46s\tremaining: 5.19s\n",
      "66:\tlearn: 1.3966402\ttotal: 1.5s\tremaining: 5.23s\n",
      "67:\tlearn: 1.3931322\ttotal: 1.55s\tremaining: 5.31s\n",
      "68:\tlearn: 1.3863990\ttotal: 1.6s\tremaining: 5.38s\n",
      "69:\tlearn: 1.3807776\ttotal: 1.65s\tremaining: 5.42s\n",
      "70:\tlearn: 1.3768618\ttotal: 1.69s\tremaining: 5.46s\n",
      "71:\tlearn: 1.3657982\ttotal: 1.74s\tremaining: 5.5s\n",
      "72:\tlearn: 1.3559536\ttotal: 1.78s\tremaining: 5.54s\n",
      "73:\tlearn: 1.3521231\ttotal: 1.82s\tremaining: 5.57s\n",
      "74:\tlearn: 1.3426096\ttotal: 1.87s\tremaining: 5.61s\n",
      "75:\tlearn: 1.3331009\ttotal: 1.92s\tremaining: 5.64s\n",
      "76:\tlearn: 1.3217527\ttotal: 1.97s\tremaining: 5.7s\n",
      "77:\tlearn: 1.3122572\ttotal: 2.02s\tremaining: 5.74s\n",
      "78:\tlearn: 1.3104856\ttotal: 2.06s\tremaining: 5.76s\n",
      "79:\tlearn: 1.3034615\ttotal: 2.1s\tremaining: 5.78s\n",
      "80:\tlearn: 1.2956348\ttotal: 2.15s\tremaining: 5.8s\n",
      "81:\tlearn: 1.2876315\ttotal: 2.19s\tremaining: 5.83s\n",
      "82:\tlearn: 1.2777509\ttotal: 2.24s\tremaining: 5.86s\n",
      "83:\tlearn: 1.2706512\ttotal: 2.28s\tremaining: 5.87s\n",
      "84:\tlearn: 1.2657385\ttotal: 2.33s\tremaining: 5.89s\n",
      "85:\tlearn: 1.2545021\ttotal: 2.37s\tremaining: 5.89s\n",
      "86:\tlearn: 1.2486268\ttotal: 2.38s\tremaining: 5.84s\n",
      "87:\tlearn: 1.2418092\ttotal: 2.4s\tremaining: 5.78s\n",
      "88:\tlearn: 1.2341709\ttotal: 2.42s\tremaining: 5.73s\n",
      "89:\tlearn: 1.2327982\ttotal: 2.42s\tremaining: 5.65s\n",
      "90:\tlearn: 1.2254184\ttotal: 2.44s\tremaining: 5.6s\n",
      "91:\tlearn: 1.2212346\ttotal: 2.47s\tremaining: 5.59s\n",
      "92:\tlearn: 1.2159567\ttotal: 2.5s\tremaining: 5.57s\n",
      "93:\tlearn: 1.2106841\ttotal: 2.52s\tremaining: 5.53s\n",
      "94:\tlearn: 1.2032925\ttotal: 2.54s\tremaining: 5.49s\n",
      "95:\tlearn: 1.1990794\ttotal: 2.56s\tremaining: 5.45s\n",
      "96:\tlearn: 1.1919569\ttotal: 2.59s\tremaining: 5.41s\n",
      "97:\tlearn: 1.1848794\ttotal: 2.61s\tremaining: 5.38s\n",
      "98:\tlearn: 1.1805637\ttotal: 2.62s\tremaining: 5.32s\n",
      "99:\tlearn: 1.1745212\ttotal: 2.64s\tremaining: 5.28s\n",
      "100:\tlearn: 1.1688463\ttotal: 2.66s\tremaining: 5.25s\n",
      "101:\tlearn: 1.1642235\ttotal: 2.68s\tremaining: 5.21s\n",
      "102:\tlearn: 1.1591829\ttotal: 2.7s\tremaining: 5.17s\n",
      "103:\tlearn: 1.1560893\ttotal: 2.72s\tremaining: 5.13s\n",
      "104:\tlearn: 1.1547502\ttotal: 2.75s\tremaining: 5.1s\n",
      "105:\tlearn: 1.1510169\ttotal: 2.77s\tremaining: 5.08s\n",
      "106:\tlearn: 1.1468555\ttotal: 2.8s\tremaining: 5.04s\n",
      "107:\tlearn: 1.1457001\ttotal: 2.82s\tremaining: 5.01s\n",
      "108:\tlearn: 1.1420701\ttotal: 2.84s\tremaining: 4.99s\n",
      "109:\tlearn: 1.1368121\ttotal: 2.87s\tremaining: 4.96s\n",
      "110:\tlearn: 1.1333011\ttotal: 2.89s\tremaining: 4.93s\n",
      "111:\tlearn: 1.1277237\ttotal: 2.92s\tremaining: 4.9s\n",
      "112:\tlearn: 1.1183943\ttotal: 2.94s\tremaining: 4.87s\n",
      "113:\tlearn: 1.1113082\ttotal: 2.97s\tremaining: 4.85s\n",
      "114:\tlearn: 1.1059808\ttotal: 3s\tremaining: 4.82s\n",
      "115:\tlearn: 1.1022166\ttotal: 3.02s\tremaining: 4.8s\n",
      "116:\tlearn: 1.0977884\ttotal: 3.05s\tremaining: 4.77s\n",
      "117:\tlearn: 1.0931108\ttotal: 3.08s\tremaining: 4.75s\n",
      "118:\tlearn: 1.0913028\ttotal: 3.08s\tremaining: 4.69s\n",
      "119:\tlearn: 1.0849236\ttotal: 3.11s\tremaining: 4.67s\n",
      "120:\tlearn: 1.0831649\ttotal: 3.14s\tremaining: 4.64s\n",
      "121:\tlearn: 1.0774836\ttotal: 3.17s\tremaining: 4.62s\n",
      "122:\tlearn: 1.0727861\ttotal: 3.19s\tremaining: 4.59s\n",
      "123:\tlearn: 1.0697867\ttotal: 3.22s\tremaining: 4.57s\n",
      "124:\tlearn: 1.0683927\ttotal: 3.25s\tremaining: 4.54s\n",
      "125:\tlearn: 1.0635526\ttotal: 3.27s\tremaining: 4.52s\n",
      "126:\tlearn: 1.0615321\ttotal: 3.3s\tremaining: 4.5s\n",
      "127:\tlearn: 1.0554701\ttotal: 3.33s\tremaining: 4.47s\n",
      "128:\tlearn: 1.0534169\ttotal: 3.35s\tremaining: 4.45s\n",
      "129:\tlearn: 1.0488477\ttotal: 3.38s\tremaining: 4.42s\n",
      "130:\tlearn: 1.0467235\ttotal: 3.41s\tremaining: 4.39s\n",
      "131:\tlearn: 1.0423138\ttotal: 3.43s\tremaining: 4.37s\n",
      "132:\tlearn: 1.0375913\ttotal: 3.46s\tremaining: 4.35s\n",
      "133:\tlearn: 1.0348715\ttotal: 3.49s\tremaining: 4.32s\n",
      "134:\tlearn: 1.0313752\ttotal: 3.52s\tremaining: 4.3s\n",
      "135:\tlearn: 1.0266968\ttotal: 3.55s\tremaining: 4.28s\n",
      "136:\tlearn: 1.0239910\ttotal: 3.57s\tremaining: 4.25s\n",
      "137:\tlearn: 1.0224373\ttotal: 3.58s\tremaining: 4.21s\n",
      "138:\tlearn: 1.0191406\ttotal: 3.6s\tremaining: 4.17s\n",
      "139:\tlearn: 1.0183908\ttotal: 3.6s\tremaining: 4.12s\n",
      "140:\tlearn: 1.0117048\ttotal: 3.62s\tremaining: 4.08s\n",
      "141:\tlearn: 1.0097293\ttotal: 3.64s\tremaining: 4.05s\n",
      "142:\tlearn: 1.0077391\ttotal: 3.65s\tremaining: 4.01s\n",
      "143:\tlearn: 1.0067560\ttotal: 3.67s\tremaining: 3.97s\n",
      "144:\tlearn: 1.0038493\ttotal: 3.69s\tremaining: 3.94s\n",
      "145:\tlearn: 1.0001058\ttotal: 3.7s\tremaining: 3.9s\n",
      "146:\tlearn: 0.9938441\ttotal: 3.72s\tremaining: 3.87s\n",
      "147:\tlearn: 0.9890696\ttotal: 3.73s\tremaining: 3.83s\n",
      "148:\tlearn: 0.9851051\ttotal: 3.75s\tremaining: 3.8s\n",
      "149:\tlearn: 0.9839195\ttotal: 3.77s\tremaining: 3.77s\n",
      "150:\tlearn: 0.9804666\ttotal: 3.78s\tremaining: 3.73s\n",
      "151:\tlearn: 0.9761860\ttotal: 3.8s\tremaining: 3.7s\n",
      "152:\tlearn: 0.9681719\ttotal: 3.82s\tremaining: 3.67s\n",
      "153:\tlearn: 0.9623376\ttotal: 3.83s\tremaining: 3.63s\n",
      "154:\tlearn: 0.9603324\ttotal: 3.85s\tremaining: 3.6s\n",
      "155:\tlearn: 0.9586154\ttotal: 3.87s\tremaining: 3.57s\n",
      "156:\tlearn: 0.9563047\ttotal: 3.89s\tremaining: 3.55s\n",
      "157:\tlearn: 0.9513401\ttotal: 3.91s\tremaining: 3.52s\n",
      "158:\tlearn: 0.9488181\ttotal: 3.94s\tremaining: 3.49s\n",
      "159:\tlearn: 0.9433838\ttotal: 3.96s\tremaining: 3.47s\n",
      "160:\tlearn: 0.9403396\ttotal: 3.99s\tremaining: 3.44s\n",
      "161:\tlearn: 0.9395498\ttotal: 4.01s\tremaining: 3.42s\n",
      "162:\tlearn: 0.9361551\ttotal: 4.04s\tremaining: 3.4s\n",
      "163:\tlearn: 0.9347517\ttotal: 4.07s\tremaining: 3.37s\n",
      "164:\tlearn: 0.9334843\ttotal: 4.1s\tremaining: 3.35s\n",
      "165:\tlearn: 0.9330575\ttotal: 4.1s\tremaining: 3.31s\n",
      "166:\tlearn: 0.9313986\ttotal: 4.13s\tremaining: 3.29s\n",
      "167:\tlearn: 0.9268582\ttotal: 4.16s\tremaining: 3.26s\n",
      "168:\tlearn: 0.9230010\ttotal: 4.18s\tremaining: 3.24s\n",
      "169:\tlearn: 0.9211164\ttotal: 4.21s\tremaining: 3.22s\n",
      "170:\tlearn: 0.9148932\ttotal: 4.24s\tremaining: 3.19s\n",
      "171:\tlearn: 0.9109006\ttotal: 4.26s\tremaining: 3.17s\n",
      "172:\tlearn: 0.9096646\ttotal: 4.29s\tremaining: 3.15s\n",
      "173:\tlearn: 0.9070573\ttotal: 4.32s\tremaining: 3.12s\n",
      "174:\tlearn: 0.9045344\ttotal: 4.34s\tremaining: 3.1s\n",
      "175:\tlearn: 0.9031712\ttotal: 4.37s\tremaining: 3.08s\n",
      "176:\tlearn: 0.9027078\ttotal: 4.4s\tremaining: 3.06s\n",
      "177:\tlearn: 0.8991375\ttotal: 4.42s\tremaining: 3.03s\n",
      "178:\tlearn: 0.8969644\ttotal: 4.45s\tremaining: 3.01s\n",
      "179:\tlearn: 0.8935499\ttotal: 4.48s\tremaining: 2.99s\n",
      "180:\tlearn: 0.8915058\ttotal: 4.51s\tremaining: 2.96s\n",
      "181:\tlearn: 0.8894215\ttotal: 4.54s\tremaining: 2.94s\n",
      "182:\tlearn: 0.8865536\ttotal: 4.56s\tremaining: 2.92s\n",
      "183:\tlearn: 0.8825794\ttotal: 4.59s\tremaining: 2.89s\n",
      "184:\tlearn: 0.8814401\ttotal: 4.62s\tremaining: 2.87s\n",
      "185:\tlearn: 0.8782065\ttotal: 4.64s\tremaining: 2.85s\n",
      "186:\tlearn: 0.8759470\ttotal: 4.67s\tremaining: 2.82s\n",
      "187:\tlearn: 0.8730274\ttotal: 4.7s\tremaining: 2.8s\n",
      "188:\tlearn: 0.8703684\ttotal: 4.73s\tremaining: 2.78s\n",
      "189:\tlearn: 0.8656421\ttotal: 4.75s\tremaining: 2.75s\n",
      "190:\tlearn: 0.8620526\ttotal: 4.76s\tremaining: 2.72s\n",
      "191:\tlearn: 0.8583467\ttotal: 4.78s\tremaining: 2.69s\n",
      "192:\tlearn: 0.8560670\ttotal: 4.8s\tremaining: 2.66s\n",
      "193:\tlearn: 0.8528864\ttotal: 4.82s\tremaining: 2.63s\n",
      "194:\tlearn: 0.8503320\ttotal: 4.83s\tremaining: 2.6s\n",
      "195:\tlearn: 0.8464767\ttotal: 4.85s\tremaining: 2.57s\n",
      "196:\tlearn: 0.8450015\ttotal: 4.87s\tremaining: 2.54s\n",
      "197:\tlearn: 0.8427826\ttotal: 4.88s\tremaining: 2.51s\n",
      "198:\tlearn: 0.8415170\ttotal: 4.9s\tremaining: 2.49s\n",
      "199:\tlearn: 0.8394574\ttotal: 4.92s\tremaining: 2.46s\n",
      "200:\tlearn: 0.8360519\ttotal: 4.94s\tremaining: 2.43s\n",
      "201:\tlearn: 0.8340450\ttotal: 4.96s\tremaining: 2.4s\n",
      "202:\tlearn: 0.8329815\ttotal: 4.98s\tremaining: 2.38s\n",
      "203:\tlearn: 0.8301717\ttotal: 5s\tremaining: 2.35s\n",
      "204:\tlearn: 0.8284840\ttotal: 5.02s\tremaining: 2.33s\n",
      "205:\tlearn: 0.8264431\ttotal: 5.04s\tremaining: 2.3s\n",
      "206:\tlearn: 0.8241208\ttotal: 5.07s\tremaining: 2.28s\n",
      "207:\tlearn: 0.8225231\ttotal: 5.09s\tremaining: 2.25s\n",
      "208:\tlearn: 0.8185778\ttotal: 5.14s\tremaining: 2.24s\n",
      "209:\tlearn: 0.8169441\ttotal: 5.2s\tremaining: 2.23s\n",
      "210:\tlearn: 0.8155941\ttotal: 5.25s\tremaining: 2.21s\n",
      "211:\tlearn: 0.8132789\ttotal: 5.27s\tremaining: 2.19s\n",
      "212:\tlearn: 0.8099501\ttotal: 5.29s\tremaining: 2.16s\n",
      "213:\tlearn: 0.8072593\ttotal: 5.32s\tremaining: 2.14s\n",
      "214:\tlearn: 0.8051425\ttotal: 5.34s\tremaining: 2.11s\n",
      "215:\tlearn: 0.8039136\ttotal: 5.37s\tremaining: 2.09s\n",
      "216:\tlearn: 0.8019610\ttotal: 5.4s\tremaining: 2.06s\n",
      "217:\tlearn: 0.8000897\ttotal: 5.42s\tremaining: 2.04s\n",
      "218:\tlearn: 0.7962948\ttotal: 5.45s\tremaining: 2.02s\n",
      "219:\tlearn: 0.7925247\ttotal: 5.48s\tremaining: 1.99s\n",
      "220:\tlearn: 0.7900005\ttotal: 5.51s\tremaining: 1.97s\n",
      "221:\tlearn: 0.7864900\ttotal: 5.54s\tremaining: 1.95s\n",
      "222:\tlearn: 0.7834706\ttotal: 5.57s\tremaining: 1.92s\n",
      "223:\tlearn: 0.7816678\ttotal: 5.59s\tremaining: 1.9s\n",
      "224:\tlearn: 0.7802290\ttotal: 5.62s\tremaining: 1.87s\n",
      "225:\tlearn: 0.7787959\ttotal: 5.65s\tremaining: 1.85s\n",
      "226:\tlearn: 0.7774234\ttotal: 5.7s\tremaining: 1.83s\n",
      "227:\tlearn: 0.7764810\ttotal: 5.75s\tremaining: 1.81s\n",
      "228:\tlearn: 0.7733115\ttotal: 5.79s\tremaining: 1.79s\n",
      "229:\tlearn: 0.7713969\ttotal: 5.84s\tremaining: 1.78s\n",
      "230:\tlearn: 0.7693803\ttotal: 5.88s\tremaining: 1.76s\n",
      "231:\tlearn: 0.7682649\ttotal: 5.93s\tremaining: 1.74s\n",
      "232:\tlearn: 0.7669492\ttotal: 5.98s\tremaining: 1.72s\n",
      "233:\tlearn: 0.7654284\ttotal: 6s\tremaining: 1.69s\n",
      "234:\tlearn: 0.7631874\ttotal: 6.02s\tremaining: 1.67s\n",
      "235:\tlearn: 0.7606795\ttotal: 6.04s\tremaining: 1.64s\n",
      "236:\tlearn: 0.7592652\ttotal: 6.05s\tremaining: 1.61s\n",
      "237:\tlearn: 0.7567921\ttotal: 6.07s\tremaining: 1.58s\n",
      "238:\tlearn: 0.7542900\ttotal: 6.09s\tremaining: 1.55s\n",
      "239:\tlearn: 0.7523299\ttotal: 6.1s\tremaining: 1.52s\n",
      "240:\tlearn: 0.7507866\ttotal: 6.12s\tremaining: 1.5s\n",
      "241:\tlearn: 0.7499622\ttotal: 6.14s\tremaining: 1.47s\n",
      "242:\tlearn: 0.7470666\ttotal: 6.15s\tremaining: 1.44s\n",
      "243:\tlearn: 0.7457588\ttotal: 6.17s\tremaining: 1.42s\n",
      "244:\tlearn: 0.7425239\ttotal: 6.18s\tremaining: 1.39s\n",
      "245:\tlearn: 0.7392294\ttotal: 6.2s\tremaining: 1.36s\n",
      "246:\tlearn: 0.7381489\ttotal: 6.22s\tremaining: 1.33s\n",
      "247:\tlearn: 0.7351547\ttotal: 6.24s\tremaining: 1.31s\n",
      "248:\tlearn: 0.7345041\ttotal: 6.26s\tremaining: 1.28s\n",
      "249:\tlearn: 0.7324273\ttotal: 6.28s\tremaining: 1.26s\n",
      "250:\tlearn: 0.7297360\ttotal: 6.31s\tremaining: 1.23s\n",
      "251:\tlearn: 0.7276309\ttotal: 6.34s\tremaining: 1.21s\n",
      "252:\tlearn: 0.7240016\ttotal: 6.37s\tremaining: 1.18s\n",
      "253:\tlearn: 0.7228019\ttotal: 6.39s\tremaining: 1.16s\n",
      "254:\tlearn: 0.7212148\ttotal: 6.42s\tremaining: 1.13s\n",
      "255:\tlearn: 0.7196742\ttotal: 6.45s\tremaining: 1.11s\n",
      "256:\tlearn: 0.7192896\ttotal: 6.48s\tremaining: 1.08s\n",
      "257:\tlearn: 0.7179333\ttotal: 6.5s\tremaining: 1.06s\n",
      "258:\tlearn: 0.7164298\ttotal: 6.53s\tremaining: 1.03s\n",
      "259:\tlearn: 0.7143105\ttotal: 6.56s\tremaining: 1.01s\n",
      "260:\tlearn: 0.7114351\ttotal: 6.59s\tremaining: 985ms\n",
      "261:\tlearn: 0.7103388\ttotal: 6.62s\tremaining: 960ms\n",
      "262:\tlearn: 0.7080206\ttotal: 6.64s\tremaining: 935ms\n",
      "263:\tlearn: 0.7045824\ttotal: 6.67s\tremaining: 910ms\n",
      "264:\tlearn: 0.7029918\ttotal: 6.7s\tremaining: 885ms\n",
      "265:\tlearn: 0.7010292\ttotal: 6.73s\tremaining: 860ms\n",
      "266:\tlearn: 0.6994883\ttotal: 6.75s\tremaining: 835ms\n",
      "267:\tlearn: 0.6979174\ttotal: 6.78s\tremaining: 810ms\n",
      "268:\tlearn: 0.6964000\ttotal: 6.81s\tremaining: 785ms\n",
      "269:\tlearn: 0.6946344\ttotal: 6.84s\tremaining: 760ms\n",
      "270:\tlearn: 0.6941227\ttotal: 6.87s\tremaining: 735ms\n",
      "271:\tlearn: 0.6918421\ttotal: 6.9s\tremaining: 710ms\n",
      "272:\tlearn: 0.6892723\ttotal: 6.93s\tremaining: 685ms\n",
      "273:\tlearn: 0.6885049\ttotal: 6.96s\tremaining: 660ms\n",
      "274:\tlearn: 0.6879903\ttotal: 6.99s\tremaining: 635ms\n",
      "275:\tlearn: 0.6858093\ttotal: 7.01s\tremaining: 610ms\n",
      "276:\tlearn: 0.6853523\ttotal: 7.04s\tremaining: 585ms\n",
      "277:\tlearn: 0.6845040\ttotal: 7.07s\tremaining: 560ms\n",
      "278:\tlearn: 0.6834588\ttotal: 7.1s\tremaining: 535ms\n",
      "279:\tlearn: 0.6814175\ttotal: 7.13s\tremaining: 509ms\n",
      "280:\tlearn: 0.6795883\ttotal: 7.16s\tremaining: 484ms\n",
      "281:\tlearn: 0.6784800\ttotal: 7.18s\tremaining: 458ms\n",
      "282:\tlearn: 0.6759859\ttotal: 7.2s\tremaining: 432ms\n",
      "283:\tlearn: 0.6741425\ttotal: 7.21s\tremaining: 406ms\n",
      "284:\tlearn: 0.6738644\ttotal: 7.23s\tremaining: 381ms\n",
      "285:\tlearn: 0.6726771\ttotal: 7.25s\tremaining: 355ms\n",
      "286:\tlearn: 0.6720161\ttotal: 7.26s\tremaining: 329ms\n",
      "287:\tlearn: 0.6704259\ttotal: 7.28s\tremaining: 303ms\n",
      "288:\tlearn: 0.6695372\ttotal: 7.3s\tremaining: 278ms\n",
      "289:\tlearn: 0.6682127\ttotal: 7.32s\tremaining: 252ms\n",
      "290:\tlearn: 0.6667027\ttotal: 7.33s\tremaining: 227ms\n",
      "291:\tlearn: 0.6658941\ttotal: 7.35s\tremaining: 201ms\n",
      "292:\tlearn: 0.6648297\ttotal: 7.37s\tremaining: 176ms\n",
      "293:\tlearn: 0.6640524\ttotal: 7.38s\tremaining: 151ms\n",
      "294:\tlearn: 0.6630638\ttotal: 7.4s\tremaining: 125ms\n",
      "295:\tlearn: 0.6609934\ttotal: 7.42s\tremaining: 100ms\n",
      "296:\tlearn: 0.6596784\ttotal: 7.44s\tremaining: 75.1ms\n",
      "297:\tlearn: 0.6586496\ttotal: 7.46s\tremaining: 50ms\n",
      "298:\tlearn: 0.6576292\ttotal: 7.47s\tremaining: 25ms\n",
      "299:\tlearn: 0.6565734\ttotal: 7.49s\tremaining: 0us\n"
     ]
    }
   ],
   "source": [
    "from sklearn.datasets import make_regression\n",
    "from sklearn.multioutput import MultiOutputRegressor\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "from sklearn.metrics import explained_variance_score,r2_score\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.ensemble import ExtraTreesRegressor\n",
    "from catboost import CatBoostRegressor\n",
    "from sklearn.metrics import mean_absolute_percentage_error\n",
    "from sklearn.ensemble import VotingRegressor\n",
    "from sklearn.ensemble import VotingClassifier\n",
    "from sklearn.gaussian_process import GaussianProcessRegressor\n",
    "from sklearn.gaussian_process.kernels import DotProduct, WhiteKernel\n",
    "\n",
    "import xgboost  as xgb\n",
    "\n",
    "# Generate dataset\n",
    "# X, y = make_regression(n_samples=25000, n_features=3, n_targets=2, random_state=33)\n",
    "name_X = [\"Temperature_of_system1\",\t\"Temperature_of_system2\",\t\"Mineral_parameter1\",\t\"Mineral_parameter2\",\t\"Mineral_parameter3\",\t\"Mineral_parameter4\",\"Process_parameter3\",\"Process_parameter4\"]\n",
    "name_y = [\"index_A\",\"index_B\",\"index_C\",\"index_D\"]\n",
    "# name_X = [\"index_A\",\"index_B\",\"index_C\",\"index_D\",\t\"Mineral_parameter1\",\t\"Mineral_parameter2\",\t\"Mineral_parameter3\",\t\"Mineral_parameter4\"]\n",
    "# name_y = [\"Temperature_of_system1\",\t\"Temperature_of_system2\"]\n",
    "name = [\"index_A\",\"index_B\",\"index_C\",\"index_D\",\"Temperature_of_system1\",\t\"Temperature_of_system2\",\t\"Mineral_parameter1\",\t\"Mineral_parameter2\",\t\"Mineral_parameter3\",\t\"Mineral_parameter4\"]\n",
    "X = date[name_X]\n",
    "y = date[name_y]\n",
    "\n",
    "# Train/test split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=30)\n",
    "\n",
    "# Create the SVR regressor\n",
    "# svr = SVR(epsilon=0.01,C=1.0,kernel='poly')\n",
    "# svr = SVR(epsilon=0.2,kernel='rbf')\n",
    "# svr = RandomForestRegressor(max_depth=2, random_state=0)\n",
    "# # svr = ExtraTreesRegressor(n_estimators=100, random_state=0)\n",
    "# other_params = {'learning_rate': 0.1, 'n_estimators': 300, 'max_depth': 5, 'min_child_weight': 1, 'seed': 0, 'subsample': 0.8, 'colsample_bytree': 0.8, 'gamma': 0, 'reg_alpha': 0, 'reg_lambda': 1}\n",
    "# svr = xgb.XGBRegressor(objective='reg:squarederror',**other_params)\n",
    "\n",
    "# params = {\n",
    "#     'iterations':330,\n",
    "#     'learning_rate':0.1,\n",
    "#     'depth':10,\n",
    "#     'loss_function':'RMSE'\n",
    "\n",
    "# }\n",
    "\n",
    "\n",
    "# svr = CatBoostRegressor(**params)\n",
    "\n",
    "\n",
    "# Create the Multioutput Regressor\n",
    "# mor = MultiOutputRegressor(svr)\n",
    "\n",
    "\n",
    "# svr1 = SVR(epsilon=0.2,kernel='rbf')\n",
    "other_params = {'learning_rate': 0.1, 'n_estimators': 300, 'max_depth': 5, 'min_child_weight': 1, 'seed': 0, 'subsample': 0.8, 'colsample_bytree': 0.8, 'gamma': 0, 'reg_alpha': 0, 'reg_lambda': 1}\n",
    "svr2 = xgb.XGBRegressor(objective='reg:squarederror',**other_params)\n",
    "# svr3 = ExtraTreesRegressor(n_estimators=400, random_state=0)\n",
    "# kernel = DotProduct() + WhiteKernel()\n",
    "# svr4 = GaussianProcessRegressor(kernel=kernel,random_state=0)\n",
    "params = {\n",
    "    'iterations':300,\n",
    "    'learning_rate':0.1,\n",
    "    'depth':10,\n",
    "    'loss_function':'RMSE'\n",
    "\n",
    "}\n",
    "\n",
    "\n",
    "svr4 = CatBoostRegressor(**params)\n",
    "\n",
    "# models = list()\n",
    "# models.append(('xg', MultiOutputRegressor(svr2)))\n",
    "# models.append(('svr', MultiOutputRegressor(svr1)))\n",
    "# models.append(('RFR', MultiOutputRegressor(svr3)))\n",
    "\n",
    "models = list()\n",
    "models.append(('xg', svr2))\n",
    "# models.append(('svr', svr1))\n",
    "# models.append(('RFR', svr3))\n",
    "models.append(('CATBOOST', svr4))\n",
    "    # define the voting ensemble\n",
    "svr = VotingRegressor(estimators=models)\n",
    "\n",
    "mor = MultiOutputRegressor(svr)\n",
    "\n",
    "\n",
    "# # Train the regressor\n",
    "mor = mor.fit(X_train, y_train)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE:\n",
      "1  0.4723346929506919\n",
      "2  0.8289606288947904\n",
      "3  0.28831513595446234\n",
      "4  2.0489838374249176\n",
      "MAE:\n",
      "1  0.5055064155858265\n",
      "2  0.6568059898649369\n",
      "3  0.4135665874203815\n",
      "4  0.9580151293228684\n",
      "可解释的方差分数:\n",
      "1  0.4778962771998032\n",
      "2  0.5393486523104958\n",
      "3  0.8121783178100314\n",
      "4  0.8039681153773822\n",
      "r2_score:\n",
      "1  0.4700354613730384\n",
      "2  0.5362651784982123\n",
      "3  0.8108728729727801\n",
      "4  0.803788072638778\n",
      "mean_absolute_percentage_error:\n",
      "1  0.006383776911476698\n",
      "2  0.027776726419015237\n",
      "3  0.037487449788125876\n",
      "4  0.06517875501938798\n"
     ]
    }
   ],
   "source": [
    "# Generate predictions for testing data\n",
    "y_pred = mor.predict(X_test)\n",
    "# Evaluate the regressor\n",
    "y_test = y_test.values\n",
    "# y_test\n",
    "mse1 = mean_squared_error(y_test[:,0], y_pred[:,0])\n",
    "mse2 = mean_squared_error(y_test[:,1], y_pred[:,1])\n",
    "mse3 = mean_squared_error(y_test[:,2], y_pred[:,2])\n",
    "mse4 = mean_squared_error(y_test[:,3], y_pred[:,3])\n",
    "# print(f'MSE for first regressor: {mse_one} -second regressor: {mse_two}')\n",
    "print(\"MSE:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "print(\"2  \" + str(mse2))\n",
    "print(\"3  \" + str(mse3))\n",
    "print(\"4  \" + str(mse4))\n",
    "\n",
    "mse1 = mean_absolute_error(y_test[:,0], y_pred[:,0])\n",
    "mse2 = mean_absolute_error(y_test[:,1], y_pred[:,1])\n",
    "mse3 = mean_absolute_error(y_test[:,2], y_pred[:,2])\n",
    "mse4 = mean_absolute_error(y_test[:,3], y_pred[:,3])\n",
    "# print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "print(\"MAE:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "print(\"2  \" + str(mse2))\n",
    "print(\"3  \" + str(mse3))\n",
    "print(\"4  \" + str(mse4))\n",
    "\n",
    "mse1 = explained_variance_score(y_test[:,0], y_pred[:,0])\n",
    "mse2 = explained_variance_score(y_test[:,1], y_pred[:,1])\n",
    "mse3 = explained_variance_score(y_test[:,2], y_pred[:,2])\n",
    "mse4 = explained_variance_score(y_test[:,3], y_pred[:,3])\n",
    "# print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "print(\"可解释的方差分数:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "print(\"2  \" + str(mse2))\n",
    "print(\"3  \" + str(mse3))\n",
    "print(\"4  \" + str(mse4))\n",
    "\n",
    "\n",
    "mse1 = r2_score(y_test[:,0], y_pred[:,0])\n",
    "mse2 = r2_score(y_test[:,1], y_pred[:,1])\n",
    "mse3 = r2_score(y_test[:,2], y_pred[:,2])\n",
    "mse4 = r2_score(y_test[:,3], y_pred[:,3])\n",
    "# print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "print(\"r2_score:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "print(\"2  \" + str(mse2))\n",
    "print(\"3  \" + str(mse3))\n",
    "print(\"4  \" + str(mse4))\n",
    "\n",
    "mse1 = mean_absolute_percentage_error(y_test[:,0], y_pred[:,0])\n",
    "mse2 = mean_absolute_percentage_error(y_test[:,1], y_pred[:,1])\n",
    "mse3 = mean_absolute_percentage_error(y_test[:,2], y_pred[:,2])\n",
    "mse4 = mean_absolute_percentage_error(y_test[:,3], y_pred[:,3])\n",
    "# print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "print(\"mean_absolute_percentage_error:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "print(\"2  \" + str(mse2))\n",
    "print(\"3  \" + str(mse3))\n",
    "print(\"4  \" + str(mse4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[79.10238439, 23.53131854, 10.84906526, 11.81289273],\n",
       "       [78.15687691, 24.28452319, 11.13914712, 22.52460576],\n",
       "       [78.89772924, 24.46402818, 11.14143577, 12.67375379],\n",
       "       ...,\n",
       "       [79.18291998, 24.65843098, 10.50359605, 15.70681486],\n",
       "       [79.01325949, 23.00443084,  8.56137563, 11.08651401],\n",
       "       [78.7674884 , 25.45722689, 11.1482952 , 14.59279145]])"
      ]
     },
     "execution_count": 252,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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>Time</th>\n",
       "      <th>index_A</th>\n",
       "      <th>index_B</th>\n",
       "      <th>index_C</th>\n",
       "      <th>index_D</th>\n",
       "      <th>Temperature_of_system1</th>\n",
       "      <th>Temperature_of_system2</th>\n",
       "      <th>Mineral_parameter1</th>\n",
       "      <th>Mineral_parameter2</th>\n",
       "      <th>Mineral_parameter3</th>\n",
       "      <th>Mineral_parameter4</th>\n",
       "      <th>Process_parameter1</th>\n",
       "      <th>Process_parameter2</th>\n",
       "      <th>Process_parameter3</th>\n",
       "      <th>Process_parameter4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>319.02</td>\n",
       "      <td>151.29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>333.77</td>\n",
       "      <td>145.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>331.19</td>\n",
       "      <td>138.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>307.62</td>\n",
       "      <td>136.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>292.61</td>\n",
       "      <td>152.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>315.84</td>\n",
       "      <td>133.83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>324.77</td>\n",
       "      <td>132.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2022-02-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.40</td>\n",
       "      <td>665.04</td>\n",
       "      <td>52.88</td>\n",
       "      <td>91.27</td>\n",
       "      <td>47.22</td>\n",
       "      <td>22.26</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>300.09</td>\n",
       "      <td>137.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>315.21</td>\n",
       "      <td>138.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>305.73</td>\n",
       "      <td>145.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>268.69</td>\n",
       "      <td>168.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>281.49</td>\n",
       "      <td>145.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>277.82</td>\n",
       "      <td>143.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>285.18</td>\n",
       "      <td>148.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>226.16</td>\n",
       "      <td>177.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>2022-02-09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1010.32</td>\n",
       "      <td>874.47</td>\n",
       "      <td>54.44</td>\n",
       "      <td>92.12</td>\n",
       "      <td>48.85</td>\n",
       "      <td>21.83</td>\n",
       "      <td>1.25</td>\n",
       "      <td>3.09</td>\n",
       "      <td>290.75</td>\n",
       "      <td>164.45</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Time  index_A  index_B  index_C  index_D  Temperature_of_system1  \\\n",
       "0  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "1  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "2  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "3  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "4  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "5  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "6  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "7  2022-02-08      NaN      NaN      NaN      NaN                  341.40   \n",
       "8  2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "9  2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "10 2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "11 2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "12 2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "13 2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "14 2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "15 2022-02-09      NaN      NaN      NaN      NaN                 1010.32   \n",
       "\n",
       "    Temperature_of_system2  Mineral_parameter1  Mineral_parameter2  \\\n",
       "0                   665.04               52.88               91.27   \n",
       "1                   665.04               52.88               91.27   \n",
       "2                   665.04               52.88               91.27   \n",
       "3                   665.04               52.88               91.27   \n",
       "4                   665.04               52.88               91.27   \n",
       "5                   665.04               52.88               91.27   \n",
       "6                   665.04               52.88               91.27   \n",
       "7                   665.04               52.88               91.27   \n",
       "8                   874.47               54.44               92.12   \n",
       "9                   874.47               54.44               92.12   \n",
       "10                  874.47               54.44               92.12   \n",
       "11                  874.47               54.44               92.12   \n",
       "12                  874.47               54.44               92.12   \n",
       "13                  874.47               54.44               92.12   \n",
       "14                  874.47               54.44               92.12   \n",
       "15                  874.47               54.44               92.12   \n",
       "\n",
       "    Mineral_parameter3  Mineral_parameter4  Process_parameter1  \\\n",
       "0                47.22               22.26                1.25   \n",
       "1                47.22               22.26                1.25   \n",
       "2                47.22               22.26                1.25   \n",
       "3                47.22               22.26                1.25   \n",
       "4                47.22               22.26                1.25   \n",
       "5                47.22               22.26                1.25   \n",
       "6                47.22               22.26                1.25   \n",
       "7                47.22               22.26                1.25   \n",
       "8                48.85               21.83                1.25   \n",
       "9                48.85               21.83                1.25   \n",
       "10               48.85               21.83                1.25   \n",
       "11               48.85               21.83                1.25   \n",
       "12               48.85               21.83                1.25   \n",
       "13               48.85               21.83                1.25   \n",
       "14               48.85               21.83                1.25   \n",
       "15               48.85               21.83                1.25   \n",
       "\n",
       "    Process_parameter2  Process_parameter3  Process_parameter4  \n",
       "0                 3.09              319.02              151.29  \n",
       "1                 3.09              333.77              145.94  \n",
       "2                 3.09              331.19              138.85  \n",
       "3                 3.09              307.62              136.84  \n",
       "4                 3.09              292.61              152.06  \n",
       "5                 3.09              315.84              133.83  \n",
       "6                 3.09              324.77              132.33  \n",
       "7                 3.09              300.09              137.59  \n",
       "8                 3.09              315.21              138.35  \n",
       "9                 3.09              305.73              145.17  \n",
       "10                3.09              268.69              168.35  \n",
       "11                3.09              281.49              145.17  \n",
       "12                3.09              277.82              143.65  \n",
       "13                3.09              285.18              148.23  \n",
       "14                3.09              226.16              177.01  \n",
       "15                3.09              290.75              164.45  "
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr = pd.read_excel(\"preid.xlsx\")\n",
    "pr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[79.47439711, 24.27770824, 11.71189039, 15.03433334],\n",
       "       [79.22777676, 24.66348463, 11.54171116, 14.4551339 ],\n",
       "       [78.95797952, 24.76847917, 11.61367525, 15.92299974],\n",
       "       [79.12681098, 24.0363533 , 11.45236372, 16.96087889],\n",
       "       [79.72956477, 24.28899566, 11.39415784, 15.45609367],\n",
       "       [78.88882351, 24.04703547, 11.81438415, 17.70026795],\n",
       "       [78.83036817, 24.74563454, 11.80285874, 17.83484488],\n",
       "       [79.2012836 , 23.91116521, 11.40659266, 16.39372428],\n",
       "       [78.98294226, 24.41137266, 11.82464274, 15.41203913],\n",
       "       [79.11967191, 24.29373734, 11.69533041, 14.93372384],\n",
       "       [78.80457089, 24.49413731, 11.31602636, 14.74050283],\n",
       "       [79.30815878, 24.19679527, 11.54582875, 14.22787834],\n",
       "       [79.13364343, 24.01063086, 11.4344021 , 13.84025854],\n",
       "       [79.2129765 , 24.26539687, 11.57100003, 14.7542066 ],\n",
       "       [78.9194657 , 24.33188684, 11.49379293, 15.40727007],\n",
       "       [78.9661081 , 24.07773893, 11.52274753, 15.92187465]])"
      ]
     },
     "execution_count": 254,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pt_val = mor.predict(pr[name_X])\n",
    "pt_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[78.52907443, 23.53389076, 12.16676367, 17.90383046],\n",
       "       [78.70487837, 23.5807542 , 12.32632263, 17.94183124],\n",
       "       [78.83500266, 23.59347112, 12.3357402 , 17.94183124],\n",
       "       ...,\n",
       "       [79.18569753, 25.33627021, 11.2817251 , 11.98531295],\n",
       "       [79.18569753, 25.33627021, 11.2817251 , 11.98531295],\n",
       "       [79.20981635, 24.80135755, 11.12085537, 12.09515475]])"
      ]
     },
     "execution_count": 255,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### 开始对抗\n",
    "Virtual_val = mor.predict(X)\n",
    "Virtual_val\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>is_qualified</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>78.31</td>\n",
       "      <td>23.66</td>\n",
       "      <td>12.24</td>\n",
       "      <td>17.81</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>78.46</td>\n",
       "      <td>23.88</td>\n",
       "      <td>12.41</td>\n",
       "      <td>17.99</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>79.08</td>\n",
       "      <td>23.52</td>\n",
       "      <td>12.41</td>\n",
       "      <td>17.86</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>79.29</td>\n",
       "      <td>22.94</td>\n",
       "      <td>11.72</td>\n",
       "      <td>17.86</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.95</td>\n",
       "      <td>21.42</td>\n",
       "      <td>10.68</td>\n",
       "      <td>17.63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1648</th>\n",
       "      <td>79.82</td>\n",
       "      <td>23.84</td>\n",
       "      <td>11.03</td>\n",
       "      <td>13.52</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1649</th>\n",
       "      <td>78.98</td>\n",
       "      <td>25.36</td>\n",
       "      <td>11.37</td>\n",
       "      <td>12.85</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1650</th>\n",
       "      <td>78.86</td>\n",
       "      <td>25.40</td>\n",
       "      <td>11.37</td>\n",
       "      <td>11.42</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1651</th>\n",
       "      <td>79.10</td>\n",
       "      <td>25.58</td>\n",
       "      <td>11.37</td>\n",
       "      <td>11.55</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1652</th>\n",
       "      <td>79.32</td>\n",
       "      <td>24.82</td>\n",
       "      <td>11.03</td>\n",
       "      <td>11.55</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1653 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          A      B      C      D  is_qualified\n",
       "0     78.31  23.66  12.24  17.81             0\n",
       "1     78.46  23.88  12.41  17.99             0\n",
       "2     79.08  23.52  12.41  17.86             0\n",
       "3     79.29  22.94  11.72  17.86             0\n",
       "4     79.95  21.42  10.68  17.63             0\n",
       "...     ...    ...    ...    ...           ...\n",
       "1648  79.82  23.84  11.03  13.52             1\n",
       "1649  78.98  25.36  11.37  12.85             0\n",
       "1650  78.86  25.40  11.37  11.42             0\n",
       "1651  79.10  25.58  11.37  11.55             0\n",
       "1652  79.32  24.82  11.03  11.55             0\n",
       "\n",
       "[1653 rows x 5 columns]"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 构建一个分类模型吧 A,B,C,D -> 0/1\n",
    "# 这里做逻辑吧\n",
    "df = pd.DataFrame(Virtual_val)\n",
    "# df.to_excel(\"Virtual_val.xlsx\")\n",
    "df.columns = ['A','B','C','D']\n",
    "df.to_excel(\"Virtual_val.xlsx\")\n",
    "ral = pd.read_excel(\"./gan.xlsx\")\n",
    "ral = pd.DataFrame(ral)\n",
    "ral\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>is_qualified</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>78.529074</td>\n",
       "      <td>23.533891</td>\n",
       "      <td>12.166764</td>\n",
       "      <td>17.903830</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>78.704878</td>\n",
       "      <td>23.580754</td>\n",
       "      <td>12.326323</td>\n",
       "      <td>17.941831</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>78.835003</td>\n",
       "      <td>23.593471</td>\n",
       "      <td>12.335740</td>\n",
       "      <td>17.941831</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>79.749106</td>\n",
       "      <td>22.076600</td>\n",
       "      <td>10.877720</td>\n",
       "      <td>17.209333</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.762681</td>\n",
       "      <td>21.951956</td>\n",
       "      <td>10.863350</td>\n",
       "      <td>17.222794</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1648</th>\n",
       "      <td>79.526425</td>\n",
       "      <td>24.175692</td>\n",
       "      <td>11.102574</td>\n",
       "      <td>13.345164</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1649</th>\n",
       "      <td>79.340402</td>\n",
       "      <td>24.673273</td>\n",
       "      <td>11.206851</td>\n",
       "      <td>13.223382</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1650</th>\n",
       "      <td>79.185698</td>\n",
       "      <td>25.336270</td>\n",
       "      <td>11.281725</td>\n",
       "      <td>11.985313</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1651</th>\n",
       "      <td>79.185698</td>\n",
       "      <td>25.336270</td>\n",
       "      <td>11.281725</td>\n",
       "      <td>11.985313</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1652</th>\n",
       "      <td>79.209816</td>\n",
       "      <td>24.801358</td>\n",
       "      <td>11.120855</td>\n",
       "      <td>12.095155</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1653 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              A          B          C          D  is_qualified\n",
       "0     78.529074  23.533891  12.166764  17.903830             0\n",
       "1     78.704878  23.580754  12.326323  17.941831             0\n",
       "2     78.835003  23.593471  12.335740  17.941831             0\n",
       "3     79.749106  22.076600  10.877720  17.209333             0\n",
       "4     79.762681  21.951956  10.863350  17.222794             0\n",
       "...         ...        ...        ...        ...           ...\n",
       "1648  79.526425  24.175692  11.102574  13.345164             1\n",
       "1649  79.340402  24.673273  11.206851  13.223382             0\n",
       "1650  79.185698  25.336270  11.281725  11.985313             0\n",
       "1651  79.185698  25.336270  11.281725  11.985313             0\n",
       "1652  79.209816  24.801358  11.120855  12.095155             0\n",
       "\n",
       "[1653 rows x 5 columns]"
      ]
     },
     "execution_count": 257,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gan_data = df\n",
    "df[\"is_qualified\"] = ral[\"is_qualified\"]\n",
    "df = pd.DataFrame(df)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>is_qualified</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>78.529074</td>\n",
       "      <td>23.533891</td>\n",
       "      <td>12.166764</td>\n",
       "      <td>17.903830</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>78.704878</td>\n",
       "      <td>23.580754</td>\n",
       "      <td>12.326323</td>\n",
       "      <td>17.941831</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>78.835003</td>\n",
       "      <td>23.593471</td>\n",
       "      <td>12.335740</td>\n",
       "      <td>17.941831</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>79.749106</td>\n",
       "      <td>22.076600</td>\n",
       "      <td>10.877720</td>\n",
       "      <td>17.209333</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.762681</td>\n",
       "      <td>21.951956</td>\n",
       "      <td>10.863350</td>\n",
       "      <td>17.222794</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1648</th>\n",
       "      <td>79.820000</td>\n",
       "      <td>23.840000</td>\n",
       "      <td>11.030000</td>\n",
       "      <td>13.520000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1649</th>\n",
       "      <td>78.980000</td>\n",
       "      <td>25.360000</td>\n",
       "      <td>11.370000</td>\n",
       "      <td>12.850000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1650</th>\n",
       "      <td>78.860000</td>\n",
       "      <td>25.400000</td>\n",
       "      <td>11.370000</td>\n",
       "      <td>11.420000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1651</th>\n",
       "      <td>79.100000</td>\n",
       "      <td>25.580000</td>\n",
       "      <td>11.370000</td>\n",
       "      <td>11.550000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1652</th>\n",
       "      <td>79.320000</td>\n",
       "      <td>24.820000</td>\n",
       "      <td>11.030000</td>\n",
       "      <td>11.550000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3306 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              A          B          C          D  is_qualified\n",
       "0     78.529074  23.533891  12.166764  17.903830             0\n",
       "1     78.704878  23.580754  12.326323  17.941831             0\n",
       "2     78.835003  23.593471  12.335740  17.941831             0\n",
       "3     79.749106  22.076600  10.877720  17.209333             0\n",
       "4     79.762681  21.951956  10.863350  17.222794             0\n",
       "...         ...        ...        ...        ...           ...\n",
       "1648  79.820000  23.840000  11.030000  13.520000             1\n",
       "1649  78.980000  25.360000  11.370000  12.850000             0\n",
       "1650  78.860000  25.400000  11.370000  11.420000             0\n",
       "1651  79.100000  25.580000  11.370000  11.550000             0\n",
       "1652  79.320000  24.820000  11.030000  11.550000             0\n",
       "\n",
       "[3306 rows x 5 columns]"
      ]
     },
     "execution_count": 258,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 拼接\n",
    "gan_data = pd.concat([df,ral])\n",
    "gan_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7b00b71b42c04dbe984e2c1123f5f9b9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "MetricVisualizer(layout=Layout(align_self='stretch', height='500px'))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 0.6869115\ttest: 0.6872658\tbest: 0.6872658 (0)\ttotal: 1.97s\tremaining: 3m 15s\n",
      "1:\tlearn: 0.6809767\ttest: 0.6816941\tbest: 0.6816941 (1)\ttotal: 3.55s\tremaining: 2m 54s\n",
      "2:\tlearn: 0.6753148\ttest: 0.6765647\tbest: 0.6765647 (2)\ttotal: 5.3s\tremaining: 2m 51s\n",
      "3:\tlearn: 0.6692505\ttest: 0.6710677\tbest: 0.6710677 (3)\ttotal: 6.69s\tremaining: 2m 40s\n",
      "4:\tlearn: 0.6633071\ttest: 0.6654666\tbest: 0.6654666 (4)\ttotal: 7.91s\tremaining: 2m 30s\n",
      "5:\tlearn: 0.6575462\ttest: 0.6601447\tbest: 0.6601447 (5)\ttotal: 9.42s\tremaining: 2m 27s\n",
      "6:\tlearn: 0.6522680\ttest: 0.6552635\tbest: 0.6552635 (6)\ttotal: 10.6s\tremaining: 2m 20s\n",
      "7:\tlearn: 0.6464293\ttest: 0.6498304\tbest: 0.6498304 (7)\ttotal: 11.8s\tremaining: 2m 15s\n",
      "8:\tlearn: 0.6410022\ttest: 0.6447897\tbest: 0.6447897 (8)\ttotal: 13.2s\tremaining: 2m 13s\n",
      "9:\tlearn: 0.6354787\ttest: 0.6396832\tbest: 0.6396832 (9)\ttotal: 14.4s\tremaining: 2m 9s\n",
      "10:\tlearn: 0.6304396\ttest: 0.6350947\tbest: 0.6350947 (10)\ttotal: 15.6s\tremaining: 2m 6s\n",
      "11:\tlearn: 0.6252064\ttest: 0.6302481\tbest: 0.6302481 (11)\ttotal: 16.7s\tremaining: 2m 2s\n",
      "12:\tlearn: 0.6201411\ttest: 0.6255987\tbest: 0.6255987 (12)\ttotal: 17.9s\tremaining: 1m 59s\n",
      "13:\tlearn: 0.6149101\ttest: 0.6208024\tbest: 0.6208024 (13)\ttotal: 19.1s\tremaining: 1m 57s\n",
      "14:\tlearn: 0.6099210\ttest: 0.6161314\tbest: 0.6161314 (14)\ttotal: 20.2s\tremaining: 1m 54s\n",
      "15:\tlearn: 0.6047839\ttest: 0.6112703\tbest: 0.6112703 (15)\ttotal: 21.3s\tremaining: 1m 51s\n",
      "16:\tlearn: 0.5997086\ttest: 0.6066187\tbest: 0.6066187 (16)\ttotal: 22.3s\tremaining: 1m 49s\n",
      "17:\tlearn: 0.5947242\ttest: 0.6021227\tbest: 0.6021227 (17)\ttotal: 23.3s\tremaining: 1m 46s\n",
      "18:\tlearn: 0.5901249\ttest: 0.5979165\tbest: 0.5979165 (18)\ttotal: 24.4s\tremaining: 1m 44s\n",
      "19:\tlearn: 0.5855733\ttest: 0.5936642\tbest: 0.5936642 (19)\ttotal: 25.6s\tremaining: 1m 42s\n",
      "20:\tlearn: 0.5812974\ttest: 0.5898637\tbest: 0.5898637 (20)\ttotal: 26.6s\tremaining: 1m 40s\n",
      "21:\tlearn: 0.5766159\ttest: 0.5855985\tbest: 0.5855985 (21)\ttotal: 27.7s\tremaining: 1m 38s\n",
      "22:\tlearn: 0.5723265\ttest: 0.5817872\tbest: 0.5817872 (22)\ttotal: 28.7s\tremaining: 1m 36s\n",
      "23:\tlearn: 0.5678331\ttest: 0.5776886\tbest: 0.5776886 (23)\ttotal: 29.8s\tremaining: 1m 34s\n",
      "24:\tlearn: 0.5636475\ttest: 0.5740428\tbest: 0.5740428 (24)\ttotal: 30.7s\tremaining: 1m 32s\n",
      "25:\tlearn: 0.5598170\ttest: 0.5705052\tbest: 0.5705052 (25)\ttotal: 31.6s\tremaining: 1m 29s\n",
      "26:\tlearn: 0.5555783\ttest: 0.5667099\tbest: 0.5667099 (26)\ttotal: 32.4s\tremaining: 1m 27s\n",
      "27:\tlearn: 0.5516227\ttest: 0.5630333\tbest: 0.5630333 (27)\ttotal: 33.3s\tremaining: 1m 25s\n",
      "28:\tlearn: 0.5476579\ttest: 0.5593985\tbest: 0.5593985 (28)\ttotal: 34.2s\tremaining: 1m 23s\n",
      "29:\tlearn: 0.5437288\ttest: 0.5559027\tbest: 0.5559027 (29)\ttotal: 35.1s\tremaining: 1m 21s\n",
      "30:\tlearn: 0.5394313\ttest: 0.5520175\tbest: 0.5520175 (30)\ttotal: 35.3s\tremaining: 1m 18s\n",
      "31:\tlearn: 0.5353723\ttest: 0.5482182\tbest: 0.5482182 (31)\ttotal: 36.1s\tremaining: 1m 16s\n",
      "32:\tlearn: 0.5318345\ttest: 0.5450699\tbest: 0.5450699 (32)\ttotal: 37s\tremaining: 1m 15s\n",
      "33:\tlearn: 0.5282922\ttest: 0.5419152\tbest: 0.5419152 (33)\ttotal: 37.8s\tremaining: 1m 13s\n",
      "34:\tlearn: 0.5243551\ttest: 0.5384447\tbest: 0.5384447 (34)\ttotal: 38.6s\tremaining: 1m 11s\n",
      "35:\tlearn: 0.5206316\ttest: 0.5350869\tbest: 0.5350869 (35)\ttotal: 39.5s\tremaining: 1m 10s\n",
      "36:\tlearn: 0.5168004\ttest: 0.5316489\tbest: 0.5316489 (36)\ttotal: 40.3s\tremaining: 1m 8s\n",
      "37:\tlearn: 0.5135963\ttest: 0.5288078\tbest: 0.5288078 (37)\ttotal: 41.2s\tremaining: 1m 7s\n",
      "38:\tlearn: 0.5100409\ttest: 0.5256296\tbest: 0.5256296 (38)\ttotal: 42.2s\tremaining: 1m 6s\n",
      "39:\tlearn: 0.5066563\ttest: 0.5225571\tbest: 0.5225571 (39)\ttotal: 43s\tremaining: 1m 4s\n",
      "40:\tlearn: 0.5029757\ttest: 0.5191716\tbest: 0.5191716 (40)\ttotal: 43.9s\tremaining: 1m 3s\n",
      "41:\tlearn: 0.4995625\ttest: 0.5160728\tbest: 0.5160728 (41)\ttotal: 44.7s\tremaining: 1m 1s\n",
      "42:\tlearn: 0.4959485\ttest: 0.5125496\tbest: 0.5125496 (42)\ttotal: 45.7s\tremaining: 1m\n",
      "43:\tlearn: 0.4926030\ttest: 0.5095810\tbest: 0.5095810 (43)\ttotal: 46.6s\tremaining: 59.3s\n",
      "44:\tlearn: 0.4889644\ttest: 0.5062600\tbest: 0.5062600 (44)\ttotal: 47.5s\tremaining: 58s\n",
      "45:\tlearn: 0.4856550\ttest: 0.5031392\tbest: 0.5031392 (45)\ttotal: 48.4s\tremaining: 56.8s\n",
      "46:\tlearn: 0.4821638\ttest: 0.5000039\tbest: 0.5000039 (46)\ttotal: 49.3s\tremaining: 55.6s\n",
      "47:\tlearn: 0.4785681\ttest: 0.4967573\tbest: 0.4967573 (47)\ttotal: 49.7s\tremaining: 53.9s\n",
      "48:\tlearn: 0.4753882\ttest: 0.4939350\tbest: 0.4939350 (48)\ttotal: 50.6s\tremaining: 52.7s\n",
      "49:\tlearn: 0.4721711\ttest: 0.4910776\tbest: 0.4910776 (49)\ttotal: 51.4s\tremaining: 51.4s\n",
      "50:\tlearn: 0.4690704\ttest: 0.4881795\tbest: 0.4881795 (50)\ttotal: 52.3s\tremaining: 50.3s\n",
      "51:\tlearn: 0.4660914\ttest: 0.4855139\tbest: 0.4855139 (51)\ttotal: 53.2s\tremaining: 49.1s\n",
      "52:\tlearn: 0.4629397\ttest: 0.4826550\tbest: 0.4826550 (52)\ttotal: 54.1s\tremaining: 48s\n",
      "53:\tlearn: 0.4598625\ttest: 0.4797963\tbest: 0.4797963 (53)\ttotal: 54.9s\tremaining: 46.8s\n",
      "54:\tlearn: 0.4570486\ttest: 0.4773108\tbest: 0.4773108 (54)\ttotal: 55.8s\tremaining: 45.6s\n",
      "55:\tlearn: 0.4541214\ttest: 0.4748051\tbest: 0.4748051 (55)\ttotal: 56.6s\tremaining: 44.4s\n",
      "56:\tlearn: 0.4514412\ttest: 0.4724805\tbest: 0.4724805 (56)\ttotal: 57.4s\tremaining: 43.3s\n",
      "57:\tlearn: 0.4485952\ttest: 0.4699380\tbest: 0.4699380 (57)\ttotal: 58.2s\tremaining: 42.1s\n",
      "58:\tlearn: 0.4458563\ttest: 0.4674831\tbest: 0.4674831 (58)\ttotal: 59s\tremaining: 41s\n",
      "59:\tlearn: 0.4427050\ttest: 0.4645309\tbest: 0.4645309 (59)\ttotal: 59.8s\tremaining: 39.9s\n",
      "60:\tlearn: 0.4398643\ttest: 0.4619081\tbest: 0.4619081 (60)\ttotal: 1m\tremaining: 38.8s\n",
      "61:\tlearn: 0.4372153\ttest: 0.4595447\tbest: 0.4595447 (61)\ttotal: 1m 1s\tremaining: 37.7s\n",
      "62:\tlearn: 0.4345175\ttest: 0.4570562\tbest: 0.4570562 (62)\ttotal: 1m 2s\tremaining: 36.7s\n",
      "63:\tlearn: 0.4319105\ttest: 0.4546930\tbest: 0.4546930 (63)\ttotal: 1m 3s\tremaining: 35.6s\n",
      "64:\tlearn: 0.4292869\ttest: 0.4524067\tbest: 0.4524067 (64)\ttotal: 1m 4s\tremaining: 34.5s\n",
      "65:\tlearn: 0.4267896\ttest: 0.4501538\tbest: 0.4501538 (65)\ttotal: 1m 4s\tremaining: 33.5s\n",
      "66:\tlearn: 0.4242789\ttest: 0.4478897\tbest: 0.4478897 (66)\ttotal: 1m 5s\tremaining: 32.4s\n",
      "67:\tlearn: 0.4217160\ttest: 0.4454862\tbest: 0.4454862 (67)\ttotal: 1m 6s\tremaining: 31.3s\n",
      "68:\tlearn: 0.4189194\ttest: 0.4429957\tbest: 0.4429957 (68)\ttotal: 1m 7s\tremaining: 30.3s\n",
      "69:\tlearn: 0.4165521\ttest: 0.4407690\tbest: 0.4407690 (69)\ttotal: 1m 8s\tremaining: 29.2s\n",
      "70:\tlearn: 0.4142188\ttest: 0.4387029\tbest: 0.4387029 (70)\ttotal: 1m 8s\tremaining: 28.2s\n",
      "71:\tlearn: 0.4116437\ttest: 0.4362938\tbest: 0.4362938 (71)\ttotal: 1m 9s\tremaining: 27.1s\n",
      "72:\tlearn: 0.4092176\ttest: 0.4341052\tbest: 0.4341052 (72)\ttotal: 1m 10s\tremaining: 26.1s\n",
      "73:\tlearn: 0.4068831\ttest: 0.4319929\tbest: 0.4319929 (73)\ttotal: 1m 11s\tremaining: 25.1s\n",
      "74:\tlearn: 0.4044595\ttest: 0.4298948\tbest: 0.4298948 (74)\ttotal: 1m 12s\tremaining: 24.1s\n",
      "75:\tlearn: 0.4018594\ttest: 0.4274179\tbest: 0.4274179 (75)\ttotal: 1m 12s\tremaining: 22.9s\n",
      "76:\tlearn: 0.3995753\ttest: 0.4253017\tbest: 0.4253017 (76)\ttotal: 1m 13s\tremaining: 21.9s\n",
      "77:\tlearn: 0.3975949\ttest: 0.4236432\tbest: 0.4236432 (77)\ttotal: 1m 14s\tremaining: 20.9s\n",
      "78:\tlearn: 0.3954559\ttest: 0.4217145\tbest: 0.4217145 (78)\ttotal: 1m 14s\tremaining: 19.9s\n",
      "79:\tlearn: 0.3932946\ttest: 0.4197591\tbest: 0.4197591 (79)\ttotal: 1m 15s\tremaining: 18.9s\n",
      "80:\tlearn: 0.3910509\ttest: 0.4178075\tbest: 0.4178075 (80)\ttotal: 1m 16s\tremaining: 17.9s\n",
      "81:\tlearn: 0.3888328\ttest: 0.4159132\tbest: 0.4159132 (81)\ttotal: 1m 17s\tremaining: 16.9s\n",
      "82:\tlearn: 0.3867010\ttest: 0.4139917\tbest: 0.4139917 (82)\ttotal: 1m 17s\tremaining: 16s\n",
      "83:\tlearn: 0.3844654\ttest: 0.4118893\tbest: 0.4118893 (83)\ttotal: 1m 18s\tremaining: 15s\n",
      "84:\tlearn: 0.3823673\ttest: 0.4102251\tbest: 0.4102251 (84)\ttotal: 1m 19s\tremaining: 14s\n",
      "85:\tlearn: 0.3801895\ttest: 0.4083032\tbest: 0.4083032 (85)\ttotal: 1m 20s\tremaining: 13.1s\n",
      "86:\tlearn: 0.3780199\ttest: 0.4063127\tbest: 0.4063127 (86)\ttotal: 1m 21s\tremaining: 12.1s\n",
      "87:\tlearn: 0.3762000\ttest: 0.4047969\tbest: 0.4047969 (87)\ttotal: 1m 21s\tremaining: 11.2s\n",
      "88:\tlearn: 0.3741914\ttest: 0.4030959\tbest: 0.4030959 (88)\ttotal: 1m 22s\tremaining: 10.2s\n",
      "89:\tlearn: 0.3723898\ttest: 0.4015908\tbest: 0.4015908 (89)\ttotal: 1m 23s\tremaining: 9.27s\n",
      "90:\tlearn: 0.3702596\ttest: 0.3996258\tbest: 0.3996258 (90)\ttotal: 1m 24s\tremaining: 8.33s\n",
      "91:\tlearn: 0.3685128\ttest: 0.3981395\tbest: 0.3981395 (91)\ttotal: 1m 24s\tremaining: 7.39s\n",
      "92:\tlearn: 0.3664985\ttest: 0.3962599\tbest: 0.3962599 (92)\ttotal: 1m 25s\tremaining: 6.46s\n",
      "93:\tlearn: 0.3646945\ttest: 0.3948036\tbest: 0.3948036 (93)\ttotal: 1m 26s\tremaining: 5.53s\n",
      "94:\tlearn: 0.3626270\ttest: 0.3928865\tbest: 0.3928865 (94)\ttotal: 1m 27s\tremaining: 4.6s\n",
      "95:\tlearn: 0.3606934\ttest: 0.3911498\tbest: 0.3911498 (95)\ttotal: 1m 28s\tremaining: 3.67s\n",
      "96:\tlearn: 0.3587495\ttest: 0.3894611\tbest: 0.3894611 (96)\ttotal: 1m 28s\tremaining: 2.75s\n",
      "97:\tlearn: 0.3569909\ttest: 0.3880753\tbest: 0.3880753 (97)\ttotal: 1m 29s\tremaining: 1.83s\n",
      "98:\tlearn: 0.3550727\ttest: 0.3864498\tbest: 0.3864498 (98)\ttotal: 1m 30s\tremaining: 914ms\n",
      "99:\tlearn: 0.3530484\ttest: 0.3846233\tbest: 0.3846233 (99)\ttotal: 1m 31s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.3846232625\n",
      "bestIteration = 99\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1]], dtype=int64)"
      ]
     },
     "execution_count": 276,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.isotonic import IsotonicRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import balanced_accuracy_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import zero_one_loss\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from catboost import CatBoostClassifier\n",
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "name_X1 = ['A','B','C','D']\n",
    "name_y1 = \"is_qualified\"\n",
    "# name_X = [\"index_A\",\"index_B\",\"index_C\",\"index_D\",\t\"Mineral_parameter1\",\t\"Mineral_parameter2\",\t\"Mineral_parameter3\",\t\"Mineral_parameter4\"]\n",
    "# name_y = [\"Temperature_of_system1\",\t\"Temperature_of_system2\"]\n",
    "# name = [\"index_A\",\"index_B\",\"index_C\",\"index_D\",\"Temperature_of_system1\",\t\"Temperature_of_system2\",\t\"Mineral_parameter1\",\t\"Mineral_parameter2\",\t\"Mineral_parameter3\",\t\"Mineral_parameter4\"]\n",
    "X1 = gan_data[name_X1]\n",
    "y1 = gan_data[name_y1]\n",
    "X1 = X1.values\n",
    "y1 = y1.values\n",
    "\n",
    "# Train/test split\n",
    "X_train1, X_test1, y_train1, y_test1 = train_test_split(X1, y1, test_size=0.3, random_state=35)\n",
    "\n",
    "# clf = RandomForestClassifier(n_estimators=10)\n",
    "# clf = GaussianNB()\n",
    "# clf = SVC()\n",
    "# clf = AdaBoostClassifier(n_estimators=100, random_state=0)\n",
    "# clf.fit(X_train1, y_train1)\n",
    "\n",
    "# categorical_features_indices = np.where(X_train1.dtypes != np.float)[0]\n",
    "clf = CatBoostClassifier(iterations=100, \n",
    "                           depth=15,\n",
    "                           learning_rate=0.01,\n",
    "                           loss_function='MultiClass',\n",
    "                           logging_level='Verbose')\n",
    "clf.fit(X_train1,y_train1,eval_set=(X_test1, y_test1),plot=True)\n",
    "\n",
    "# Generate predictions for testing data\n",
    "y_pred1 = clf.predict(X_test1)\n",
    "# y_test1 = y_test1.values\n",
    "# Evaluate the regressor\n",
    "y_pred1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "平衡精度:\n",
      "1  0.9178974351684696\n",
      "混淆矩阵:\n",
      "1  [[646  37]\n",
      " [ 34 275]]\n",
      "根据预测分数计算平均精度:\n",
      "1  0.818700797149748\n",
      "精度分类得分。:\n",
      "1  0.9284274193548387\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import average_precision_score\n",
    "\n",
    "\n",
    "# y_test\n",
    "mse1 = balanced_accuracy_score(y_test1, y_pred1)\n",
    "# print(f'MSE for first regressor: {mse_one} -second regressor: {mse_two}')\n",
    "print(\"平衡精度:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "\n",
    "mse1 = confusion_matrix(y_test1, y_pred1)\n",
    "# print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "print(\"混淆矩阵:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "\n",
    "mse1 = average_precision_score(y_test1, y_pred1)\n",
    "\n",
    "# print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "print(\"根据预测分数计算平均精度:\")\n",
    "print(\"1  \" + str(mse1))\n",
    "\n",
    "\n",
    "\n",
    "# precision, recall, thresholds = zero_one_loss(y_test1, y_pred1)\n",
    "\n",
    "# # print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "# print(\"0-1精度损失:\")\n",
    "# print(\"precision\")\n",
    "# print(precision)\n",
    "# print(\"recall\")\n",
    "# print(recall)\n",
    "# print(\"thresholds\")\n",
    "# print(thresholds)\n",
    "\n",
    "\n",
    "mse1 = accuracy_score(y_test1, y_pred1)\n",
    "\n",
    "# print(f'MAE for first regressor: {mae_one} - second regressor: {mae_two}')\n",
    "print(\"精度分类得分。:\")\n",
    "print(\"1  \" + str(mse1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>79.474397</td>\n",
       "      <td>24.277708</td>\n",
       "      <td>11.711890</td>\n",
       "      <td>15.034333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>79.227777</td>\n",
       "      <td>24.663485</td>\n",
       "      <td>11.541711</td>\n",
       "      <td>14.455134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>78.957980</td>\n",
       "      <td>24.768479</td>\n",
       "      <td>11.613675</td>\n",
       "      <td>15.923000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>79.126811</td>\n",
       "      <td>24.036353</td>\n",
       "      <td>11.452364</td>\n",
       "      <td>16.960879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>79.729565</td>\n",
       "      <td>24.288996</td>\n",
       "      <td>11.394158</td>\n",
       "      <td>15.456094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>78.888824</td>\n",
       "      <td>24.047035</td>\n",
       "      <td>11.814384</td>\n",
       "      <td>17.700268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>78.830368</td>\n",
       "      <td>24.745635</td>\n",
       "      <td>11.802859</td>\n",
       "      <td>17.834845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>79.201284</td>\n",
       "      <td>23.911165</td>\n",
       "      <td>11.406593</td>\n",
       "      <td>16.393724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>78.982942</td>\n",
       "      <td>24.411373</td>\n",
       "      <td>11.824643</td>\n",
       "      <td>15.412039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>79.119672</td>\n",
       "      <td>24.293737</td>\n",
       "      <td>11.695330</td>\n",
       "      <td>14.933724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>78.804571</td>\n",
       "      <td>24.494137</td>\n",
       "      <td>11.316026</td>\n",
       "      <td>14.740503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>79.308159</td>\n",
       "      <td>24.196795</td>\n",
       "      <td>11.545829</td>\n",
       "      <td>14.227878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>79.133643</td>\n",
       "      <td>24.010631</td>\n",
       "      <td>11.434402</td>\n",
       "      <td>13.840259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>79.212976</td>\n",
       "      <td>24.265397</td>\n",
       "      <td>11.571000</td>\n",
       "      <td>14.754207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>78.919466</td>\n",
       "      <td>24.331887</td>\n",
       "      <td>11.493793</td>\n",
       "      <td>15.407270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>78.966108</td>\n",
       "      <td>24.077739</td>\n",
       "      <td>11.522748</td>\n",
       "      <td>15.921875</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            A          B          C          D\n",
       "0   79.474397  24.277708  11.711890  15.034333\n",
       "1   79.227777  24.663485  11.541711  14.455134\n",
       "2   78.957980  24.768479  11.613675  15.923000\n",
       "3   79.126811  24.036353  11.452364  16.960879\n",
       "4   79.729565  24.288996  11.394158  15.456094\n",
       "5   78.888824  24.047035  11.814384  17.700268\n",
       "6   78.830368  24.745635  11.802859  17.834845\n",
       "7   79.201284  23.911165  11.406593  16.393724\n",
       "8   78.982942  24.411373  11.824643  15.412039\n",
       "9   79.119672  24.293737  11.695330  14.933724\n",
       "10  78.804571  24.494137  11.316026  14.740503\n",
       "11  79.308159  24.196795  11.545829  14.227878\n",
       "12  79.133643  24.010631  11.434402  13.840259\n",
       "13  79.212976  24.265397  11.571000  14.754207\n",
       "14  78.919466  24.331887  11.493793  15.407270\n",
       "15  78.966108  24.077739  11.522748  15.921875"
      ]
     },
     "execution_count": 278,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_val = pd.DataFrame(pt_val)\n",
    "y_pred_val.columns = ['A','B','C','D']\n",
    "y_pred_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0],\n",
       "       [1],\n",
       "       [0],\n",
       "       [0],\n",
       "       [0]], dtype=int64)"
      ]
     },
     "execution_count": 279,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result = clf.predict(y_pred_val)\n",
    "result"
   ]
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}