TimeSeries_ML.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Project description\n",
    "\n",
    "Sweet Lift Taxi company has collected historical data on taxi orders at airports. To attract more drivers during peak hours, we need to predict the amount of taxi orders for the next hour. Build a model for such a prediction.\n",
    "\n",
    "The RMSE metric on the test set should not be more than 48.\n",
    "\n",
    "### Project instructions\n",
    "\n",
    "1. Download the data and resample it by one hour.\n",
    "2. Analyze the data.\n",
    "3. Train different models with different hyperparameters. The test sample should be 10% of the initial dataset. \n",
    "4. Test the data using the test sample and provide a conclusion.\n",
    "\n",
    "### Data description\n",
    "\n",
    "The data is stored in file `taxi.csv`. The number of orders is in the '*num_orders*' column."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from catboost import CatBoostRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('/datasets/taxi.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>num_orders</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>2018-03-01 00:00:00</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>2018-03-01 00:10:00</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2018-03-01 00:20:00</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>2018-03-01 00:30:00</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>2018-03-01 00:40:00</td>\n",
       "      <td>32</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              datetime  num_orders\n",
       "0  2018-03-01 00:00:00           9\n",
       "1  2018-03-01 00:10:00          14\n",
       "2  2018-03-01 00:20:00          28\n",
       "3  2018-03-01 00:30:00          20\n",
       "4  2018-03-01 00:40:00          32"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 26496 entries, 0 to 26495\n",
      "Data columns (total 2 columns):\n",
      "datetime      26496 non-null object\n",
      "num_orders    26496 non-null int64\n",
      "dtypes: int64(1), object(1)\n",
      "memory usage: 414.1+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['datetime'] = pd.to_datetime(df['datetime'], format='%Y-%m-%d %H:%M:%S')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.set_index('datetime')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.resample('1H').sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "decomposed = seasonal_decompose(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1 = df.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x1008 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 14))\n",
    "plt.subplot(311)\n",
    "decomposed.trend.plot(ax=plt.gca())\n",
    "plt.title('Trend')\n",
    "plt.ylabel('Number of Orders')\n",
    "plt.tight_layout(pad=0.4, w_pad=0.5, h_pad=0.8)\n",
    "decomposed.seasonal['2018-03-01 00:00:00': '2018-03-03 00:00:00'].plot()\n",
    "plt.title('DAY')\n",
    "plt.ylabel('Number of Orders')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we can see that there are more orders during the summer and also If we look at the daily breakdown we can see a trand that there are more orders during the night, which peaks at midnight."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1['mean'] = df1['num_orders'].rolling(15).mean()\n",
    "df1['std'] = df1['num_orders'].rolling(15).std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f746f6d7810>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df1.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "He we can see that the time series are not stationary, because mean and standart deviation change."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_features(data, lag, rolling_mean_size):\n",
    "    data['year'] = data.index.year\n",
    "    data['month'] = data.index.month\n",
    "    data['day'] = data.index.day\n",
    "    data['dayofweek'] = data.index.dayofweek\n",
    "    data['lag_{}'.format(lag)] = data['num_orders'].shift(lag)  \n",
    "    data['rolling_mean'] = data['num_orders'].shift().rolling(rolling_mean_size).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "make_features(df, 24, 6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "train, test = train_test_split(df, shuffle = False, test_size = 0.1)\n",
    "train = train.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_x = train.drop('num_orders', axis = 1)\n",
    "train_y = train['num_orders']\n",
    "test_x = test.drop('num_orders', axis = 1)\n",
    "test_y = test['num_orders']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for the testing set: 49.733198597464636\n"
     ]
    }
   ],
   "source": [
    "lr = LinearRegression()\n",
    "lr.fit(train_x, train_y) \n",
    "lr_test_pred = lr.predict(test_x)\n",
    "print(\"RMSE for the testing set:\", mean_squared_error(test_y, lr_test_pred) ** 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for the testing set: 46.23719218525786\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "model = RandomForestRegressor()\n",
    "model.fit(train_x, train_y)\n",
    "pred = model.predict(test_x)\n",
    "print(\"RMSE for the testing set:\", mean_squared_error(test_y, pred) ** 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = CatBoostRegressor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 38.1484442\ttotal: 52.1ms\tremaining: 52.1s\n",
      "1:\tlearn: 37.6760579\ttotal: 57.8ms\tremaining: 28.8s\n",
      "2:\tlearn: 37.1899882\ttotal: 62.2ms\tremaining: 20.7s\n",
      "3:\tlearn: 36.7349048\ttotal: 88.2ms\tremaining: 22s\n",
      "4:\tlearn: 36.2861411\ttotal: 94.9ms\tremaining: 18.9s\n",
      "5:\tlearn: 35.8405745\ttotal: 102ms\tremaining: 16.9s\n",
      "6:\tlearn: 35.4246274\ttotal: 190ms\tremaining: 26.9s\n",
      "7:\tlearn: 35.0230022\ttotal: 199ms\tremaining: 24.6s\n",
      "8:\tlearn: 34.6388315\ttotal: 282ms\tremaining: 31.1s\n",
      "9:\tlearn: 34.2658737\ttotal: 290ms\tremaining: 28.7s\n",
      "10:\tlearn: 33.9352096\ttotal: 296ms\tremaining: 26.6s\n",
      "11:\tlearn: 33.6058237\ttotal: 384ms\tremaining: 31.6s\n",
      "12:\tlearn: 33.2854179\ttotal: 391ms\tremaining: 29.7s\n",
      "13:\tlearn: 33.0311488\ttotal: 479ms\tremaining: 33.7s\n",
      "14:\tlearn: 32.7317796\ttotal: 486ms\tremaining: 31.9s\n",
      "15:\tlearn: 32.4413858\ttotal: 491ms\tremaining: 30.2s\n",
      "16:\tlearn: 32.1560038\ttotal: 576ms\tremaining: 33.3s\n",
      "17:\tlearn: 31.8943890\ttotal: 584ms\tremaining: 31.9s\n",
      "18:\tlearn: 31.6409002\ttotal: 593ms\tremaining: 30.6s\n",
      "19:\tlearn: 31.4166977\ttotal: 681ms\tremaining: 33.4s\n",
      "20:\tlearn: 31.1943221\ttotal: 692ms\tremaining: 32.3s\n",
      "21:\tlearn: 30.9743273\ttotal: 777ms\tremaining: 34.5s\n",
      "22:\tlearn: 30.7784206\ttotal: 783ms\tremaining: 33.3s\n",
      "23:\tlearn: 30.5844734\ttotal: 788ms\tremaining: 32s\n",
      "24:\tlearn: 30.3935133\ttotal: 793ms\tremaining: 30.9s\n",
      "25:\tlearn: 30.2080288\ttotal: 797ms\tremaining: 29.9s\n",
      "26:\tlearn: 30.0331450\ttotal: 801ms\tremaining: 28.9s\n",
      "27:\tlearn: 29.8662464\ttotal: 873ms\tremaining: 30.3s\n",
      "28:\tlearn: 29.7042175\ttotal: 878ms\tremaining: 29.4s\n",
      "29:\tlearn: 29.5439922\ttotal: 882ms\tremaining: 28.5s\n",
      "30:\tlearn: 29.3786188\ttotal: 888ms\tremaining: 27.7s\n",
      "31:\tlearn: 29.2369566\ttotal: 977ms\tremaining: 29.5s\n",
      "32:\tlearn: 29.1055218\ttotal: 985ms\tremaining: 28.9s\n",
      "33:\tlearn: 28.9759336\ttotal: 1.07s\tremaining: 30.3s\n",
      "34:\tlearn: 28.8555467\ttotal: 1.07s\tremaining: 29.6s\n",
      "35:\tlearn: 28.7318739\ttotal: 1.08s\tremaining: 28.9s\n",
      "36:\tlearn: 28.6220673\ttotal: 1.09s\tremaining: 28.3s\n",
      "37:\tlearn: 28.5292817\ttotal: 1.17s\tremaining: 29.6s\n",
      "38:\tlearn: 28.4258848\ttotal: 1.18s\tremaining: 29.1s\n",
      "39:\tlearn: 28.3194946\ttotal: 1.26s\tremaining: 30.3s\n",
      "40:\tlearn: 28.2234780\ttotal: 1.27s\tremaining: 29.7s\n",
      "41:\tlearn: 28.1331142\ttotal: 1.27s\tremaining: 29.1s\n",
      "42:\tlearn: 28.0473883\ttotal: 1.28s\tremaining: 28.5s\n",
      "43:\tlearn: 27.9555615\ttotal: 1.36s\tremaining: 29.6s\n",
      "44:\tlearn: 27.8730654\ttotal: 1.37s\tremaining: 29.1s\n",
      "45:\tlearn: 27.7948228\ttotal: 1.38s\tremaining: 28.6s\n",
      "46:\tlearn: 27.7249827\ttotal: 1.47s\tremaining: 29.8s\n",
      "47:\tlearn: 27.6521604\ttotal: 1.48s\tremaining: 29.3s\n",
      "48:\tlearn: 27.5882115\ttotal: 1.49s\tremaining: 28.9s\n",
      "49:\tlearn: 27.5149988\ttotal: 1.56s\tremaining: 29.6s\n",
      "50:\tlearn: 27.4473899\ttotal: 1.57s\tremaining: 29.2s\n",
      "51:\tlearn: 27.3888268\ttotal: 1.57s\tremaining: 28.7s\n",
      "52:\tlearn: 27.3286568\ttotal: 1.58s\tremaining: 28.3s\n",
      "53:\tlearn: 27.2692996\ttotal: 1.66s\tremaining: 29s\n",
      "54:\tlearn: 27.2107377\ttotal: 1.67s\tremaining: 28.7s\n",
      "55:\tlearn: 27.1573530\ttotal: 1.75s\tremaining: 29.5s\n",
      "56:\tlearn: 27.1042692\ttotal: 1.76s\tremaining: 29.1s\n",
      "57:\tlearn: 27.0536974\ttotal: 1.76s\tremaining: 28.7s\n",
      "58:\tlearn: 27.0000156\ttotal: 1.85s\tremaining: 29.6s\n",
      "59:\tlearn: 26.9552271\ttotal: 1.86s\tremaining: 29.2s\n",
      "60:\tlearn: 26.9109788\ttotal: 1.87s\tremaining: 28.7s\n",
      "61:\tlearn: 26.8668726\ttotal: 1.88s\tremaining: 28.5s\n",
      "62:\tlearn: 26.8284912\ttotal: 1.95s\tremaining: 29s\n",
      "63:\tlearn: 26.7889599\ttotal: 1.95s\tremaining: 28.6s\n",
      "64:\tlearn: 26.7534829\ttotal: 1.96s\tremaining: 28.2s\n",
      "65:\tlearn: 26.7141261\ttotal: 1.97s\tremaining: 27.9s\n",
      "66:\tlearn: 26.6799289\ttotal: 2.05s\tremaining: 28.5s\n",
      "67:\tlearn: 26.6383726\ttotal: 2.06s\tremaining: 28.2s\n",
      "68:\tlearn: 26.6041983\ttotal: 2.07s\tremaining: 27.9s\n",
      "69:\tlearn: 26.5777548\ttotal: 2.07s\tremaining: 27.5s\n",
      "70:\tlearn: 26.5423219\ttotal: 2.15s\tremaining: 28.2s\n",
      "71:\tlearn: 26.5175871\ttotal: 2.17s\tremaining: 28s\n",
      "72:\tlearn: 26.4925416\ttotal: 2.17s\tremaining: 27.6s\n",
      "73:\tlearn: 26.4698308\ttotal: 2.25s\tremaining: 28.1s\n",
      "74:\tlearn: 26.4457236\ttotal: 2.26s\tremaining: 27.9s\n",
      "75:\tlearn: 26.4173662\ttotal: 2.27s\tremaining: 27.6s\n",
      "76:\tlearn: 26.3890380\ttotal: 2.34s\tremaining: 28.1s\n",
      "77:\tlearn: 26.3653726\ttotal: 2.35s\tremaining: 27.8s\n",
      "78:\tlearn: 26.3364787\ttotal: 2.36s\tremaining: 27.5s\n",
      "79:\tlearn: 26.3087568\ttotal: 2.37s\tremaining: 27.2s\n",
      "80:\tlearn: 26.2851768\ttotal: 2.44s\tremaining: 27.7s\n",
      "81:\tlearn: 26.2639674\ttotal: 2.45s\tremaining: 27.4s\n",
      "82:\tlearn: 26.2427490\ttotal: 2.46s\tremaining: 27.1s\n",
      "83:\tlearn: 26.2237713\ttotal: 2.53s\tremaining: 27.6s\n",
      "84:\tlearn: 26.1974466\ttotal: 2.54s\tremaining: 27.3s\n",
      "85:\tlearn: 26.1733114\ttotal: 2.54s\tremaining: 27.1s\n",
      "86:\tlearn: 26.1528710\ttotal: 2.63s\tremaining: 27.6s\n",
      "87:\tlearn: 26.1384166\ttotal: 2.64s\tremaining: 27.4s\n",
      "88:\tlearn: 26.1180084\ttotal: 2.65s\tremaining: 27.1s\n",
      "89:\tlearn: 26.0982486\ttotal: 2.73s\tremaining: 27.7s\n",
      "90:\tlearn: 26.0761787\ttotal: 2.74s\tremaining: 27.4s\n",
      "91:\tlearn: 26.0602633\ttotal: 2.83s\tremaining: 27.9s\n",
      "92:\tlearn: 26.0380461\ttotal: 2.83s\tremaining: 27.6s\n",
      "93:\tlearn: 26.0214693\ttotal: 2.84s\tremaining: 27.4s\n",
      "94:\tlearn: 26.0053651\ttotal: 2.93s\tremaining: 27.9s\n",
      "95:\tlearn: 25.9805926\ttotal: 2.94s\tremaining: 27.7s\n",
      "96:\tlearn: 25.9604232\ttotal: 2.95s\tremaining: 27.4s\n",
      "97:\tlearn: 25.9390706\ttotal: 2.95s\tremaining: 27.2s\n",
      "98:\tlearn: 25.9280333\ttotal: 2.96s\tremaining: 27s\n",
      "99:\tlearn: 25.9162861\ttotal: 2.97s\tremaining: 26.8s\n",
      "100:\tlearn: 25.9144502\ttotal: 2.98s\tremaining: 26.5s\n",
      "101:\tlearn: 25.9014089\ttotal: 3.06s\tremaining: 26.9s\n",
      "102:\tlearn: 25.8914503\ttotal: 3.07s\tremaining: 26.7s\n",
      "103:\tlearn: 25.8728702\ttotal: 3.08s\tremaining: 26.5s\n",
      "104:\tlearn: 25.8635552\ttotal: 3.15s\tremaining: 26.9s\n",
      "105:\tlearn: 25.8533349\ttotal: 3.16s\tremaining: 26.6s\n",
      "106:\tlearn: 25.8418560\ttotal: 3.16s\tremaining: 26.4s\n",
      "107:\tlearn: 25.8262857\ttotal: 3.25s\tremaining: 26.8s\n",
      "108:\tlearn: 25.8178230\ttotal: 3.26s\tremaining: 26.6s\n",
      "109:\tlearn: 25.7985372\ttotal: 3.27s\tremaining: 26.4s\n",
      "110:\tlearn: 25.7836105\ttotal: 3.27s\tremaining: 26.2s\n",
      "111:\tlearn: 25.7688317\ttotal: 3.35s\tremaining: 26.5s\n",
      "112:\tlearn: 25.7500965\ttotal: 3.35s\tremaining: 26.3s\n",
      "113:\tlearn: 25.7389037\ttotal: 3.35s\tremaining: 26.1s\n",
      "114:\tlearn: 25.7235221\ttotal: 3.36s\tremaining: 25.9s\n",
      "115:\tlearn: 25.7130715\ttotal: 3.44s\tremaining: 26.3s\n",
      "116:\tlearn: 25.7005782\ttotal: 3.45s\tremaining: 26.1s\n",
      "117:\tlearn: 25.6897761\ttotal: 3.46s\tremaining: 25.9s\n",
      "118:\tlearn: 25.6709567\ttotal: 3.54s\tremaining: 26.2s\n",
      "119:\tlearn: 25.6574028\ttotal: 3.55s\tremaining: 26s\n",
      "120:\tlearn: 25.6464160\ttotal: 3.56s\tremaining: 25.8s\n",
      "121:\tlearn: 25.6320756\ttotal: 3.64s\tremaining: 26.2s\n",
      "122:\tlearn: 25.6217845\ttotal: 3.65s\tremaining: 26s\n",
      "123:\tlearn: 25.6037886\ttotal: 3.74s\tremaining: 26.4s\n",
      "124:\tlearn: 25.5912853\ttotal: 3.75s\tremaining: 26.2s\n",
      "125:\tlearn: 25.5854254\ttotal: 3.75s\tremaining: 26s\n",
      "126:\tlearn: 25.5726147\ttotal: 3.83s\tremaining: 26.4s\n",
      "127:\tlearn: 25.5627788\ttotal: 3.85s\tremaining: 26.2s\n",
      "128:\tlearn: 25.5573434\ttotal: 3.86s\tremaining: 26.1s\n",
      "129:\tlearn: 25.5440298\ttotal: 3.93s\tremaining: 26.3s\n",
      "130:\tlearn: 25.5328676\ttotal: 3.94s\tremaining: 26.1s\n",
      "131:\tlearn: 25.5185825\ttotal: 3.94s\tremaining: 25.9s\n",
      "132:\tlearn: 25.5133169\ttotal: 3.95s\tremaining: 25.8s\n",
      "133:\tlearn: 25.5046121\ttotal: 4.03s\tremaining: 26s\n",
      "134:\tlearn: 25.4998452\ttotal: 4.03s\tremaining: 25.9s\n",
      "135:\tlearn: 25.4874562\ttotal: 4.04s\tremaining: 25.7s\n",
      "136:\tlearn: 25.4719391\ttotal: 4.05s\tremaining: 25.5s\n",
      "137:\tlearn: 25.4613214\ttotal: 4.05s\tremaining: 25.3s\n",
      "138:\tlearn: 25.4531905\ttotal: 4.13s\tremaining: 25.6s\n",
      "139:\tlearn: 25.4414568\ttotal: 4.13s\tremaining: 25.4s\n",
      "140:\tlearn: 25.4360811\ttotal: 4.14s\tremaining: 25.2s\n",
      "141:\tlearn: 25.4224629\ttotal: 4.15s\tremaining: 25.1s\n",
      "142:\tlearn: 25.4115159\ttotal: 4.22s\tremaining: 25.3s\n",
      "143:\tlearn: 25.4039729\ttotal: 4.23s\tremaining: 25.1s\n",
      "144:\tlearn: 25.3948001\ttotal: 4.23s\tremaining: 25s\n",
      "145:\tlearn: 25.3891149\ttotal: 4.33s\tremaining: 25.3s\n",
      "146:\tlearn: 25.3790827\ttotal: 4.33s\tremaining: 25.1s\n",
      "147:\tlearn: 25.3701681\ttotal: 4.34s\tremaining: 25s\n",
      "148:\tlearn: 25.3614426\ttotal: 4.43s\tremaining: 25.3s\n",
      "149:\tlearn: 25.3499658\ttotal: 4.44s\tremaining: 25.2s\n",
      "150:\tlearn: 25.3401881\ttotal: 4.44s\tremaining: 25s\n",
      "151:\tlearn: 25.3319433\ttotal: 4.52s\tremaining: 25.2s\n",
      "152:\tlearn: 25.3230748\ttotal: 4.54s\tremaining: 25.1s\n",
      "153:\tlearn: 25.3081675\ttotal: 4.54s\tremaining: 25s\n",
      "154:\tlearn: 25.3013423\ttotal: 4.62s\tremaining: 25.2s\n",
      "155:\tlearn: 25.2919155\ttotal: 4.64s\tremaining: 25.1s\n",
      "156:\tlearn: 25.2846458\ttotal: 4.65s\tremaining: 25s\n",
      "157:\tlearn: 25.2758861\ttotal: 4.66s\tremaining: 24.8s\n",
      "158:\tlearn: 25.2648125\ttotal: 4.71s\tremaining: 24.9s\n",
      "159:\tlearn: 25.2579772\ttotal: 4.72s\tremaining: 24.8s\n",
      "160:\tlearn: 25.2510847\ttotal: 4.72s\tremaining: 24.6s\n",
      "161:\tlearn: 25.2442116\ttotal: 4.8s\tremaining: 24.8s\n",
      "162:\tlearn: 25.2336540\ttotal: 4.81s\tremaining: 24.7s\n",
      "163:\tlearn: 25.2285584\ttotal: 4.82s\tremaining: 24.6s\n",
      "164:\tlearn: 25.2173162\ttotal: 4.9s\tremaining: 24.8s\n",
      "165:\tlearn: 25.2135244\ttotal: 4.91s\tremaining: 24.7s\n",
      "166:\tlearn: 25.2039359\ttotal: 4.91s\tremaining: 24.5s\n",
      "167:\tlearn: 25.1972748\ttotal: 4.92s\tremaining: 24.4s\n",
      "168:\tlearn: 25.1835969\ttotal: 5s\tremaining: 24.6s\n",
      "169:\tlearn: 25.1746720\ttotal: 5.01s\tremaining: 24.5s\n",
      "170:\tlearn: 25.1681146\ttotal: 5.02s\tremaining: 24.3s\n",
      "171:\tlearn: 25.1583252\ttotal: 5.09s\tremaining: 24.5s\n",
      "172:\tlearn: 25.1521895\ttotal: 5.12s\tremaining: 24.5s\n",
      "173:\tlearn: 25.1403484\ttotal: 5.13s\tremaining: 24.3s\n",
      "174:\tlearn: 25.1342751\ttotal: 5.21s\tremaining: 24.6s\n",
      "175:\tlearn: 25.1279686\ttotal: 5.22s\tremaining: 24.4s\n",
      "176:\tlearn: 25.1236199\ttotal: 5.22s\tremaining: 24.3s\n",
      "177:\tlearn: 25.1158772\ttotal: 5.29s\tremaining: 24.4s\n",
      "178:\tlearn: 25.1112983\ttotal: 5.3s\tremaining: 24.3s\n",
      "179:\tlearn: 25.1075645\ttotal: 5.3s\tremaining: 24.2s\n",
      "180:\tlearn: 25.0951656\ttotal: 5.39s\tremaining: 24.4s\n",
      "181:\tlearn: 25.0859537\ttotal: 5.4s\tremaining: 24.3s\n",
      "182:\tlearn: 25.0759628\ttotal: 5.41s\tremaining: 24.2s\n",
      "183:\tlearn: 25.0714963\ttotal: 5.5s\tremaining: 24.4s\n",
      "184:\tlearn: 25.0666674\ttotal: 5.51s\tremaining: 24.3s\n",
      "185:\tlearn: 25.0601976\ttotal: 5.59s\tremaining: 24.5s\n",
      "186:\tlearn: 25.0564390\ttotal: 5.6s\tremaining: 24.3s\n",
      "187:\tlearn: 25.0518678\ttotal: 5.6s\tremaining: 24.2s\n",
      "188:\tlearn: 25.0461240\ttotal: 5.69s\tremaining: 24.4s\n",
      "189:\tlearn: 25.0393771\ttotal: 5.69s\tremaining: 24.3s\n",
      "190:\tlearn: 25.0325607\ttotal: 5.7s\tremaining: 24.1s\n",
      "191:\tlearn: 25.0286353\ttotal: 5.71s\tremaining: 24s\n",
      "192:\tlearn: 25.0244827\ttotal: 5.79s\tremaining: 24.2s\n",
      "193:\tlearn: 25.0218487\ttotal: 5.79s\tremaining: 24.1s\n",
      "194:\tlearn: 25.0177300\ttotal: 5.8s\tremaining: 24s\n",
      "195:\tlearn: 25.0085500\ttotal: 5.89s\tremaining: 24.2s\n",
      "196:\tlearn: 25.0052220\ttotal: 5.91s\tremaining: 24.1s\n",
      "197:\tlearn: 24.9948169\ttotal: 5.91s\tremaining: 23.9s\n",
      "198:\tlearn: 24.9834022\ttotal: 5.99s\tremaining: 24.1s\n",
      "199:\tlearn: 24.9777099\ttotal: 6s\tremaining: 24s\n",
      "200:\tlearn: 24.9709353\ttotal: 6.01s\tremaining: 23.9s\n",
      "201:\tlearn: 24.9661035\ttotal: 6.08s\tremaining: 24s\n",
      "202:\tlearn: 24.9592304\ttotal: 6.09s\tremaining: 23.9s\n",
      "203:\tlearn: 24.9509640\ttotal: 6.11s\tremaining: 23.8s\n",
      "204:\tlearn: 24.9430724\ttotal: 6.11s\tremaining: 23.7s\n",
      "205:\tlearn: 24.9407887\ttotal: 6.12s\tremaining: 23.6s\n",
      "206:\tlearn: 24.9386063\ttotal: 6.17s\tremaining: 23.7s\n",
      "207:\tlearn: 24.9309998\ttotal: 6.18s\tremaining: 23.5s\n",
      "208:\tlearn: 24.9262406\ttotal: 6.19s\tremaining: 23.4s\n",
      "209:\tlearn: 24.9153022\ttotal: 6.2s\tremaining: 23.3s\n",
      "210:\tlearn: 24.9086737\ttotal: 6.29s\tremaining: 23.5s\n",
      "211:\tlearn: 24.9061882\ttotal: 6.3s\tremaining: 23.4s\n",
      "212:\tlearn: 24.9011940\ttotal: 6.3s\tremaining: 23.3s\n",
      "213:\tlearn: 24.8968147\ttotal: 6.31s\tremaining: 23.2s\n",
      "214:\tlearn: 24.8910118\ttotal: 6.38s\tremaining: 23.3s\n",
      "215:\tlearn: 24.8872711\ttotal: 6.38s\tremaining: 23.2s\n",
      "216:\tlearn: 24.8817740\ttotal: 6.4s\tremaining: 23.1s\n",
      "217:\tlearn: 24.8766692\ttotal: 6.47s\tremaining: 23.2s\n",
      "218:\tlearn: 24.8723086\ttotal: 6.48s\tremaining: 23.1s\n",
      "219:\tlearn: 24.8618900\ttotal: 6.49s\tremaining: 23s\n",
      "220:\tlearn: 24.8588025\ttotal: 6.5s\tremaining: 22.9s\n",
      "221:\tlearn: 24.8543442\ttotal: 6.56s\tremaining: 23s\n",
      "222:\tlearn: 24.8477072\ttotal: 6.57s\tremaining: 22.9s\n",
      "223:\tlearn: 24.8467493\ttotal: 6.57s\tremaining: 22.8s\n",
      "224:\tlearn: 24.8429895\ttotal: 6.58s\tremaining: 22.7s\n",
      "225:\tlearn: 24.8346787\ttotal: 6.59s\tremaining: 22.6s\n",
      "226:\tlearn: 24.8316036\ttotal: 6.59s\tremaining: 22.5s\n",
      "227:\tlearn: 24.8244216\ttotal: 6.67s\tremaining: 22.6s\n",
      "228:\tlearn: 24.8203137\ttotal: 6.67s\tremaining: 22.5s\n",
      "229:\tlearn: 24.8124504\ttotal: 6.69s\tremaining: 22.4s\n",
      "230:\tlearn: 24.8069019\ttotal: 6.7s\tremaining: 22.3s\n",
      "231:\tlearn: 24.8041817\ttotal: 6.78s\tremaining: 22.4s\n",
      "232:\tlearn: 24.8016642\ttotal: 6.78s\tremaining: 22.3s\n",
      "233:\tlearn: 24.7965399\ttotal: 6.79s\tremaining: 22.2s\n",
      "234:\tlearn: 24.7890554\ttotal: 6.85s\tremaining: 22.3s\n",
      "235:\tlearn: 24.7837624\ttotal: 6.86s\tremaining: 22.2s\n",
      "236:\tlearn: 24.7776445\ttotal: 6.88s\tremaining: 22.1s\n",
      "237:\tlearn: 24.7734667\ttotal: 6.88s\tremaining: 22s\n",
      "238:\tlearn: 24.7650403\ttotal: 6.96s\tremaining: 22.2s\n",
      "239:\tlearn: 24.7553572\ttotal: 6.96s\tremaining: 22.1s\n",
      "240:\tlearn: 24.7477044\ttotal: 6.97s\tremaining: 22s\n",
      "241:\tlearn: 24.7420183\ttotal: 6.97s\tremaining: 21.8s\n",
      "242:\tlearn: 24.7372852\ttotal: 7.06s\tremaining: 22s\n",
      "243:\tlearn: 24.7291986\ttotal: 7.07s\tremaining: 21.9s\n",
      "244:\tlearn: 24.7223903\ttotal: 7.08s\tremaining: 21.8s\n",
      "245:\tlearn: 24.7188615\ttotal: 7.08s\tremaining: 21.7s\n",
      "246:\tlearn: 24.7129512\ttotal: 7.15s\tremaining: 21.8s\n",
      "247:\tlearn: 24.7065430\ttotal: 7.15s\tremaining: 21.7s\n",
      "248:\tlearn: 24.7042113\ttotal: 7.16s\tremaining: 21.6s\n",
      "249:\tlearn: 24.7001915\ttotal: 7.17s\tremaining: 21.5s\n",
      "250:\tlearn: 24.6945282\ttotal: 7.25s\tremaining: 21.6s\n",
      "251:\tlearn: 24.6874289\ttotal: 7.25s\tremaining: 21.5s\n",
      "252:\tlearn: 24.6818526\ttotal: 7.26s\tremaining: 21.5s\n",
      "253:\tlearn: 24.6771990\ttotal: 7.34s\tremaining: 21.6s\n",
      "254:\tlearn: 24.6747997\ttotal: 7.35s\tremaining: 21.5s\n",
      "255:\tlearn: 24.6659420\ttotal: 7.35s\tremaining: 21.4s\n",
      "256:\tlearn: 24.6571439\ttotal: 7.44s\tremaining: 21.5s\n",
      "257:\tlearn: 24.6527356\ttotal: 7.44s\tremaining: 21.4s\n",
      "258:\tlearn: 24.6471294\ttotal: 7.45s\tremaining: 21.3s\n",
      "259:\tlearn: 24.6419494\ttotal: 7.47s\tremaining: 21.3s\n",
      "260:\tlearn: 24.6324329\ttotal: 7.55s\tremaining: 21.4s\n",
      "261:\tlearn: 24.6266266\ttotal: 7.56s\tremaining: 21.3s\n",
      "262:\tlearn: 24.6238818\ttotal: 7.57s\tremaining: 21.2s\n",
      "263:\tlearn: 24.6187253\ttotal: 7.64s\tremaining: 21.3s\n",
      "264:\tlearn: 24.6100556\ttotal: 7.65s\tremaining: 21.2s\n",
      "265:\tlearn: 24.6065293\ttotal: 7.66s\tremaining: 21.1s\n",
      "266:\tlearn: 24.6019498\ttotal: 7.75s\tremaining: 21.3s\n",
      "267:\tlearn: 24.5955710\ttotal: 7.75s\tremaining: 21.2s\n",
      "268:\tlearn: 24.5923225\ttotal: 7.76s\tremaining: 21.1s\n",
      "269:\tlearn: 24.5895096\ttotal: 7.77s\tremaining: 21s\n",
      "270:\tlearn: 24.5824403\ttotal: 7.84s\tremaining: 21.1s\n",
      "271:\tlearn: 24.5745895\ttotal: 7.84s\tremaining: 21s\n",
      "272:\tlearn: 24.5698295\ttotal: 7.85s\tremaining: 20.9s\n",
      "273:\tlearn: 24.5655306\ttotal: 7.86s\tremaining: 20.8s\n",
      "274:\tlearn: 24.5594031\ttotal: 7.94s\tremaining: 20.9s\n",
      "275:\tlearn: 24.5527820\ttotal: 7.95s\tremaining: 20.9s\n",
      "276:\tlearn: 24.5451777\ttotal: 7.97s\tremaining: 20.8s\n",
      "277:\tlearn: 24.5431147\ttotal: 8.03s\tremaining: 20.9s\n",
      "278:\tlearn: 24.5365857\ttotal: 8.04s\tremaining: 20.8s\n",
      "279:\tlearn: 24.5309171\ttotal: 8.04s\tremaining: 20.7s\n",
      "280:\tlearn: 24.5275664\ttotal: 8.05s\tremaining: 20.6s\n",
      "281:\tlearn: 24.5207862\ttotal: 8.12s\tremaining: 20.7s\n",
      "282:\tlearn: 24.5166710\ttotal: 8.13s\tremaining: 20.6s\n",
      "283:\tlearn: 24.5150521\ttotal: 8.13s\tremaining: 20.5s\n",
      "284:\tlearn: 24.5081686\ttotal: 8.14s\tremaining: 20.4s\n",
      "285:\tlearn: 24.4974030\ttotal: 8.23s\tremaining: 20.5s\n",
      "286:\tlearn: 24.4916729\ttotal: 8.24s\tremaining: 20.5s\n",
      "287:\tlearn: 24.4873315\ttotal: 8.24s\tremaining: 20.4s\n",
      "288:\tlearn: 24.4838365\ttotal: 8.25s\tremaining: 20.3s\n",
      "289:\tlearn: 24.4735204\ttotal: 8.32s\tremaining: 20.4s\n",
      "290:\tlearn: 24.4664626\ttotal: 8.33s\tremaining: 20.3s\n",
      "291:\tlearn: 24.4569383\ttotal: 8.33s\tremaining: 20.2s\n",
      "292:\tlearn: 24.4536375\ttotal: 8.42s\tremaining: 20.3s\n",
      "293:\tlearn: 24.4463150\ttotal: 8.43s\tremaining: 20.2s\n",
      "294:\tlearn: 24.4436397\ttotal: 8.43s\tremaining: 20.2s\n",
      "295:\tlearn: 24.4356479\ttotal: 8.52s\tremaining: 20.3s\n",
      "296:\tlearn: 24.4289245\ttotal: 8.53s\tremaining: 20.2s\n",
      "297:\tlearn: 24.4227149\ttotal: 8.53s\tremaining: 20.1s\n",
      "298:\tlearn: 24.4121723\ttotal: 8.54s\tremaining: 20s\n",
      "299:\tlearn: 24.4054488\ttotal: 8.61s\tremaining: 20.1s\n",
      "300:\tlearn: 24.4020670\ttotal: 8.62s\tremaining: 20s\n",
      "301:\tlearn: 24.3950846\ttotal: 8.63s\tremaining: 19.9s\n",
      "302:\tlearn: 24.3894591\ttotal: 8.63s\tremaining: 19.9s\n",
      "303:\tlearn: 24.3859492\ttotal: 8.64s\tremaining: 19.8s\n",
      "304:\tlearn: 24.3822081\ttotal: 8.64s\tremaining: 19.7s\n",
      "305:\tlearn: 24.3773991\ttotal: 8.65s\tremaining: 19.6s\n",
      "306:\tlearn: 24.3705849\ttotal: 8.65s\tremaining: 19.5s\n",
      "307:\tlearn: 24.3675937\ttotal: 8.66s\tremaining: 19.5s\n",
      "308:\tlearn: 24.3604684\ttotal: 8.66s\tremaining: 19.4s\n",
      "309:\tlearn: 24.3556696\ttotal: 8.71s\tremaining: 19.4s\n",
      "310:\tlearn: 24.3497584\ttotal: 8.72s\tremaining: 19.3s\n",
      "311:\tlearn: 24.3440021\ttotal: 8.72s\tremaining: 19.2s\n",
      "312:\tlearn: 24.3396933\ttotal: 8.81s\tremaining: 19.3s\n",
      "313:\tlearn: 24.3320206\ttotal: 8.82s\tremaining: 19.3s\n",
      "314:\tlearn: 24.3221890\ttotal: 8.82s\tremaining: 19.2s\n",
      "315:\tlearn: 24.3153600\ttotal: 8.83s\tremaining: 19.1s\n",
      "316:\tlearn: 24.3099954\ttotal: 8.91s\tremaining: 19.2s\n",
      "317:\tlearn: 24.3048907\ttotal: 8.91s\tremaining: 19.1s\n",
      "318:\tlearn: 24.3006753\ttotal: 8.92s\tremaining: 19s\n",
      "319:\tlearn: 24.2983371\ttotal: 9.01s\tremaining: 19.1s\n",
      "320:\tlearn: 24.2948527\ttotal: 9.01s\tremaining: 19.1s\n",
      "321:\tlearn: 24.2868720\ttotal: 9.1s\tremaining: 19.2s\n",
      "322:\tlearn: 24.2779707\ttotal: 9.11s\tremaining: 19.1s\n",
      "323:\tlearn: 24.2726352\ttotal: 9.12s\tremaining: 19s\n",
      "324:\tlearn: 24.2661100\ttotal: 9.2s\tremaining: 19.1s\n",
      "325:\tlearn: 24.2638684\ttotal: 9.21s\tremaining: 19s\n",
      "326:\tlearn: 24.2576780\ttotal: 9.3s\tremaining: 19.1s\n",
      "327:\tlearn: 24.2496003\ttotal: 9.31s\tremaining: 19.1s\n",
      "328:\tlearn: 24.2454715\ttotal: 9.32s\tremaining: 19s\n",
      "329:\tlearn: 24.2405572\ttotal: 9.33s\tremaining: 18.9s\n",
      "330:\tlearn: 24.2367534\ttotal: 9.4s\tremaining: 19s\n",
      "331:\tlearn: 24.2312094\ttotal: 9.41s\tremaining: 18.9s\n",
      "332:\tlearn: 24.2295337\ttotal: 9.42s\tremaining: 18.9s\n",
      "333:\tlearn: 24.2245133\ttotal: 9.49s\tremaining: 18.9s\n",
      "334:\tlearn: 24.2187462\ttotal: 9.51s\tremaining: 18.9s\n",
      "335:\tlearn: 24.2112425\ttotal: 9.52s\tremaining: 18.8s\n",
      "336:\tlearn: 24.2014839\ttotal: 9.6s\tremaining: 18.9s\n",
      "337:\tlearn: 24.1976451\ttotal: 9.61s\tremaining: 18.8s\n",
      "338:\tlearn: 24.1894410\ttotal: 9.61s\tremaining: 18.7s\n",
      "339:\tlearn: 24.1824804\ttotal: 9.72s\tremaining: 18.9s\n",
      "340:\tlearn: 24.1815069\ttotal: 9.72s\tremaining: 18.8s\n",
      "341:\tlearn: 24.1761920\ttotal: 9.73s\tremaining: 18.7s\n",
      "342:\tlearn: 24.1754537\ttotal: 9.8s\tremaining: 18.8s\n",
      "343:\tlearn: 24.1715798\ttotal: 9.8s\tremaining: 18.7s\n",
      "344:\tlearn: 24.1680434\ttotal: 9.81s\tremaining: 18.6s\n",
      "345:\tlearn: 24.1617604\ttotal: 9.83s\tremaining: 18.6s\n",
      "346:\tlearn: 24.1531226\ttotal: 9.9s\tremaining: 18.6s\n",
      "347:\tlearn: 24.1471200\ttotal: 9.91s\tremaining: 18.6s\n",
      "348:\tlearn: 24.1413804\ttotal: 9.92s\tremaining: 18.5s\n",
      "349:\tlearn: 24.1361476\ttotal: 9.93s\tremaining: 18.4s\n",
      "350:\tlearn: 24.1321226\ttotal: 9.98s\tremaining: 18.5s\n",
      "351:\tlearn: 24.1247054\ttotal: 9.99s\tremaining: 18.4s\n",
      "352:\tlearn: 24.1196019\ttotal: 10s\tremaining: 18.3s\n",
      "353:\tlearn: 24.1114212\ttotal: 10.1s\tremaining: 18.4s\n",
      "354:\tlearn: 24.1069451\ttotal: 10.1s\tremaining: 18.4s\n",
      "355:\tlearn: 24.1034871\ttotal: 10.1s\tremaining: 18.3s\n",
      "356:\tlearn: 24.1007546\ttotal: 10.1s\tremaining: 18.2s\n",
      "357:\tlearn: 24.0927717\ttotal: 10.2s\tremaining: 18.3s\n",
      "358:\tlearn: 24.0886945\ttotal: 10.2s\tremaining: 18.2s\n",
      "359:\tlearn: 24.0856247\ttotal: 10.2s\tremaining: 18.1s\n",
      "360:\tlearn: 24.0756645\ttotal: 10.3s\tremaining: 18.2s\n",
      "361:\tlearn: 24.0718372\ttotal: 10.3s\tremaining: 18.2s\n",
      "362:\tlearn: 24.0691696\ttotal: 10.3s\tremaining: 18.1s\n",
      "363:\tlearn: 24.0639515\ttotal: 10.5s\tremaining: 18.3s\n",
      "364:\tlearn: 24.0582232\ttotal: 10.5s\tremaining: 18.3s\n",
      "365:\tlearn: 24.0549492\ttotal: 10.5s\tremaining: 18.2s\n",
      "366:\tlearn: 24.0474706\ttotal: 10.5s\tremaining: 18.1s\n",
      "367:\tlearn: 24.0406103\ttotal: 10.6s\tremaining: 18.2s\n",
      "368:\tlearn: 24.0360987\ttotal: 10.6s\tremaining: 18.1s\n",
      "369:\tlearn: 24.0312539\ttotal: 10.6s\tremaining: 18s\n",
      "370:\tlearn: 24.0306620\ttotal: 10.6s\tremaining: 18s\n",
      "371:\tlearn: 24.0249952\ttotal: 10.6s\tremaining: 17.9s\n",
      "372:\tlearn: 24.0187241\ttotal: 10.7s\tremaining: 18s\n",
      "373:\tlearn: 24.0133800\ttotal: 10.7s\tremaining: 17.9s\n",
      "374:\tlearn: 24.0082316\ttotal: 10.7s\tremaining: 17.8s\n",
      "375:\tlearn: 24.0074984\ttotal: 10.8s\tremaining: 17.9s\n",
      "376:\tlearn: 24.0040463\ttotal: 10.8s\tremaining: 17.8s\n",
      "377:\tlearn: 23.9985291\ttotal: 10.8s\tremaining: 17.7s\n",
      "378:\tlearn: 23.9956088\ttotal: 10.8s\tremaining: 17.7s\n",
      "379:\tlearn: 23.9869360\ttotal: 10.9s\tremaining: 17.7s\n",
      "380:\tlearn: 23.9808220\ttotal: 10.9s\tremaining: 17.7s\n",
      "381:\tlearn: 23.9762876\ttotal: 10.9s\tremaining: 17.6s\n",
      "382:\tlearn: 23.9687039\ttotal: 10.9s\tremaining: 17.5s\n",
      "383:\tlearn: 23.9679261\ttotal: 11s\tremaining: 17.6s\n",
      "384:\tlearn: 23.9631936\ttotal: 11s\tremaining: 17.5s\n",
      "385:\tlearn: 23.9624476\ttotal: 11s\tremaining: 17.5s\n",
      "386:\tlearn: 23.9574024\ttotal: 11.1s\tremaining: 17.5s\n",
      "387:\tlearn: 23.9526802\ttotal: 11.1s\tremaining: 17.5s\n",
      "388:\tlearn: 23.9468504\ttotal: 11.1s\tremaining: 17.4s\n",
      "389:\tlearn: 23.9395891\ttotal: 11.1s\tremaining: 17.3s\n",
      "390:\tlearn: 23.9316688\ttotal: 11.1s\tremaining: 17.3s\n",
      "391:\tlearn: 23.9263816\ttotal: 11.1s\tremaining: 17.2s\n",
      "392:\tlearn: 23.9182307\ttotal: 11.2s\tremaining: 17.3s\n",
      "393:\tlearn: 23.9125826\ttotal: 11.2s\tremaining: 17.2s\n",
      "394:\tlearn: 23.9053407\ttotal: 11.2s\tremaining: 17.1s\n",
      "395:\tlearn: 23.8996972\ttotal: 11.3s\tremaining: 17.2s\n",
      "396:\tlearn: 23.8907561\ttotal: 11.3s\tremaining: 17.1s\n",
      "397:\tlearn: 23.8902328\ttotal: 11.3s\tremaining: 17.1s\n",
      "398:\tlearn: 23.8858669\ttotal: 11.3s\tremaining: 17s\n",
      "399:\tlearn: 23.8804069\ttotal: 11.4s\tremaining: 17s\n",
      "400:\tlearn: 23.8762956\ttotal: 11.4s\tremaining: 17s\n",
      "401:\tlearn: 23.8726148\ttotal: 11.5s\tremaining: 17.1s\n",
      "402:\tlearn: 23.8674107\ttotal: 11.5s\tremaining: 17s\n",
      "403:\tlearn: 23.8588664\ttotal: 11.5s\tremaining: 16.9s\n",
      "404:\tlearn: 23.8516303\ttotal: 11.5s\tremaining: 16.9s\n",
      "405:\tlearn: 23.8473960\ttotal: 11.6s\tremaining: 16.9s\n",
      "406:\tlearn: 23.8398953\ttotal: 11.6s\tremaining: 16.9s\n",
      "407:\tlearn: 23.8359097\ttotal: 11.6s\tremaining: 16.8s\n",
      "408:\tlearn: 23.8319487\ttotal: 11.7s\tremaining: 16.9s\n",
      "409:\tlearn: 23.8266591\ttotal: 11.7s\tremaining: 16.8s\n",
      "410:\tlearn: 23.8169086\ttotal: 11.7s\tremaining: 16.8s\n",
      "411:\tlearn: 23.8146776\ttotal: 11.8s\tremaining: 16.8s\n",
      "412:\tlearn: 23.8086476\ttotal: 11.8s\tremaining: 16.7s\n",
      "413:\tlearn: 23.8025931\ttotal: 11.8s\tremaining: 16.8s\n",
      "414:\tlearn: 23.7978320\ttotal: 11.9s\tremaining: 16.7s\n",
      "415:\tlearn: 23.7944114\ttotal: 11.9s\tremaining: 16.7s\n",
      "416:\tlearn: 23.7905351\ttotal: 11.9s\tremaining: 16.6s\n",
      "417:\tlearn: 23.7876937\ttotal: 12s\tremaining: 16.6s\n",
      "418:\tlearn: 23.7821838\ttotal: 12s\tremaining: 16.6s\n",
      "419:\tlearn: 23.7776130\ttotal: 12s\tremaining: 16.5s\n",
      "420:\tlearn: 23.7710639\ttotal: 12s\tremaining: 16.5s\n",
      "421:\tlearn: 23.7660184\ttotal: 12s\tremaining: 16.5s\n",
      "422:\tlearn: 23.7654856\ttotal: 12.1s\tremaining: 16.4s\n",
      "423:\tlearn: 23.7604418\ttotal: 12.1s\tremaining: 16.5s\n",
      "424:\tlearn: 23.7542056\ttotal: 12.1s\tremaining: 16.4s\n",
      "425:\tlearn: 23.7501939\ttotal: 12.2s\tremaining: 16.4s\n",
      "426:\tlearn: 23.7451968\ttotal: 12.2s\tremaining: 16.4s\n",
      "427:\tlearn: 23.7407015\ttotal: 12.2s\tremaining: 16.4s\n",
      "428:\tlearn: 23.7349635\ttotal: 12.3s\tremaining: 16.3s\n",
      "429:\tlearn: 23.7291150\ttotal: 12.3s\tremaining: 16.4s\n",
      "430:\tlearn: 23.7229154\ttotal: 12.3s\tremaining: 16.3s\n",
      "431:\tlearn: 23.7160244\ttotal: 12.3s\tremaining: 16.2s\n",
      "432:\tlearn: 23.7116579\ttotal: 12.4s\tremaining: 16.3s\n",
      "433:\tlearn: 23.7110602\ttotal: 12.4s\tremaining: 16.2s\n",
      "434:\tlearn: 23.7071297\ttotal: 12.5s\tremaining: 16.2s\n",
      "435:\tlearn: 23.7028851\ttotal: 12.5s\tremaining: 16.2s\n",
      "436:\tlearn: 23.6991509\ttotal: 12.5s\tremaining: 16.2s\n",
      "437:\tlearn: 23.6927240\ttotal: 12.6s\tremaining: 16.2s\n",
      "438:\tlearn: 23.6921772\ttotal: 12.6s\tremaining: 16.2s\n",
      "439:\tlearn: 23.6871049\ttotal: 12.7s\tremaining: 16.1s\n",
      "440:\tlearn: 23.6836408\ttotal: 12.7s\tremaining: 16.1s\n",
      "441:\tlearn: 23.6775419\ttotal: 12.7s\tremaining: 16.1s\n",
      "442:\tlearn: 23.6770481\ttotal: 12.8s\tremaining: 16s\n",
      "443:\tlearn: 23.6715652\ttotal: 12.8s\tremaining: 16.1s\n",
      "444:\tlearn: 23.6652996\ttotal: 12.8s\tremaining: 16s\n",
      "445:\tlearn: 23.6601195\ttotal: 12.8s\tremaining: 16s\n",
      "446:\tlearn: 23.6561876\ttotal: 12.9s\tremaining: 16s\n",
      "447:\tlearn: 23.6498457\ttotal: 12.9s\tremaining: 15.9s\n",
      "448:\tlearn: 23.6494420\ttotal: 12.9s\tremaining: 15.9s\n",
      "449:\tlearn: 23.6455675\ttotal: 13s\tremaining: 15.9s\n",
      "450:\tlearn: 23.6418195\ttotal: 13s\tremaining: 15.9s\n",
      "451:\tlearn: 23.6363951\ttotal: 13s\tremaining: 15.8s\n",
      "452:\tlearn: 23.6358515\ttotal: 13s\tremaining: 15.8s\n",
      "453:\tlearn: 23.6321919\ttotal: 13.1s\tremaining: 15.8s\n",
      "454:\tlearn: 23.6257933\ttotal: 13.1s\tremaining: 15.7s\n",
      "455:\tlearn: 23.6221527\ttotal: 13.1s\tremaining: 15.7s\n",
      "456:\tlearn: 23.6167540\ttotal: 13.2s\tremaining: 15.7s\n",
      "457:\tlearn: 23.6127173\ttotal: 13.2s\tremaining: 15.7s\n",
      "458:\tlearn: 23.6106805\ttotal: 13.2s\tremaining: 15.6s\n",
      "459:\tlearn: 23.6097220\ttotal: 13.3s\tremaining: 15.6s\n",
      "460:\tlearn: 23.6057039\ttotal: 13.3s\tremaining: 15.6s\n",
      "461:\tlearn: 23.6025142\ttotal: 13.3s\tremaining: 15.5s\n",
      "462:\tlearn: 23.5980564\ttotal: 13.4s\tremaining: 15.6s\n",
      "463:\tlearn: 23.5936305\ttotal: 13.4s\tremaining: 15.5s\n",
      "464:\tlearn: 23.5889941\ttotal: 13.4s\tremaining: 15.5s\n",
      "465:\tlearn: 23.5845225\ttotal: 13.5s\tremaining: 15.5s\n",
      "466:\tlearn: 23.5802609\ttotal: 13.5s\tremaining: 15.4s\n",
      "467:\tlearn: 23.5758496\ttotal: 13.5s\tremaining: 15.4s\n",
      "468:\tlearn: 23.5714594\ttotal: 13.6s\tremaining: 15.4s\n",
      "469:\tlearn: 23.5654847\ttotal: 13.6s\tremaining: 15.4s\n",
      "470:\tlearn: 23.5625365\ttotal: 13.6s\tremaining: 15.3s\n",
      "471:\tlearn: 23.5573591\ttotal: 13.7s\tremaining: 15.3s\n",
      "472:\tlearn: 23.5569803\ttotal: 13.7s\tremaining: 15.3s\n",
      "473:\tlearn: 23.5542729\ttotal: 13.7s\tremaining: 15.2s\n",
      "474:\tlearn: 23.5505195\ttotal: 13.7s\tremaining: 15.2s\n",
      "475:\tlearn: 23.5463659\ttotal: 13.8s\tremaining: 15.2s\n",
      "476:\tlearn: 23.5409861\ttotal: 13.8s\tremaining: 15.2s\n",
      "477:\tlearn: 23.5383768\ttotal: 13.8s\tremaining: 15.1s\n",
      "478:\tlearn: 23.5347111\ttotal: 13.9s\tremaining: 15.1s\n",
      "479:\tlearn: 23.5285303\ttotal: 13.9s\tremaining: 15.1s\n",
      "480:\tlearn: 23.5260127\ttotal: 13.9s\tremaining: 15s\n",
      "481:\tlearn: 23.5192631\ttotal: 13.9s\tremaining: 15s\n",
      "482:\tlearn: 23.5188296\ttotal: 14s\tremaining: 15s\n",
      "483:\tlearn: 23.5133792\ttotal: 14s\tremaining: 14.9s\n",
      "484:\tlearn: 23.5089425\ttotal: 14s\tremaining: 14.9s\n",
      "485:\tlearn: 23.5012933\ttotal: 14.1s\tremaining: 14.9s\n",
      "486:\tlearn: 23.4981696\ttotal: 14.1s\tremaining: 14.9s\n",
      "487:\tlearn: 23.4976994\ttotal: 14.1s\tremaining: 14.8s\n",
      "488:\tlearn: 23.4972298\ttotal: 14.1s\tremaining: 14.8s\n",
      "489:\tlearn: 23.4931706\ttotal: 14.2s\tremaining: 14.8s\n",
      "490:\tlearn: 23.4856977\ttotal: 14.2s\tremaining: 14.7s\n",
      "491:\tlearn: 23.4791851\ttotal: 14.2s\tremaining: 14.7s\n",
      "492:\tlearn: 23.4712130\ttotal: 14.2s\tremaining: 14.6s\n",
      "493:\tlearn: 23.4668767\ttotal: 14.3s\tremaining: 14.7s\n",
      "494:\tlearn: 23.4624953\ttotal: 14.3s\tremaining: 14.6s\n",
      "495:\tlearn: 23.4572925\ttotal: 14.3s\tremaining: 14.6s\n",
      "496:\tlearn: 23.4551047\ttotal: 14.3s\tremaining: 14.5s\n",
      "497:\tlearn: 23.4487823\ttotal: 14.4s\tremaining: 14.5s\n",
      "498:\tlearn: 23.4409063\ttotal: 14.4s\tremaining: 14.5s\n",
      "499:\tlearn: 23.4345722\ttotal: 14.4s\tremaining: 14.4s\n",
      "500:\tlearn: 23.4280537\ttotal: 14.4s\tremaining: 14.4s\n",
      "501:\tlearn: 23.4252452\ttotal: 14.5s\tremaining: 14.4s\n",
      "502:\tlearn: 23.4166778\ttotal: 14.5s\tremaining: 14.3s\n",
      "503:\tlearn: 23.4124428\ttotal: 14.5s\tremaining: 14.3s\n",
      "504:\tlearn: 23.4081677\ttotal: 14.6s\tremaining: 14.3s\n",
      "505:\tlearn: 23.4043698\ttotal: 14.6s\tremaining: 14.3s\n",
      "506:\tlearn: 23.4008351\ttotal: 14.6s\tremaining: 14.2s\n",
      "507:\tlearn: 23.3939142\ttotal: 14.6s\tremaining: 14.2s\n",
      "508:\tlearn: 23.3903687\ttotal: 14.6s\tremaining: 14.1s\n",
      "509:\tlearn: 23.3853934\ttotal: 14.7s\tremaining: 14.1s\n",
      "510:\tlearn: 23.3801199\ttotal: 14.7s\tremaining: 14.1s\n",
      "511:\tlearn: 23.3766424\ttotal: 14.7s\tremaining: 14s\n",
      "512:\tlearn: 23.3717880\ttotal: 14.8s\tremaining: 14s\n",
      "513:\tlearn: 23.3645530\ttotal: 14.8s\tremaining: 14s\n",
      "514:\tlearn: 23.3568354\ttotal: 14.8s\tremaining: 14s\n",
      "515:\tlearn: 23.3506779\ttotal: 14.8s\tremaining: 13.9s\n",
      "516:\tlearn: 23.3459295\ttotal: 14.9s\tremaining: 13.9s\n",
      "517:\tlearn: 23.3423544\ttotal: 14.9s\tremaining: 13.9s\n",
      "518:\tlearn: 23.3358388\ttotal: 14.9s\tremaining: 13.8s\n",
      "519:\tlearn: 23.3323136\ttotal: 14.9s\tremaining: 13.8s\n",
      "520:\tlearn: 23.3299127\ttotal: 15s\tremaining: 13.8s\n",
      "521:\tlearn: 23.3251910\ttotal: 15s\tremaining: 13.7s\n",
      "522:\tlearn: 23.3247203\ttotal: 15s\tremaining: 13.7s\n",
      "523:\tlearn: 23.3222534\ttotal: 15s\tremaining: 13.6s\n",
      "524:\tlearn: 23.3154805\ttotal: 15.1s\tremaining: 13.7s\n",
      "525:\tlearn: 23.3089594\ttotal: 15.1s\tremaining: 13.6s\n",
      "526:\tlearn: 23.3047194\ttotal: 15.1s\tremaining: 13.6s\n",
      "527:\tlearn: 23.2974305\ttotal: 15.2s\tremaining: 13.6s\n",
      "528:\tlearn: 23.2897389\ttotal: 15.2s\tremaining: 13.5s\n",
      "529:\tlearn: 23.2821035\ttotal: 15.2s\tremaining: 13.5s\n",
      "530:\tlearn: 23.2767904\ttotal: 15.3s\tremaining: 13.5s\n",
      "531:\tlearn: 23.2710861\ttotal: 15.3s\tremaining: 13.4s\n",
      "532:\tlearn: 23.2660854\ttotal: 15.3s\tremaining: 13.4s\n",
      "533:\tlearn: 23.2595206\ttotal: 15.4s\tremaining: 13.4s\n",
      "534:\tlearn: 23.2533355\ttotal: 15.4s\tremaining: 13.4s\n",
      "535:\tlearn: 23.2447637\ttotal: 15.4s\tremaining: 13.3s\n",
      "536:\tlearn: 23.2382375\ttotal: 15.5s\tremaining: 13.3s\n",
      "537:\tlearn: 23.2298186\ttotal: 15.5s\tremaining: 13.3s\n",
      "538:\tlearn: 23.2263018\ttotal: 15.5s\tremaining: 13.3s\n",
      "539:\tlearn: 23.2197422\ttotal: 15.6s\tremaining: 13.3s\n",
      "540:\tlearn: 23.2138638\ttotal: 15.6s\tremaining: 13.2s\n",
      "541:\tlearn: 23.2107875\ttotal: 15.6s\tremaining: 13.2s\n",
      "542:\tlearn: 23.2103985\ttotal: 15.7s\tremaining: 13.2s\n",
      "543:\tlearn: 23.2015049\ttotal: 15.7s\tremaining: 13.1s\n",
      "544:\tlearn: 23.1922445\ttotal: 15.7s\tremaining: 13.1s\n",
      "545:\tlearn: 23.1870647\ttotal: 15.8s\tremaining: 13.1s\n",
      "546:\tlearn: 23.1819699\ttotal: 15.8s\tremaining: 13.1s\n",
      "547:\tlearn: 23.1763164\ttotal: 15.8s\tremaining: 13s\n",
      "548:\tlearn: 23.1716816\ttotal: 15.9s\tremaining: 13s\n",
      "549:\tlearn: 23.1693008\ttotal: 15.9s\tremaining: 13s\n",
      "550:\tlearn: 23.1631676\ttotal: 15.9s\tremaining: 12.9s\n",
      "551:\tlearn: 23.1597589\ttotal: 16s\tremaining: 13s\n",
      "552:\tlearn: 23.1566145\ttotal: 16s\tremaining: 12.9s\n",
      "553:\tlearn: 23.1506486\ttotal: 16s\tremaining: 12.9s\n",
      "554:\tlearn: 23.1473336\ttotal: 16.1s\tremaining: 12.9s\n",
      "555:\tlearn: 23.1421379\ttotal: 16.1s\tremaining: 12.8s\n",
      "556:\tlearn: 23.1355716\ttotal: 16.1s\tremaining: 12.8s\n",
      "557:\tlearn: 23.1290761\ttotal: 16.1s\tremaining: 12.7s\n",
      "558:\tlearn: 23.1207489\ttotal: 16.2s\tremaining: 12.7s\n",
      "559:\tlearn: 23.1147125\ttotal: 16.2s\tremaining: 12.7s\n",
      "560:\tlearn: 23.1142849\ttotal: 16.2s\tremaining: 12.7s\n",
      "561:\tlearn: 23.1090368\ttotal: 16.3s\tremaining: 12.7s\n",
      "562:\tlearn: 23.1055922\ttotal: 16.3s\tremaining: 12.6s\n",
      "563:\tlearn: 23.1016103\ttotal: 16.3s\tremaining: 12.6s\n",
      "564:\tlearn: 23.0980171\ttotal: 16.4s\tremaining: 12.6s\n",
      "565:\tlearn: 23.0934162\ttotal: 16.4s\tremaining: 12.6s\n",
      "566:\tlearn: 23.0881884\ttotal: 16.4s\tremaining: 12.6s\n",
      "567:\tlearn: 23.0816701\ttotal: 16.5s\tremaining: 12.5s\n",
      "568:\tlearn: 23.0760309\ttotal: 16.5s\tremaining: 12.5s\n",
      "569:\tlearn: 23.0702990\ttotal: 16.5s\tremaining: 12.5s\n",
      "570:\tlearn: 23.0649528\ttotal: 16.5s\tremaining: 12.4s\n",
      "571:\tlearn: 23.0626277\ttotal: 16.6s\tremaining: 12.4s\n",
      "572:\tlearn: 23.0569376\ttotal: 16.6s\tremaining: 12.3s\n",
      "573:\tlearn: 23.0565404\ttotal: 16.6s\tremaining: 12.4s\n",
      "574:\tlearn: 23.0485524\ttotal: 16.7s\tremaining: 12.3s\n",
      "575:\tlearn: 23.0452781\ttotal: 16.7s\tremaining: 12.3s\n",
      "576:\tlearn: 23.0427016\ttotal: 16.7s\tremaining: 12.3s\n",
      "577:\tlearn: 23.0423720\ttotal: 16.7s\tremaining: 12.2s\n",
      "578:\tlearn: 23.0394808\ttotal: 16.8s\tremaining: 12.2s\n",
      "579:\tlearn: 23.0364392\ttotal: 16.8s\tremaining: 12.2s\n",
      "580:\tlearn: 23.0323057\ttotal: 16.9s\tremaining: 12.2s\n",
      "581:\tlearn: 23.0292443\ttotal: 16.9s\tremaining: 12.2s\n",
      "582:\tlearn: 23.0267008\ttotal: 16.9s\tremaining: 12.1s\n",
      "583:\tlearn: 23.0237507\ttotal: 16.9s\tremaining: 12.1s\n",
      "584:\tlearn: 23.0216812\ttotal: 17s\tremaining: 12.1s\n",
      "585:\tlearn: 23.0187430\ttotal: 17s\tremaining: 12s\n",
      "586:\tlearn: 23.0136308\ttotal: 17.1s\tremaining: 12s\n",
      "587:\tlearn: 23.0111115\ttotal: 17.1s\tremaining: 12s\n",
      "588:\tlearn: 23.0090890\ttotal: 17.2s\tremaining: 12s\n",
      "589:\tlearn: 23.0024689\ttotal: 17.2s\tremaining: 12s\n",
      "590:\tlearn: 22.9992672\ttotal: 17.2s\tremaining: 11.9s\n",
      "591:\tlearn: 22.9989794\ttotal: 17.2s\tremaining: 11.9s\n",
      "592:\tlearn: 22.9954751\ttotal: 17.3s\tremaining: 11.8s\n",
      "593:\tlearn: 22.9906145\ttotal: 17.3s\tremaining: 11.8s\n",
      "594:\tlearn: 22.9849192\ttotal: 17.3s\tremaining: 11.8s\n",
      "595:\tlearn: 22.9763901\ttotal: 17.3s\tremaining: 11.8s\n",
      "596:\tlearn: 22.9731592\ttotal: 17.4s\tremaining: 11.7s\n",
      "597:\tlearn: 22.9659286\ttotal: 17.4s\tremaining: 11.7s\n",
      "598:\tlearn: 22.9601069\ttotal: 17.4s\tremaining: 11.7s\n",
      "599:\tlearn: 22.9543956\ttotal: 17.4s\tremaining: 11.6s\n",
      "600:\tlearn: 22.9447906\ttotal: 17.4s\tremaining: 11.6s\n",
      "601:\tlearn: 22.9382159\ttotal: 17.5s\tremaining: 11.5s\n",
      "602:\tlearn: 22.9337847\ttotal: 17.5s\tremaining: 11.5s\n",
      "603:\tlearn: 22.9288736\ttotal: 17.5s\tremaining: 11.5s\n",
      "604:\tlearn: 22.9251212\ttotal: 17.5s\tremaining: 11.5s\n",
      "605:\tlearn: 22.9212390\ttotal: 17.5s\tremaining: 11.4s\n",
      "606:\tlearn: 22.9192555\ttotal: 17.6s\tremaining: 11.4s\n",
      "607:\tlearn: 22.9168914\ttotal: 17.6s\tremaining: 11.4s\n",
      "608:\tlearn: 22.9103785\ttotal: 17.6s\tremaining: 11.3s\n",
      "609:\tlearn: 22.9068261\ttotal: 17.7s\tremaining: 11.3s\n",
      "610:\tlearn: 22.9044647\ttotal: 17.7s\tremaining: 11.2s\n",
      "611:\tlearn: 22.8968371\ttotal: 17.7s\tremaining: 11.2s\n",
      "612:\tlearn: 22.8888310\ttotal: 17.7s\tremaining: 11.2s\n",
      "613:\tlearn: 22.8853747\ttotal: 17.7s\tremaining: 11.2s\n",
      "614:\tlearn: 22.8773617\ttotal: 17.8s\tremaining: 11.2s\n",
      "615:\tlearn: 22.8747217\ttotal: 17.8s\tremaining: 11.1s\n",
      "616:\tlearn: 22.8689170\ttotal: 17.8s\tremaining: 11.1s\n",
      "617:\tlearn: 22.8650565\ttotal: 17.8s\tremaining: 11s\n",
      "618:\tlearn: 22.8590458\ttotal: 17.9s\tremaining: 11s\n",
      "619:\tlearn: 22.8570650\ttotal: 17.9s\tremaining: 11s\n",
      "620:\tlearn: 22.8507189\ttotal: 17.9s\tremaining: 10.9s\n",
      "621:\tlearn: 22.8468786\ttotal: 17.9s\tremaining: 10.9s\n",
      "622:\tlearn: 22.8417945\ttotal: 18s\tremaining: 10.9s\n",
      "623:\tlearn: 22.8367841\ttotal: 18s\tremaining: 10.9s\n",
      "624:\tlearn: 22.8297389\ttotal: 18.1s\tremaining: 10.8s\n",
      "625:\tlearn: 22.8271132\ttotal: 18.1s\tremaining: 10.8s\n",
      "626:\tlearn: 22.8220885\ttotal: 18.1s\tremaining: 10.8s\n",
      "627:\tlearn: 22.8149031\ttotal: 18.1s\tremaining: 10.7s\n",
      "628:\tlearn: 22.8130340\ttotal: 18.2s\tremaining: 10.7s\n",
      "629:\tlearn: 22.8088028\ttotal: 18.2s\tremaining: 10.7s\n",
      "630:\tlearn: 22.8050632\ttotal: 18.2s\tremaining: 10.7s\n",
      "631:\tlearn: 22.8013414\ttotal: 18.2s\tremaining: 10.6s\n",
      "632:\tlearn: 22.7963904\ttotal: 18.3s\tremaining: 10.6s\n",
      "633:\tlearn: 22.7887947\ttotal: 18.3s\tremaining: 10.6s\n",
      "634:\tlearn: 22.7859405\ttotal: 18.3s\tremaining: 10.5s\n",
      "635:\tlearn: 22.7755027\ttotal: 18.4s\tremaining: 10.5s\n",
      "636:\tlearn: 22.7699854\ttotal: 18.4s\tremaining: 10.5s\n",
      "637:\tlearn: 22.7661726\ttotal: 18.4s\tremaining: 10.4s\n",
      "638:\tlearn: 22.7612053\ttotal: 18.4s\tremaining: 10.4s\n",
      "639:\tlearn: 22.7593350\ttotal: 18.5s\tremaining: 10.4s\n",
      "640:\tlearn: 22.7556047\ttotal: 18.5s\tremaining: 10.4s\n",
      "641:\tlearn: 22.7510065\ttotal: 18.5s\tremaining: 10.3s\n",
      "642:\tlearn: 22.7459132\ttotal: 18.6s\tremaining: 10.3s\n",
      "643:\tlearn: 22.7422799\ttotal: 18.6s\tremaining: 10.3s\n",
      "644:\tlearn: 22.7370214\ttotal: 18.6s\tremaining: 10.2s\n",
      "645:\tlearn: 22.7344558\ttotal: 18.6s\tremaining: 10.2s\n",
      "646:\tlearn: 22.7341428\ttotal: 18.7s\tremaining: 10.2s\n",
      "647:\tlearn: 22.7311581\ttotal: 18.7s\tremaining: 10.2s\n",
      "648:\tlearn: 22.7309018\ttotal: 18.7s\tremaining: 10.1s\n",
      "649:\tlearn: 22.7224317\ttotal: 18.7s\tremaining: 10.1s\n",
      "650:\tlearn: 22.7221333\ttotal: 18.8s\tremaining: 10.1s\n",
      "651:\tlearn: 22.7182389\ttotal: 18.8s\tremaining: 10s\n",
      "652:\tlearn: 22.7159235\ttotal: 18.8s\tremaining: 10s\n",
      "653:\tlearn: 22.7128998\ttotal: 18.9s\tremaining: 9.99s\n",
      "654:\tlearn: 22.7098690\ttotal: 18.9s\tremaining: 9.95s\n",
      "655:\tlearn: 22.7079055\ttotal: 18.9s\tremaining: 9.91s\n",
      "656:\tlearn: 22.7032998\ttotal: 19s\tremaining: 9.92s\n",
      "657:\tlearn: 22.6990763\ttotal: 19s\tremaining: 9.88s\n",
      "658:\tlearn: 22.6944033\ttotal: 19s\tremaining: 9.84s\n",
      "659:\tlearn: 22.6900763\ttotal: 19.1s\tremaining: 9.83s\n",
      "660:\tlearn: 22.6852678\ttotal: 19.1s\tremaining: 9.79s\n",
      "661:\tlearn: 22.6831796\ttotal: 19.1s\tremaining: 9.75s\n",
      "662:\tlearn: 22.6786614\ttotal: 19.1s\tremaining: 9.71s\n",
      "663:\tlearn: 22.6732536\ttotal: 19.2s\tremaining: 9.71s\n",
      "664:\tlearn: 22.6648700\ttotal: 19.2s\tremaining: 9.67s\n",
      "665:\tlearn: 22.6576553\ttotal: 19.2s\tremaining: 9.63s\n",
      "666:\tlearn: 22.6515341\ttotal: 19.3s\tremaining: 9.62s\n",
      "667:\tlearn: 22.6470321\ttotal: 19.3s\tremaining: 9.58s\n",
      "668:\tlearn: 22.6440303\ttotal: 19.3s\tremaining: 9.54s\n",
      "669:\tlearn: 22.6369348\ttotal: 19.4s\tremaining: 9.54s\n",
      "670:\tlearn: 22.6315377\ttotal: 19.4s\tremaining: 9.5s\n",
      "671:\tlearn: 22.6291666\ttotal: 19.4s\tremaining: 9.48s\n",
      "672:\tlearn: 22.6270854\ttotal: 19.5s\tremaining: 9.46s\n",
      "673:\tlearn: 22.6203053\ttotal: 19.5s\tremaining: 9.42s\n",
      "674:\tlearn: 22.6153901\ttotal: 19.5s\tremaining: 9.38s\n",
      "675:\tlearn: 22.6109846\ttotal: 19.5s\tremaining: 9.34s\n",
      "676:\tlearn: 22.6055491\ttotal: 19.5s\tremaining: 9.3s\n",
      "677:\tlearn: 22.6033139\ttotal: 19.6s\tremaining: 9.3s\n",
      "678:\tlearn: 22.6009706\ttotal: 19.6s\tremaining: 9.26s\n",
      "679:\tlearn: 22.5978466\ttotal: 19.6s\tremaining: 9.22s\n",
      "680:\tlearn: 22.5932739\ttotal: 19.7s\tremaining: 9.22s\n",
      "681:\tlearn: 22.5888137\ttotal: 19.7s\tremaining: 9.19s\n",
      "682:\tlearn: 22.5875853\ttotal: 19.7s\tremaining: 9.15s\n",
      "683:\tlearn: 22.5848843\ttotal: 19.8s\tremaining: 9.13s\n",
      "684:\tlearn: 22.5808822\ttotal: 19.8s\tremaining: 9.1s\n",
      "685:\tlearn: 22.5761212\ttotal: 19.8s\tremaining: 9.06s\n",
      "686:\tlearn: 22.5725371\ttotal: 19.9s\tremaining: 9.05s\n",
      "687:\tlearn: 22.5676334\ttotal: 19.9s\tremaining: 9.02s\n",
      "688:\tlearn: 22.5649605\ttotal: 19.9s\tremaining: 8.97s\n",
      "689:\tlearn: 22.5605504\ttotal: 19.9s\tremaining: 8.94s\n",
      "690:\tlearn: 22.5573483\ttotal: 19.9s\tremaining: 8.9s\n",
      "691:\tlearn: 22.5539928\ttotal: 19.9s\tremaining: 8.86s\n",
      "692:\tlearn: 22.5498499\ttotal: 20s\tremaining: 8.85s\n",
      "693:\tlearn: 22.5419544\ttotal: 20s\tremaining: 8.81s\n",
      "694:\tlearn: 22.5412484\ttotal: 20s\tremaining: 8.77s\n",
      "695:\tlearn: 22.5340559\ttotal: 20s\tremaining: 8.73s\n",
      "696:\tlearn: 22.5304428\ttotal: 20.1s\tremaining: 8.72s\n",
      "697:\tlearn: 22.5231595\ttotal: 20.1s\tremaining: 8.69s\n",
      "698:\tlearn: 22.5175756\ttotal: 20.1s\tremaining: 8.65s\n",
      "699:\tlearn: 22.5133591\ttotal: 20.2s\tremaining: 8.64s\n",
      "700:\tlearn: 22.5058889\ttotal: 20.2s\tremaining: 8.61s\n",
      "701:\tlearn: 22.5032916\ttotal: 20.2s\tremaining: 8.57s\n",
      "702:\tlearn: 22.5008680\ttotal: 20.2s\tremaining: 8.53s\n",
      "703:\tlearn: 22.4967015\ttotal: 20.3s\tremaining: 8.52s\n",
      "704:\tlearn: 22.4918502\ttotal: 20.3s\tremaining: 8.48s\n",
      "705:\tlearn: 22.4877476\ttotal: 20.3s\tremaining: 8.44s\n",
      "706:\tlearn: 22.4833761\ttotal: 20.3s\tremaining: 8.43s\n",
      "707:\tlearn: 22.4769900\ttotal: 20.4s\tremaining: 8.4s\n",
      "708:\tlearn: 22.4733961\ttotal: 20.4s\tremaining: 8.36s\n",
      "709:\tlearn: 22.4714173\ttotal: 20.4s\tremaining: 8.32s\n",
      "710:\tlearn: 22.4678156\ttotal: 20.5s\tremaining: 8.32s\n",
      "711:\tlearn: 22.4604040\ttotal: 20.5s\tremaining: 8.28s\n",
      "712:\tlearn: 22.4570827\ttotal: 20.5s\tremaining: 8.24s\n",
      "713:\tlearn: 22.4528869\ttotal: 20.5s\tremaining: 8.23s\n",
      "714:\tlearn: 22.4469954\ttotal: 20.6s\tremaining: 8.21s\n",
      "715:\tlearn: 22.4445340\ttotal: 20.6s\tremaining: 8.17s\n",
      "716:\tlearn: 22.4397463\ttotal: 20.6s\tremaining: 8.13s\n",
      "717:\tlearn: 22.4310797\ttotal: 20.6s\tremaining: 8.11s\n",
      "718:\tlearn: 22.4268991\ttotal: 20.7s\tremaining: 8.07s\n",
      "719:\tlearn: 22.4211244\ttotal: 20.7s\tremaining: 8.03s\n",
      "720:\tlearn: 22.4160414\ttotal: 20.7s\tremaining: 8s\n",
      "721:\tlearn: 22.4104463\ttotal: 20.7s\tremaining: 7.99s\n",
      "722:\tlearn: 22.4070826\ttotal: 20.8s\tremaining: 7.95s\n",
      "723:\tlearn: 22.4045093\ttotal: 20.8s\tremaining: 7.94s\n",
      "724:\tlearn: 22.3976797\ttotal: 20.8s\tremaining: 7.91s\n",
      "725:\tlearn: 22.3973893\ttotal: 20.9s\tremaining: 7.87s\n",
      "726:\tlearn: 22.3942474\ttotal: 20.9s\tremaining: 7.83s\n",
      "727:\tlearn: 22.3869455\ttotal: 20.9s\tremaining: 7.82s\n",
      "728:\tlearn: 22.3810716\ttotal: 20.9s\tremaining: 7.79s\n",
      "729:\tlearn: 22.3808172\ttotal: 21s\tremaining: 7.75s\n",
      "730:\tlearn: 22.3770462\ttotal: 21s\tremaining: 7.71s\n",
      "731:\tlearn: 22.3727697\ttotal: 21s\tremaining: 7.7s\n",
      "732:\tlearn: 22.3673344\ttotal: 21.1s\tremaining: 7.67s\n",
      "733:\tlearn: 22.3629103\ttotal: 21.1s\tremaining: 7.63s\n",
      "734:\tlearn: 22.3605379\ttotal: 21.1s\tremaining: 7.62s\n",
      "735:\tlearn: 22.3581918\ttotal: 21.1s\tremaining: 7.58s\n",
      "736:\tlearn: 22.3539904\ttotal: 21.2s\tremaining: 7.55s\n",
      "737:\tlearn: 22.3497440\ttotal: 21.2s\tremaining: 7.54s\n",
      "738:\tlearn: 22.3444018\ttotal: 21.2s\tremaining: 7.5s\n",
      "739:\tlearn: 22.3385386\ttotal: 21.2s\tremaining: 7.46s\n",
      "740:\tlearn: 22.3362828\ttotal: 21.3s\tremaining: 7.45s\n",
      "741:\tlearn: 22.3308171\ttotal: 21.3s\tremaining: 7.42s\n",
      "742:\tlearn: 22.3270875\ttotal: 21.3s\tremaining: 7.38s\n",
      "743:\tlearn: 22.3235771\ttotal: 21.4s\tremaining: 7.35s\n",
      "744:\tlearn: 22.3202935\ttotal: 21.4s\tremaining: 7.33s\n",
      "745:\tlearn: 22.3155092\ttotal: 21.4s\tremaining: 7.3s\n",
      "746:\tlearn: 22.3130263\ttotal: 21.4s\tremaining: 7.26s\n",
      "747:\tlearn: 22.3096618\ttotal: 21.4s\tremaining: 7.22s\n",
      "748:\tlearn: 22.3066431\ttotal: 21.5s\tremaining: 7.21s\n",
      "749:\tlearn: 22.2999457\ttotal: 21.5s\tremaining: 7.18s\n",
      "750:\tlearn: 22.2921290\ttotal: 21.5s\tremaining: 7.14s\n",
      "751:\tlearn: 22.2859594\ttotal: 21.5s\tremaining: 7.1s\n",
      "752:\tlearn: 22.2780533\ttotal: 21.6s\tremaining: 7.09s\n",
      "753:\tlearn: 22.2746522\ttotal: 21.6s\tremaining: 7.05s\n",
      "754:\tlearn: 22.2687679\ttotal: 21.6s\tremaining: 7.02s\n",
      "755:\tlearn: 22.2619333\ttotal: 21.6s\tremaining: 6.98s\n",
      "756:\tlearn: 22.2575204\ttotal: 21.7s\tremaining: 6.97s\n",
      "757:\tlearn: 22.2505384\ttotal: 21.7s\tremaining: 6.94s\n",
      "758:\tlearn: 22.2482260\ttotal: 21.7s\tremaining: 6.9s\n",
      "759:\tlearn: 22.2429987\ttotal: 21.8s\tremaining: 6.89s\n",
      "760:\tlearn: 22.2384710\ttotal: 21.8s\tremaining: 6.85s\n",
      "761:\tlearn: 22.2334777\ttotal: 21.8s\tremaining: 6.82s\n",
      "762:\tlearn: 22.2310744\ttotal: 21.9s\tremaining: 6.8s\n",
      "763:\tlearn: 22.2256621\ttotal: 21.9s\tremaining: 6.77s\n",
      "764:\tlearn: 22.2229154\ttotal: 21.9s\tremaining: 6.74s\n",
      "765:\tlearn: 22.2190182\ttotal: 22s\tremaining: 6.72s\n",
      "766:\tlearn: 22.2175422\ttotal: 22s\tremaining: 6.69s\n",
      "767:\tlearn: 22.2135967\ttotal: 22s\tremaining: 6.65s\n",
      "768:\tlearn: 22.2109366\ttotal: 22s\tremaining: 6.62s\n",
      "769:\tlearn: 22.2057969\ttotal: 22.1s\tremaining: 6.6s\n",
      "770:\tlearn: 22.2028868\ttotal: 22.1s\tremaining: 6.57s\n",
      "771:\tlearn: 22.2000117\ttotal: 22.1s\tremaining: 6.53s\n",
      "772:\tlearn: 22.1922238\ttotal: 22.1s\tremaining: 6.5s\n",
      "773:\tlearn: 22.1848620\ttotal: 22.2s\tremaining: 6.48s\n",
      "774:\tlearn: 22.1800186\ttotal: 22.2s\tremaining: 6.45s\n",
      "775:\tlearn: 22.1771414\ttotal: 22.2s\tremaining: 6.42s\n",
      "776:\tlearn: 22.1696890\ttotal: 22.2s\tremaining: 6.38s\n",
      "777:\tlearn: 22.1671099\ttotal: 22.3s\tremaining: 6.37s\n",
      "778:\tlearn: 22.1651170\ttotal: 22.3s\tremaining: 6.33s\n",
      "779:\tlearn: 22.1624220\ttotal: 22.3s\tremaining: 6.3s\n",
      "780:\tlearn: 22.1603125\ttotal: 22.4s\tremaining: 6.29s\n",
      "781:\tlearn: 22.1562495\ttotal: 22.4s\tremaining: 6.25s\n",
      "782:\tlearn: 22.1525868\ttotal: 22.4s\tremaining: 6.21s\n",
      "783:\tlearn: 22.1474376\ttotal: 22.5s\tremaining: 6.2s\n",
      "784:\tlearn: 22.1398933\ttotal: 22.5s\tremaining: 6.16s\n",
      "785:\tlearn: 22.1377402\ttotal: 22.5s\tremaining: 6.13s\n",
      "786:\tlearn: 22.1327150\ttotal: 22.6s\tremaining: 6.12s\n",
      "787:\tlearn: 22.1270763\ttotal: 22.6s\tremaining: 6.09s\n",
      "788:\tlearn: 22.1238022\ttotal: 22.6s\tremaining: 6.05s\n",
      "789:\tlearn: 22.1160952\ttotal: 22.7s\tremaining: 6.03s\n",
      "790:\tlearn: 22.1158374\ttotal: 22.7s\tremaining: 6s\n",
      "791:\tlearn: 22.1138371\ttotal: 22.7s\tremaining: 5.96s\n",
      "792:\tlearn: 22.1080495\ttotal: 22.7s\tremaining: 5.93s\n",
      "793:\tlearn: 22.1025583\ttotal: 22.8s\tremaining: 5.91s\n",
      "794:\tlearn: 22.1008811\ttotal: 22.8s\tremaining: 5.88s\n",
      "795:\tlearn: 22.0958185\ttotal: 22.8s\tremaining: 5.84s\n",
      "796:\tlearn: 22.0928357\ttotal: 22.8s\tremaining: 5.81s\n",
      "797:\tlearn: 22.0864455\ttotal: 22.9s\tremaining: 5.8s\n",
      "798:\tlearn: 22.0825069\ttotal: 22.9s\tremaining: 5.76s\n",
      "799:\tlearn: 22.0769173\ttotal: 22.9s\tremaining: 5.73s\n",
      "800:\tlearn: 22.0723620\ttotal: 23s\tremaining: 5.71s\n",
      "801:\tlearn: 22.0686196\ttotal: 23s\tremaining: 5.68s\n",
      "802:\tlearn: 22.0656122\ttotal: 23s\tremaining: 5.64s\n",
      "803:\tlearn: 22.0600079\ttotal: 23s\tremaining: 5.61s\n",
      "804:\tlearn: 22.0541650\ttotal: 23.1s\tremaining: 5.59s\n",
      "805:\tlearn: 22.0484789\ttotal: 23.1s\tremaining: 5.56s\n",
      "806:\tlearn: 22.0482505\ttotal: 23.1s\tremaining: 5.53s\n",
      "807:\tlearn: 22.0433122\ttotal: 23.1s\tremaining: 5.49s\n",
      "808:\tlearn: 22.0395950\ttotal: 23.2s\tremaining: 5.47s\n",
      "809:\tlearn: 22.0355334\ttotal: 23.2s\tremaining: 5.44s\n",
      "810:\tlearn: 22.0322415\ttotal: 23.2s\tremaining: 5.41s\n",
      "811:\tlearn: 22.0297669\ttotal: 23.3s\tremaining: 5.39s\n",
      "812:\tlearn: 22.0264699\ttotal: 23.3s\tremaining: 5.36s\n",
      "813:\tlearn: 22.0215023\ttotal: 23.3s\tremaining: 5.32s\n",
      "814:\tlearn: 22.0181180\ttotal: 23.4s\tremaining: 5.3s\n",
      "815:\tlearn: 22.0151365\ttotal: 23.4s\tremaining: 5.27s\n",
      "816:\tlearn: 22.0084103\ttotal: 23.4s\tremaining: 5.24s\n",
      "817:\tlearn: 22.0038192\ttotal: 23.4s\tremaining: 5.21s\n",
      "818:\tlearn: 22.0007941\ttotal: 23.5s\tremaining: 5.19s\n",
      "819:\tlearn: 21.9957114\ttotal: 23.5s\tremaining: 5.16s\n",
      "820:\tlearn: 21.9910808\ttotal: 23.5s\tremaining: 5.12s\n",
      "821:\tlearn: 21.9871657\ttotal: 23.6s\tremaining: 5.1s\n",
      "822:\tlearn: 21.9842941\ttotal: 23.6s\tremaining: 5.07s\n",
      "823:\tlearn: 21.9813042\ttotal: 23.6s\tremaining: 5.04s\n",
      "824:\tlearn: 21.9748544\ttotal: 23.6s\tremaining: 5s\n",
      "825:\tlearn: 21.9705868\ttotal: 23.7s\tremaining: 4.99s\n",
      "826:\tlearn: 21.9674552\ttotal: 23.7s\tremaining: 4.96s\n",
      "827:\tlearn: 21.9614422\ttotal: 23.7s\tremaining: 4.92s\n",
      "828:\tlearn: 21.9564296\ttotal: 23.8s\tremaining: 4.9s\n",
      "829:\tlearn: 21.9511998\ttotal: 23.8s\tremaining: 4.87s\n",
      "830:\tlearn: 21.9471400\ttotal: 23.8s\tremaining: 4.84s\n",
      "831:\tlearn: 21.9438072\ttotal: 23.8s\tremaining: 4.8s\n",
      "832:\tlearn: 21.9387816\ttotal: 23.9s\tremaining: 4.78s\n",
      "833:\tlearn: 21.9355493\ttotal: 23.9s\tremaining: 4.75s\n",
      "834:\tlearn: 21.9315412\ttotal: 23.9s\tremaining: 4.72s\n",
      "835:\tlearn: 21.9288162\ttotal: 24s\tremaining: 4.7s\n",
      "836:\tlearn: 21.9256432\ttotal: 24s\tremaining: 4.67s\n",
      "837:\tlearn: 21.9207570\ttotal: 24s\tremaining: 4.64s\n",
      "838:\tlearn: 21.9169771\ttotal: 24s\tremaining: 4.6s\n",
      "839:\tlearn: 21.9126733\ttotal: 24.1s\tremaining: 4.58s\n",
      "840:\tlearn: 21.9113259\ttotal: 24.1s\tremaining: 4.55s\n",
      "841:\tlearn: 21.9098911\ttotal: 24.1s\tremaining: 4.52s\n",
      "842:\tlearn: 21.9080488\ttotal: 24.2s\tremaining: 4.5s\n",
      "843:\tlearn: 21.9044298\ttotal: 24.2s\tremaining: 4.47s\n",
      "844:\tlearn: 21.8984481\ttotal: 24.2s\tremaining: 4.43s\n",
      "845:\tlearn: 21.8927869\ttotal: 24.2s\tremaining: 4.4s\n",
      "846:\tlearn: 21.8887921\ttotal: 24.3s\tremaining: 4.38s\n",
      "847:\tlearn: 21.8822248\ttotal: 24.3s\tremaining: 4.35s\n",
      "848:\tlearn: 21.8780890\ttotal: 24.3s\tremaining: 4.32s\n",
      "849:\tlearn: 21.8732354\ttotal: 24.3s\tremaining: 4.28s\n",
      "850:\tlearn: 21.8666855\ttotal: 24.4s\tremaining: 4.26s\n",
      "851:\tlearn: 21.8624492\ttotal: 24.4s\tremaining: 4.23s\n",
      "852:\tlearn: 21.8576472\ttotal: 24.4s\tremaining: 4.2s\n",
      "853:\tlearn: 21.8533135\ttotal: 24.4s\tremaining: 4.17s\n",
      "854:\tlearn: 21.8497116\ttotal: 24.5s\tremaining: 4.15s\n",
      "855:\tlearn: 21.8466134\ttotal: 24.5s\tremaining: 4.12s\n",
      "856:\tlearn: 21.8425780\ttotal: 24.5s\tremaining: 4.08s\n",
      "857:\tlearn: 21.8399821\ttotal: 24.6s\tremaining: 4.06s\n",
      "858:\tlearn: 21.8320539\ttotal: 24.6s\tremaining: 4.03s\n",
      "859:\tlearn: 21.8318245\ttotal: 24.6s\tremaining: 4s\n",
      "860:\tlearn: 21.8270883\ttotal: 24.6s\tremaining: 3.97s\n",
      "861:\tlearn: 21.8209258\ttotal: 24.6s\tremaining: 3.94s\n",
      "862:\tlearn: 21.8195130\ttotal: 24.7s\tremaining: 3.92s\n",
      "863:\tlearn: 21.8150827\ttotal: 24.7s\tremaining: 3.88s\n",
      "864:\tlearn: 21.8132794\ttotal: 24.7s\tremaining: 3.85s\n",
      "865:\tlearn: 21.8088206\ttotal: 24.8s\tremaining: 3.83s\n",
      "866:\tlearn: 21.8051134\ttotal: 24.8s\tremaining: 3.8s\n",
      "867:\tlearn: 21.7993067\ttotal: 24.8s\tremaining: 3.77s\n",
      "868:\tlearn: 21.7959425\ttotal: 24.8s\tremaining: 3.73s\n",
      "869:\tlearn: 21.7898051\ttotal: 24.8s\tremaining: 3.71s\n",
      "870:\tlearn: 21.7871916\ttotal: 24.9s\tremaining: 3.68s\n",
      "871:\tlearn: 21.7812342\ttotal: 24.9s\tremaining: 3.65s\n",
      "872:\tlearn: 21.7779012\ttotal: 24.9s\tremaining: 3.63s\n",
      "873:\tlearn: 21.7755464\ttotal: 25s\tremaining: 3.6s\n",
      "874:\tlearn: 21.7707327\ttotal: 25s\tremaining: 3.57s\n",
      "875:\tlearn: 21.7662085\ttotal: 25s\tremaining: 3.54s\n",
      "876:\tlearn: 21.7634954\ttotal: 25s\tremaining: 3.51s\n",
      "877:\tlearn: 21.7591841\ttotal: 25.1s\tremaining: 3.48s\n",
      "878:\tlearn: 21.7556374\ttotal: 25.1s\tremaining: 3.45s\n",
      "879:\tlearn: 21.7554406\ttotal: 25.1s\tremaining: 3.43s\n",
      "880:\tlearn: 21.7499445\ttotal: 25.2s\tremaining: 3.4s\n",
      "881:\tlearn: 21.7459497\ttotal: 25.2s\tremaining: 3.37s\n",
      "882:\tlearn: 21.7424692\ttotal: 25.2s\tremaining: 3.33s\n",
      "883:\tlearn: 21.7385737\ttotal: 25.2s\tremaining: 3.31s\n",
      "884:\tlearn: 21.7349894\ttotal: 25.2s\tremaining: 3.28s\n",
      "885:\tlearn: 21.7309170\ttotal: 25.3s\tremaining: 3.26s\n",
      "886:\tlearn: 21.7261655\ttotal: 25.3s\tremaining: 3.23s\n",
      "887:\tlearn: 21.7231604\ttotal: 25.4s\tremaining: 3.2s\n",
      "888:\tlearn: 21.7179791\ttotal: 25.4s\tremaining: 3.17s\n",
      "889:\tlearn: 21.7155892\ttotal: 25.4s\tremaining: 3.15s\n",
      "890:\tlearn: 21.7125068\ttotal: 25.5s\tremaining: 3.12s\n",
      "891:\tlearn: 21.7053880\ttotal: 25.5s\tremaining: 3.09s\n",
      "892:\tlearn: 21.7023932\ttotal: 25.6s\tremaining: 3.06s\n",
      "893:\tlearn: 21.6986057\ttotal: 25.6s\tremaining: 3.04s\n",
      "894:\tlearn: 21.6911535\ttotal: 25.6s\tremaining: 3.01s\n",
      "895:\tlearn: 21.6852571\ttotal: 25.6s\tremaining: 2.98s\n",
      "896:\tlearn: 21.6815165\ttotal: 25.6s\tremaining: 2.94s\n",
      "897:\tlearn: 21.6769299\ttotal: 25.7s\tremaining: 2.91s\n",
      "898:\tlearn: 21.6740050\ttotal: 25.7s\tremaining: 2.89s\n",
      "899:\tlearn: 21.6712041\ttotal: 25.7s\tremaining: 2.86s\n",
      "900:\tlearn: 21.6692571\ttotal: 25.8s\tremaining: 2.83s\n",
      "901:\tlearn: 21.6647970\ttotal: 25.8s\tremaining: 2.81s\n",
      "902:\tlearn: 21.6613533\ttotal: 25.8s\tremaining: 2.78s\n",
      "903:\tlearn: 21.6557460\ttotal: 25.9s\tremaining: 2.75s\n",
      "904:\tlearn: 21.6527224\ttotal: 25.9s\tremaining: 2.72s\n",
      "905:\tlearn: 21.6495819\ttotal: 25.9s\tremaining: 2.69s\n",
      "906:\tlearn: 21.6462977\ttotal: 25.9s\tremaining: 2.66s\n",
      "907:\tlearn: 21.6429764\ttotal: 25.9s\tremaining: 2.63s\n",
      "908:\tlearn: 21.6398958\ttotal: 26s\tremaining: 2.6s\n",
      "909:\tlearn: 21.6341480\ttotal: 26s\tremaining: 2.57s\n",
      "910:\tlearn: 21.6293982\ttotal: 26s\tremaining: 2.54s\n",
      "911:\tlearn: 21.6257417\ttotal: 26s\tremaining: 2.51s\n",
      "912:\tlearn: 21.6238104\ttotal: 26.1s\tremaining: 2.49s\n",
      "913:\tlearn: 21.6215246\ttotal: 26.1s\tremaining: 2.46s\n",
      "914:\tlearn: 21.6161990\ttotal: 26.1s\tremaining: 2.43s\n",
      "915:\tlearn: 21.6127515\ttotal: 26.2s\tremaining: 2.4s\n",
      "916:\tlearn: 21.6089153\ttotal: 26.2s\tremaining: 2.37s\n",
      "917:\tlearn: 21.6037474\ttotal: 26.2s\tremaining: 2.34s\n",
      "918:\tlearn: 21.6020216\ttotal: 26.2s\tremaining: 2.31s\n",
      "919:\tlearn: 21.5988695\ttotal: 26.3s\tremaining: 2.29s\n",
      "920:\tlearn: 21.5948986\ttotal: 26.3s\tremaining: 2.26s\n",
      "921:\tlearn: 21.5910962\ttotal: 26.3s\tremaining: 2.23s\n",
      "922:\tlearn: 21.5889100\ttotal: 26.4s\tremaining: 2.2s\n",
      "923:\tlearn: 21.5873962\ttotal: 26.4s\tremaining: 2.17s\n",
      "924:\tlearn: 21.5826099\ttotal: 26.4s\tremaining: 2.14s\n",
      "925:\tlearn: 21.5766069\ttotal: 26.5s\tremaining: 2.12s\n",
      "926:\tlearn: 21.5749959\ttotal: 26.5s\tremaining: 2.09s\n",
      "927:\tlearn: 21.5717232\ttotal: 26.5s\tremaining: 2.06s\n",
      "928:\tlearn: 21.5687599\ttotal: 26.5s\tremaining: 2.03s\n",
      "929:\tlearn: 21.5644352\ttotal: 26.6s\tremaining: 2s\n",
      "930:\tlearn: 21.5597202\ttotal: 26.6s\tremaining: 1.97s\n",
      "931:\tlearn: 21.5574532\ttotal: 26.6s\tremaining: 1.94s\n",
      "932:\tlearn: 21.5492351\ttotal: 26.7s\tremaining: 1.92s\n",
      "933:\tlearn: 21.5464736\ttotal: 26.7s\tremaining: 1.89s\n",
      "934:\tlearn: 21.5442694\ttotal: 26.7s\tremaining: 1.86s\n",
      "935:\tlearn: 21.5387498\ttotal: 26.7s\tremaining: 1.83s\n",
      "936:\tlearn: 21.5365608\ttotal: 26.8s\tremaining: 1.8s\n",
      "937:\tlearn: 21.5335452\ttotal: 26.8s\tremaining: 1.77s\n",
      "938:\tlearn: 21.5289197\ttotal: 26.8s\tremaining: 1.74s\n",
      "939:\tlearn: 21.5271402\ttotal: 26.9s\tremaining: 1.72s\n",
      "940:\tlearn: 21.5230777\ttotal: 26.9s\tremaining: 1.69s\n",
      "941:\tlearn: 21.5157120\ttotal: 26.9s\tremaining: 1.66s\n",
      "942:\tlearn: 21.5123832\ttotal: 26.9s\tremaining: 1.63s\n",
      "943:\tlearn: 21.5068692\ttotal: 27s\tremaining: 1.6s\n",
      "944:\tlearn: 21.5020967\ttotal: 27s\tremaining: 1.57s\n",
      "945:\tlearn: 21.4979119\ttotal: 27s\tremaining: 1.54s\n",
      "946:\tlearn: 21.4956597\ttotal: 27s\tremaining: 1.51s\n",
      "947:\tlearn: 21.4927323\ttotal: 27.1s\tremaining: 1.49s\n",
      "948:\tlearn: 21.4867537\ttotal: 27.1s\tremaining: 1.46s\n",
      "949:\tlearn: 21.4800133\ttotal: 27.1s\tremaining: 1.43s\n",
      "950:\tlearn: 21.4773162\ttotal: 27.2s\tremaining: 1.4s\n",
      "951:\tlearn: 21.4759428\ttotal: 27.2s\tremaining: 1.37s\n",
      "952:\tlearn: 21.4729279\ttotal: 27.2s\tremaining: 1.34s\n",
      "953:\tlearn: 21.4656918\ttotal: 27.3s\tremaining: 1.31s\n",
      "954:\tlearn: 21.4584189\ttotal: 27.3s\tremaining: 1.29s\n",
      "955:\tlearn: 21.4532292\ttotal: 27.3s\tremaining: 1.26s\n",
      "956:\tlearn: 21.4490745\ttotal: 27.3s\tremaining: 1.23s\n",
      "957:\tlearn: 21.4423979\ttotal: 27.3s\tremaining: 1.2s\n",
      "958:\tlearn: 21.4386558\ttotal: 27.4s\tremaining: 1.17s\n",
      "959:\tlearn: 21.4356597\ttotal: 27.4s\tremaining: 1.14s\n",
      "960:\tlearn: 21.4328519\ttotal: 27.5s\tremaining: 1.11s\n",
      "961:\tlearn: 21.4262412\ttotal: 27.5s\tremaining: 1.08s\n",
      "962:\tlearn: 21.4217842\ttotal: 27.5s\tremaining: 1.06s\n",
      "963:\tlearn: 21.4202561\ttotal: 27.6s\tremaining: 1.03s\n",
      "964:\tlearn: 21.4164481\ttotal: 27.6s\tremaining: 1s\n",
      "965:\tlearn: 21.4123895\ttotal: 27.6s\tremaining: 972ms\n",
      "966:\tlearn: 21.4094553\ttotal: 27.6s\tremaining: 942ms\n",
      "967:\tlearn: 21.4058983\ttotal: 27.7s\tremaining: 915ms\n",
      "968:\tlearn: 21.4039994\ttotal: 27.7s\tremaining: 886ms\n",
      "969:\tlearn: 21.3999401\ttotal: 27.7s\tremaining: 857ms\n",
      "970:\tlearn: 21.3968608\ttotal: 27.8s\tremaining: 830ms\n",
      "971:\tlearn: 21.3897258\ttotal: 27.8s\tremaining: 801ms\n",
      "972:\tlearn: 21.3864099\ttotal: 27.8s\tremaining: 771ms\n",
      "973:\tlearn: 21.3838320\ttotal: 27.9s\tremaining: 744ms\n",
      "974:\tlearn: 21.3788350\ttotal: 27.9s\tremaining: 715ms\n",
      "975:\tlearn: 21.3726689\ttotal: 27.9s\tremaining: 686ms\n",
      "976:\tlearn: 21.3697108\ttotal: 27.9s\tremaining: 657ms\n",
      "977:\tlearn: 21.3695102\ttotal: 28s\tremaining: 630ms\n",
      "978:\tlearn: 21.3637786\ttotal: 28s\tremaining: 601ms\n",
      "979:\tlearn: 21.3579072\ttotal: 28s\tremaining: 571ms\n",
      "980:\tlearn: 21.3508159\ttotal: 28s\tremaining: 542ms\n",
      "981:\tlearn: 21.3459463\ttotal: 28.1s\tremaining: 515ms\n",
      "982:\tlearn: 21.3421886\ttotal: 28.1s\tremaining: 486ms\n",
      "983:\tlearn: 21.3381736\ttotal: 28.2s\tremaining: 458ms\n",
      "984:\tlearn: 21.3332893\ttotal: 28.2s\tremaining: 429ms\n",
      "985:\tlearn: 21.3302411\ttotal: 28.2s\tremaining: 400ms\n",
      "986:\tlearn: 21.3254099\ttotal: 28.3s\tremaining: 372ms\n",
      "987:\tlearn: 21.3229474\ttotal: 28.3s\tremaining: 344ms\n",
      "988:\tlearn: 21.3205393\ttotal: 28.3s\tremaining: 315ms\n",
      "989:\tlearn: 21.3203126\ttotal: 28.3s\tremaining: 286ms\n",
      "990:\tlearn: 21.3159429\ttotal: 28.4s\tremaining: 258ms\n",
      "991:\tlearn: 21.3125009\ttotal: 28.4s\tremaining: 229ms\n",
      "992:\tlearn: 21.3088990\ttotal: 28.4s\tremaining: 200ms\n",
      "993:\tlearn: 21.3058006\ttotal: 28.4s\tremaining: 171ms\n",
      "994:\tlearn: 21.3020594\ttotal: 28.5s\tremaining: 143ms\n",
      "995:\tlearn: 21.3006097\ttotal: 28.5s\tremaining: 114ms\n",
      "996:\tlearn: 21.2994328\ttotal: 28.5s\tremaining: 85.7ms\n",
      "997:\tlearn: 21.2949400\ttotal: 28.5s\tremaining: 57.1ms\n",
      "998:\tlearn: 21.2909928\ttotal: 28.5s\tremaining: 28.5ms\n",
      "999:\tlearn: 21.2888085\ttotal: 28.6s\tremaining: 0us\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<catboost.core.CatBoostRegressor at 0x7f746f606190>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(train_x, train_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "44.840442826499604"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr_test_pred = model.predict(test_x)\n",
    "mse= mean_squared_error(test_y, lr_test_pred) \n",
    "mse**0.5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Testing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After all I achieved the RMSE of 44.84 which is less then 48, also I concluded that the number of orders tends to increase during summer. Also, the number of orders tends to be increasing during the evening and peaks at midnight."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Regarding the models, I performed fit/predict onto three different models, linearregression model, random forest regressor and CatBoost, out of these three models I achived the lowest RMSE with the CatBoost model, however it also took more time, on the other hand LR model was the quickest with RMSE of 49.73, lastly random forrest regressor achieved RMSE of 46.23, which is lower then required 48 and a bit faster then CatBoost."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}