model_dev.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pylab as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from model import get_model, get_train_test_pools, get_precision_recall, load_model, get_model_path\n",
    "from catboost import CatBoostRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# path = get_model_path(N, DIRECTION)\n",
    "# model = load_model(path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from prepare_X import get_featrues, specs\n",
    "from turnpt_analysis import get_df_turnpt_measures\n",
    "\n",
    "\n",
    "import pickle, io\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from catboost import CatBoostClassifier, Pool\n",
    "from catboost import CatBoostRegressor\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "transformed OHLCV at frequency 1D\n",
      "reading cache of TXF in 1D OHLCV\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/cyyen/miniconda3/envs/stock_path_tracker/lib/python3.8/site-packages/pandas/core/groupby/groupby.py:980: FutureWarning: Sorting because non-concatenation axis is not aligned. A future version\n",
      "of pandas will change to not sort by default.\n",
      "\n",
      "To accept the future behavior, pass 'sort=False'.\n",
      "\n",
      "To retain the current behavior and silence the warning, pass 'sort=True'.\n",
      "\n",
      "  result = concat(values, axis=self.axis, keys=keys)\n",
      "<ipython-input-50-510885712db5>:16: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df_chosen_y[DIRECTION] = np.where(df_chosen_y[DIRECTION]==-1, 2, df_chosen_y[DIRECTION])\n"
     ]
    }
   ],
   "source": [
    "# X\n",
    "CACHE=True\n",
    "ADD_DIFF = 0\n",
    "PCT_TRAIN = 0.75\n",
    "df_features = get_featrues(specs, cache=CACHE, add_diff=ADD_DIFF)\n",
    "df_features.index = df_features.index.date\n",
    "df_features = df_features.shift(1)\n",
    "\n",
    "# y\n",
    "df_y = get_df_turnpt_measures(N).fillna(0)\n",
    "df_y.index = df_y.tx_datetime\n",
    "\n",
    "df_chosen_y = df_y[[DIRECTION]] # before the close of t, the agent will decide whether to enter the market by predicting if t close is a turning point \n",
    "assert type(df_chosen_y.index.values[0]) == type(df_features.index.values[0])\n",
    "\n",
    "df_chosen_y[DIRECTION] = np.where(df_chosen_y[DIRECTION]==-1, 2, df_chosen_y[DIRECTION])\n",
    "\n",
    "### DATA ###\n",
    "df = df_chosen_y.join(df_features).dropna()\n",
    "w = (\n",
    "df[[DIRECTION]].join(df_y.prc_diff).\n",
    "assign(weight = lambda x: np.where(x.is_turnpt!=0, x.prc_diff.abs(), np.nan)).\n",
    "assign(weight = lambda x: 1+x.weight.fillna(0)).\n",
    "#     assign(weight = lambda x: x.weight.fillna(x.weight.quantile(WEIGHTS_QUNATILE))).\n",
    "weight\n",
    ")\n",
    "num_train = int(df.shape[0] * PCT_TRAIN)# DONE\n",
    "\n",
    "y = df[DIRECTION] # DONE\n",
    "y_train = y[: num_train]\n",
    "y_test = y[num_train:]\n",
    "\n",
    "w_train = w[: num_train]\n",
    "w_test = w[num_train:]\n",
    "\n",
    "X = df.iloc[:, 1:]\n",
    "X_train = X.iloc[:num_train, :]\n",
    "X_test = X.iloc[num_train:, :]\n",
    "\n",
    "train_pool = Pool(X_train, y_train, weight=w_train)\n",
    "test_pool = Pool(X_test, y_test, weight=w_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 0 ns, sys: 3 µs, total: 3 µs\n",
      "Wall time: 5.72 µs\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# train_pool, test_pool =  get_train_test_pools(N,DIRECTION, WEIGHTS_QUNATILE, CACHE=True)\n",
    "# best_iter = next(map(lambda x: len(x[1])-1, model.eval_metrics(train_pool,['Precision']).items()))\n",
    "# model.plot_tree(best_iter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 1.0944834\ttest: 1.0969475\tbest: 1.0969475 (0)\ttotal: 421ms\tremaining: 11h 41m 3s\n",
      "1:\tlearn: 1.0893820\ttest: 1.0959807\tbest: 1.0959807 (1)\ttotal: 830ms\tremaining: 11h 31m 23s\n",
      "2:\tlearn: 1.0834324\ttest: 1.0950932\tbest: 1.0950932 (2)\ttotal: 1.25s\tremaining: 11h 34m 3s\n",
      "3:\tlearn: 1.0792036\ttest: 1.0953233\tbest: 1.0950932 (2)\ttotal: 1.66s\tremaining: 11h 31m\n",
      "4:\tlearn: 1.0741568\ttest: 1.0953330\tbest: 1.0950932 (2)\ttotal: 2.06s\tremaining: 11h 28m 25s\n",
      "5:\tlearn: 1.0674854\ttest: 1.0886660\tbest: 1.0886660 (5)\ttotal: 2.48s\tremaining: 11h 28m 20s\n",
      "6:\tlearn: 1.0631603\ttest: 1.0872763\tbest: 1.0872763 (6)\ttotal: 2.89s\tremaining: 11h 28m 45s\n",
      "7:\tlearn: 1.0561685\ttest: 1.0830492\tbest: 1.0830492 (7)\ttotal: 3.31s\tremaining: 11h 29m 9s\n",
      "8:\tlearn: 1.0506670\ttest: 1.0830408\tbest: 1.0830408 (8)\ttotal: 3.74s\tremaining: 11h 33m 22s\n",
      "9:\tlearn: 1.0458663\ttest: 1.0801252\tbest: 1.0801252 (9)\ttotal: 4.18s\tremaining: 11h 35m 56s\n",
      "10:\tlearn: 1.0408145\ttest: 1.0787712\tbest: 1.0787712 (10)\ttotal: 4.59s\tremaining: 11h 35m 43s\n",
      "11:\tlearn: 1.0364189\ttest: 1.0778559\tbest: 1.0778559 (11)\ttotal: 5.01s\tremaining: 11h 35m 17s\n",
      "12:\tlearn: 1.0306154\ttest: 1.0778018\tbest: 1.0778018 (12)\ttotal: 5.42s\tremaining: 11h 34m 38s\n",
      "13:\tlearn: 1.0258331\ttest: 1.0760384\tbest: 1.0760384 (13)\ttotal: 5.83s\tremaining: 11h 33m 59s\n",
      "14:\tlearn: 1.0215159\ttest: 1.0755603\tbest: 1.0755603 (14)\ttotal: 6.24s\tremaining: 11h 33m 35s\n",
      "15:\tlearn: 1.0167449\ttest: 1.0727770\tbest: 1.0727770 (15)\ttotal: 6.65s\tremaining: 11h 32m 35s\n",
      "16:\tlearn: 1.0125005\ttest: 1.0719477\tbest: 1.0719477 (16)\ttotal: 7.07s\tremaining: 11h 33m 6s\n",
      "17:\tlearn: 1.0085495\ttest: 1.0705397\tbest: 1.0705397 (17)\ttotal: 7.48s\tremaining: 11h 32m 21s\n",
      "18:\tlearn: 1.0046644\ttest: 1.0701710\tbest: 1.0701710 (18)\ttotal: 7.88s\tremaining: 11h 31m 31s\n",
      "19:\tlearn: 0.9994137\ttest: 1.0699227\tbest: 1.0699227 (19)\ttotal: 8.31s\tremaining: 11h 32m 4s\n",
      "20:\tlearn: 0.9953555\ttest: 1.0699128\tbest: 1.0699128 (20)\ttotal: 8.73s\tremaining: 11h 32m 24s\n",
      "21:\tlearn: 0.9911514\ttest: 1.0694196\tbest: 1.0694196 (21)\ttotal: 9.14s\tremaining: 11h 32m 42s\n",
      "22:\tlearn: 0.9872742\ttest: 1.0693077\tbest: 1.0693077 (22)\ttotal: 9.55s\tremaining: 11h 32m 12s\n",
      "23:\tlearn: 0.9835776\ttest: 1.0688984\tbest: 1.0688984 (23)\ttotal: 9.98s\tremaining: 11h 32m 50s\n",
      "24:\tlearn: 0.9799830\ttest: 1.0678772\tbest: 1.0678772 (24)\ttotal: 10.4s\tremaining: 11h 33m 15s\n",
      "25:\tlearn: 0.9758182\ttest: 1.0673631\tbest: 1.0673631 (25)\ttotal: 10.8s\tremaining: 11h 32m 56s\n",
      "26:\tlearn: 0.9719306\ttest: 1.0660422\tbest: 1.0660422 (26)\ttotal: 11.2s\tremaining: 11h 33m 15s\n",
      "27:\tlearn: 0.9676533\ttest: 1.0648511\tbest: 1.0648511 (27)\ttotal: 11.6s\tremaining: 11h 33m 2s\n",
      "28:\tlearn: 0.9649077\ttest: 1.0639779\tbest: 1.0639779 (28)\ttotal: 12.1s\tremaining: 11h 32m 54s\n",
      "29:\tlearn: 0.9607773\ttest: 1.0616563\tbest: 1.0616563 (29)\ttotal: 12.5s\tremaining: 11h 32m 32s\n",
      "30:\tlearn: 0.9573092\ttest: 1.0613545\tbest: 1.0613545 (30)\ttotal: 12.9s\tremaining: 11h 32m 18s\n",
      "31:\tlearn: 0.9536048\ttest: 1.0597168\tbest: 1.0597168 (31)\ttotal: 13.3s\tremaining: 11h 32m 29s\n",
      "32:\tlearn: 0.9502897\ttest: 1.0595675\tbest: 1.0595675 (32)\ttotal: 13.7s\tremaining: 11h 32m 54s\n",
      "33:\tlearn: 0.9465012\ttest: 1.0587882\tbest: 1.0587882 (33)\ttotal: 14.1s\tremaining: 11h 33m 16s\n",
      "34:\tlearn: 0.9429238\ttest: 1.0579165\tbest: 1.0579165 (34)\ttotal: 14.6s\tremaining: 11h 33m 47s\n",
      "35:\tlearn: 0.9383849\ttest: 1.0565389\tbest: 1.0565389 (35)\ttotal: 15s\tremaining: 11h 35m 6s\n",
      "36:\tlearn: 0.9355190\ttest: 1.0555769\tbest: 1.0555769 (36)\ttotal: 15.4s\tremaining: 11h 34m 44s\n",
      "37:\tlearn: 0.9310114\ttest: 1.0533933\tbest: 1.0533933 (37)\ttotal: 15.9s\tremaining: 11h 35m 1s\n",
      "38:\tlearn: 0.9275618\ttest: 1.0521265\tbest: 1.0521265 (38)\ttotal: 16.3s\tremaining: 11h 35m 14s\n",
      "39:\tlearn: 0.9243157\ttest: 1.0521439\tbest: 1.0521265 (38)\ttotal: 16.7s\tremaining: 11h 35m 34s\n",
      "40:\tlearn: 0.9214965\ttest: 1.0514586\tbest: 1.0514586 (40)\ttotal: 17.1s\tremaining: 11h 35m 20s\n",
      "41:\tlearn: 0.9184963\ttest: 1.0508829\tbest: 1.0508829 (41)\ttotal: 17.5s\tremaining: 11h 35m 14s\n",
      "42:\tlearn: 0.9155394\ttest: 1.0500915\tbest: 1.0500915 (42)\ttotal: 17.9s\tremaining: 11h 35m 14s\n",
      "43:\tlearn: 0.9125283\ttest: 1.0490662\tbest: 1.0490662 (43)\ttotal: 18.4s\tremaining: 11h 35m 9s\n",
      "44:\tlearn: 0.9075059\ttest: 1.0466961\tbest: 1.0466961 (44)\ttotal: 18.8s\tremaining: 11h 35m 1s\n",
      "45:\tlearn: 0.9022904\ttest: 1.0464405\tbest: 1.0464405 (45)\ttotal: 19.2s\tremaining: 11h 35m 9s\n",
      "46:\tlearn: 0.8992399\ttest: 1.0457199\tbest: 1.0457199 (46)\ttotal: 19.6s\tremaining: 11h 35m 53s\n",
      "47:\tlearn: 0.8963679\ttest: 1.0457128\tbest: 1.0457128 (47)\ttotal: 20s\tremaining: 11h 35m 49s\n",
      "48:\tlearn: 0.8935755\ttest: 1.0445949\tbest: 1.0445949 (48)\ttotal: 20.5s\tremaining: 11h 35m 51s\n",
      "49:\tlearn: 0.8902603\ttest: 1.0444960\tbest: 1.0444960 (49)\ttotal: 20.9s\tremaining: 11h 35m 58s\n",
      "50:\tlearn: 0.8876276\ttest: 1.0448151\tbest: 1.0444960 (49)\ttotal: 21.3s\tremaining: 11h 36m 2s\n",
      "51:\tlearn: 0.8848781\ttest: 1.0449058\tbest: 1.0444960 (49)\ttotal: 21.7s\tremaining: 11h 36m 11s\n",
      "52:\tlearn: 0.8827438\ttest: 1.0440046\tbest: 1.0440046 (52)\ttotal: 22.1s\tremaining: 11h 35m 51s\n",
      "53:\tlearn: 0.8805918\ttest: 1.0432236\tbest: 1.0432236 (53)\ttotal: 22.5s\tremaining: 11h 35m 27s\n",
      "54:\tlearn: 0.8777701\ttest: 1.0430479\tbest: 1.0430479 (54)\ttotal: 23s\tremaining: 11h 35m 19s\n",
      "55:\tlearn: 0.8752101\ttest: 1.0425754\tbest: 1.0425754 (55)\ttotal: 23.4s\tremaining: 11h 35m 22s\n",
      "56:\tlearn: 0.8726401\ttest: 1.0423557\tbest: 1.0423557 (56)\ttotal: 23.8s\tremaining: 11h 35m 23s\n",
      "57:\tlearn: 0.8697758\ttest: 1.0420492\tbest: 1.0420492 (57)\ttotal: 24.2s\tremaining: 11h 35m 34s\n",
      "58:\tlearn: 0.8663229\ttest: 1.0413459\tbest: 1.0413459 (58)\ttotal: 24.6s\tremaining: 11h 35m 55s\n",
      "59:\tlearn: 0.8638409\ttest: 1.0406610\tbest: 1.0406610 (59)\ttotal: 25.1s\tremaining: 11h 35m 57s\n",
      "60:\tlearn: 0.8612020\ttest: 1.0396909\tbest: 1.0396909 (60)\ttotal: 25.5s\tremaining: 11h 35m 58s\n",
      "61:\tlearn: 0.8585243\ttest: 1.0395162\tbest: 1.0395162 (61)\ttotal: 25.9s\tremaining: 11h 36m 11s\n",
      "62:\tlearn: 0.8557564\ttest: 1.0394908\tbest: 1.0394908 (62)\ttotal: 26.3s\tremaining: 11h 36m 1s\n",
      "63:\tlearn: 0.8534451\ttest: 1.0395648\tbest: 1.0394908 (62)\ttotal: 26.7s\tremaining: 11h 35m 59s\n",
      "64:\tlearn: 0.8511879\ttest: 1.0387930\tbest: 1.0387930 (64)\ttotal: 27.2s\tremaining: 11h 35m 53s\n",
      "65:\tlearn: 0.8485811\ttest: 1.0371197\tbest: 1.0371197 (65)\ttotal: 27.6s\tremaining: 11h 35m 25s\n",
      "66:\tlearn: 0.8458043\ttest: 1.0371055\tbest: 1.0371055 (66)\ttotal: 28s\tremaining: 11h 35m 24s\n",
      "67:\tlearn: 0.8437346\ttest: 1.0362662\tbest: 1.0362662 (67)\ttotal: 28.4s\tremaining: 11h 35m 20s\n",
      "68:\tlearn: 0.8404598\ttest: 1.0351187\tbest: 1.0351187 (68)\ttotal: 28.8s\tremaining: 11h 35m 23s\n",
      "69:\tlearn: 0.8376518\ttest: 1.0327885\tbest: 1.0327885 (69)\ttotal: 29.2s\tremaining: 11h 35m 31s\n",
      "70:\tlearn: 0.8353011\ttest: 1.0321185\tbest: 1.0321185 (70)\ttotal: 29.6s\tremaining: 11h 35m 29s\n",
      "71:\tlearn: 0.8327706\ttest: 1.0317264\tbest: 1.0317264 (71)\ttotal: 30.1s\tremaining: 11h 35m 26s\n",
      "72:\tlearn: 0.8303334\ttest: 1.0306527\tbest: 1.0306527 (72)\ttotal: 30.5s\tremaining: 11h 35m 38s\n",
      "73:\tlearn: 0.8279731\ttest: 1.0306110\tbest: 1.0306110 (73)\ttotal: 30.9s\tremaining: 11h 35m 38s\n",
      "74:\tlearn: 0.8248611\ttest: 1.0296805\tbest: 1.0296805 (74)\ttotal: 31.3s\tremaining: 11h 35m 42s\n",
      "75:\tlearn: 0.8219752\ttest: 1.0271451\tbest: 1.0271451 (75)\ttotal: 31.8s\tremaining: 11h 35m 51s\n",
      "76:\tlearn: 0.8191299\ttest: 1.0254377\tbest: 1.0254377 (76)\ttotal: 32.2s\tremaining: 11h 35m 58s\n",
      "77:\tlearn: 0.8167229\ttest: 1.0238885\tbest: 1.0238885 (77)\ttotal: 32.6s\tremaining: 11h 35m 57s\n",
      "78:\tlearn: 0.8147049\ttest: 1.0238298\tbest: 1.0238298 (78)\ttotal: 33s\tremaining: 11h 35m 53s\n",
      "79:\tlearn: 0.8123095\ttest: 1.0238458\tbest: 1.0238298 (78)\ttotal: 33.4s\tremaining: 11h 35m 55s\n",
      "80:\tlearn: 0.8102657\ttest: 1.0225126\tbest: 1.0225126 (80)\ttotal: 33.9s\tremaining: 11h 36m\n",
      "81:\tlearn: 0.8075498\ttest: 1.0220036\tbest: 1.0220036 (81)\ttotal: 34.3s\tremaining: 11h 36m 12s\n",
      "82:\tlearn: 0.8054746\ttest: 1.0218457\tbest: 1.0218457 (82)\ttotal: 34.7s\tremaining: 11h 36m 12s\n",
      "83:\tlearn: 0.8021768\ttest: 1.0201074\tbest: 1.0201074 (83)\ttotal: 35.1s\tremaining: 11h 36m 17s\n",
      "84:\tlearn: 0.8000210\ttest: 1.0194279\tbest: 1.0194279 (84)\ttotal: 35.5s\tremaining: 11h 36m 23s\n",
      "85:\tlearn: 0.7973375\ttest: 1.0185015\tbest: 1.0185015 (85)\ttotal: 36s\tremaining: 11h 36m 18s\n",
      "86:\tlearn: 0.7950889\ttest: 1.0184540\tbest: 1.0184540 (86)\ttotal: 36.4s\tremaining: 11h 36m 27s\n",
      "87:\tlearn: 0.7931501\ttest: 1.0183166\tbest: 1.0183166 (87)\ttotal: 36.8s\tremaining: 11h 36m 27s\n",
      "88:\tlearn: 0.7908579\ttest: 1.0170111\tbest: 1.0170111 (88)\ttotal: 37.2s\tremaining: 11h 36m 21s\n",
      "89:\tlearn: 0.7886089\ttest: 1.0169588\tbest: 1.0169588 (89)\ttotal: 37.6s\tremaining: 11h 36m 13s\n",
      "90:\tlearn: 0.7865530\ttest: 1.0166393\tbest: 1.0166393 (90)\ttotal: 38s\tremaining: 11h 36m 3s\n",
      "91:\tlearn: 0.7847764\ttest: 1.0168084\tbest: 1.0166393 (90)\ttotal: 38.5s\tremaining: 11h 36m 6s\n",
      "92:\tlearn: 0.7828046\ttest: 1.0161932\tbest: 1.0161932 (92)\ttotal: 38.9s\tremaining: 11h 36m 1s\n",
      "93:\tlearn: 0.7807122\ttest: 1.0159079\tbest: 1.0159079 (93)\ttotal: 39.3s\tremaining: 11h 36m 27s\n",
      "94:\tlearn: 0.7787035\ttest: 1.0157468\tbest: 1.0157468 (94)\ttotal: 39.7s\tremaining: 11h 36m 17s\n",
      "95:\tlearn: 0.7766359\ttest: 1.0144674\tbest: 1.0144674 (95)\ttotal: 40.1s\tremaining: 11h 36m 11s\n",
      "96:\tlearn: 0.7744229\ttest: 1.0142983\tbest: 1.0142983 (96)\ttotal: 40.6s\tremaining: 11h 36m 18s\n",
      "97:\tlearn: 0.7719623\ttest: 1.0134032\tbest: 1.0134032 (97)\ttotal: 41s\tremaining: 11h 36m 15s\n",
      "98:\tlearn: 0.7699865\ttest: 1.0130088\tbest: 1.0130088 (98)\ttotal: 41.4s\tremaining: 11h 36m 17s\n",
      "99:\tlearn: 0.7679048\ttest: 1.0125379\tbest: 1.0125379 (99)\ttotal: 41.8s\tremaining: 11h 36m 15s\n",
      "100:\tlearn: 0.7657587\ttest: 1.0120983\tbest: 1.0120983 (100)\ttotal: 42.2s\tremaining: 11h 36m 9s\n",
      "101:\tlearn: 0.7639898\ttest: 1.0120200\tbest: 1.0120200 (101)\ttotal: 42.6s\tremaining: 11h 36m 3s\n",
      "102:\tlearn: 0.7621314\ttest: 1.0112029\tbest: 1.0112029 (102)\ttotal: 43.1s\tremaining: 11h 35m 57s\n",
      "103:\tlearn: 0.7605800\ttest: 1.0109001\tbest: 1.0109001 (103)\ttotal: 43.5s\tremaining: 11h 36m 1s\n",
      "104:\tlearn: 0.7587200\ttest: 1.0103810\tbest: 1.0103810 (104)\ttotal: 43.9s\tremaining: 11h 36m 14s\n",
      "105:\tlearn: 0.7565032\ttest: 1.0098921\tbest: 1.0098921 (105)\ttotal: 44.3s\tremaining: 11h 36m 10s\n",
      "106:\tlearn: 0.7541989\ttest: 1.0095667\tbest: 1.0095667 (106)\ttotal: 44.7s\tremaining: 11h 36m 5s\n",
      "107:\tlearn: 0.7523229\ttest: 1.0096798\tbest: 1.0095667 (106)\ttotal: 45.2s\tremaining: 11h 36m 10s\n",
      "108:\tlearn: 0.7501493\ttest: 1.0090061\tbest: 1.0090061 (108)\ttotal: 45.6s\tremaining: 11h 36m 32s\n",
      "109:\tlearn: 0.7484419\ttest: 1.0084125\tbest: 1.0084125 (109)\ttotal: 46s\tremaining: 11h 36m 36s\n",
      "110:\tlearn: 0.7465232\ttest: 1.0078420\tbest: 1.0078420 (110)\ttotal: 46.5s\tremaining: 11h 36m 53s\n",
      "111:\tlearn: 0.7444775\ttest: 1.0067195\tbest: 1.0067195 (111)\ttotal: 46.9s\tremaining: 11h 36m 54s\n",
      "112:\tlearn: 0.7428173\ttest: 1.0060122\tbest: 1.0060122 (112)\ttotal: 47.3s\tremaining: 11h 36m 51s\n",
      "113:\tlearn: 0.7409254\ttest: 1.0060469\tbest: 1.0060122 (112)\ttotal: 47.7s\tremaining: 11h 36m 41s\n",
      "114:\tlearn: 0.7391737\ttest: 1.0055639\tbest: 1.0055639 (114)\ttotal: 48.1s\tremaining: 11h 36m 22s\n",
      "115:\tlearn: 0.7377531\ttest: 1.0051640\tbest: 1.0051640 (115)\ttotal: 48.5s\tremaining: 11h 36m 25s\n",
      "116:\tlearn: 0.7360612\ttest: 1.0045137\tbest: 1.0045137 (116)\ttotal: 49s\tremaining: 11h 36m 31s\n",
      "117:\tlearn: 0.7340067\ttest: 1.0038445\tbest: 1.0038445 (117)\ttotal: 49.4s\tremaining: 11h 36m 22s\n",
      "118:\tlearn: 0.7325483\ttest: 1.0034205\tbest: 1.0034205 (118)\ttotal: 49.8s\tremaining: 11h 36m 16s\n",
      "119:\tlearn: 0.7303293\ttest: 1.0036518\tbest: 1.0034205 (118)\ttotal: 50.2s\tremaining: 11h 36m 26s\n",
      "120:\tlearn: 0.7289033\ttest: 1.0036841\tbest: 1.0034205 (118)\ttotal: 50.6s\tremaining: 11h 36m 20s\n",
      "121:\tlearn: 0.7267588\ttest: 1.0037573\tbest: 1.0034205 (118)\ttotal: 51s\tremaining: 11h 36m 13s\n",
      "122:\tlearn: 0.7253182\ttest: 1.0032793\tbest: 1.0032793 (122)\ttotal: 51.4s\tremaining: 11h 36m 13s\n",
      "123:\tlearn: 0.7231910\ttest: 1.0017709\tbest: 1.0017709 (123)\ttotal: 51.9s\tremaining: 11h 36m 13s\n",
      "124:\tlearn: 0.7214365\ttest: 1.0020540\tbest: 1.0017709 (123)\ttotal: 52.3s\tremaining: 11h 36m 15s\n",
      "125:\tlearn: 0.7197110\ttest: 1.0013968\tbest: 1.0013968 (125)\ttotal: 52.7s\tremaining: 11h 36m 6s\n",
      "126:\tlearn: 0.7180482\ttest: 1.0007796\tbest: 1.0007796 (126)\ttotal: 53.1s\tremaining: 11h 35m 52s\n",
      "127:\tlearn: 0.7167352\ttest: 0.9999971\tbest: 0.9999971 (127)\ttotal: 53.5s\tremaining: 11h 35m 52s\n",
      "128:\tlearn: 0.7152348\ttest: 0.9996609\tbest: 0.9996609 (128)\ttotal: 53.9s\tremaining: 11h 36m\n",
      "129:\tlearn: 0.7133908\ttest: 0.9990839\tbest: 0.9990839 (129)\ttotal: 54.4s\tremaining: 11h 36m 10s\n",
      "130:\tlearn: 0.7119821\ttest: 0.9983636\tbest: 0.9983636 (130)\ttotal: 54.8s\tremaining: 11h 36m 21s\n",
      "131:\tlearn: 0.7103174\ttest: 0.9977809\tbest: 0.9977809 (131)\ttotal: 55.2s\tremaining: 11h 36m 21s\n",
      "132:\tlearn: 0.7087502\ttest: 0.9973511\tbest: 0.9973511 (132)\ttotal: 55.6s\tremaining: 11h 36m 20s\n",
      "133:\tlearn: 0.7072265\ttest: 0.9968589\tbest: 0.9968589 (133)\ttotal: 56.1s\tremaining: 11h 36m 19s\n",
      "134:\tlearn: 0.7056100\ttest: 0.9966426\tbest: 0.9966426 (134)\ttotal: 56.5s\tremaining: 11h 36m 24s\n",
      "135:\tlearn: 0.7043090\ttest: 0.9963320\tbest: 0.9963320 (135)\ttotal: 56.9s\tremaining: 11h 36m 22s\n",
      "136:\tlearn: 0.7027979\ttest: 0.9956204\tbest: 0.9956204 (136)\ttotal: 57.3s\tremaining: 11h 36m 22s\n",
      "137:\tlearn: 0.7014943\ttest: 0.9949932\tbest: 0.9949932 (137)\ttotal: 57.7s\tremaining: 11h 36m 15s\n",
      "138:\tlearn: 0.7001680\ttest: 0.9949416\tbest: 0.9949416 (138)\ttotal: 58.1s\tremaining: 11h 36m 8s\n",
      "139:\tlearn: 0.6987171\ttest: 0.9946885\tbest: 0.9946885 (139)\ttotal: 58.6s\tremaining: 11h 36m 12s\n",
      "140:\tlearn: 0.6973772\ttest: 0.9944097\tbest: 0.9944097 (140)\ttotal: 59s\tremaining: 11h 36m 5s\n",
      "141:\tlearn: 0.6960186\ttest: 0.9940342\tbest: 0.9940342 (141)\ttotal: 59.4s\tremaining: 11h 35m 54s\n",
      "142:\tlearn: 0.6946210\ttest: 0.9936609\tbest: 0.9936609 (142)\ttotal: 59.8s\tremaining: 11h 35m 45s\n",
      "143:\tlearn: 0.6929969\ttest: 0.9937854\tbest: 0.9936609 (142)\ttotal: 1m\tremaining: 11h 35m 41s\n",
      "144:\tlearn: 0.6915691\ttest: 0.9940893\tbest: 0.9936609 (142)\ttotal: 1m\tremaining: 11h 35m 31s\n",
      "145:\tlearn: 0.6902606\ttest: 0.9935846\tbest: 0.9935846 (145)\ttotal: 1m 1s\tremaining: 11h 35m 26s\n",
      "146:\tlearn: 0.6888608\ttest: 0.9934268\tbest: 0.9934268 (146)\ttotal: 1m 1s\tremaining: 11h 35m 19s\n",
      "147:\tlearn: 0.6876813\ttest: 0.9926894\tbest: 0.9926894 (147)\ttotal: 1m 1s\tremaining: 11h 35m 7s\n",
      "148:\tlearn: 0.6861917\ttest: 0.9918693\tbest: 0.9918693 (148)\ttotal: 1m 2s\tremaining: 11h 34m 58s\n",
      "149:\tlearn: 0.6845502\ttest: 0.9913751\tbest: 0.9913751 (149)\ttotal: 1m 2s\tremaining: 11h 34m 59s\n",
      "150:\tlearn: 0.6829999\ttest: 0.9905159\tbest: 0.9905159 (150)\ttotal: 1m 3s\tremaining: 11h 34m 51s\n",
      "151:\tlearn: 0.6814990\ttest: 0.9899391\tbest: 0.9899391 (151)\ttotal: 1m 3s\tremaining: 11h 34m 44s\n",
      "152:\tlearn: 0.6801263\ttest: 0.9897641\tbest: 0.9897641 (152)\ttotal: 1m 3s\tremaining: 11h 34m 42s\n",
      "153:\tlearn: 0.6785462\ttest: 0.9894048\tbest: 0.9894048 (153)\ttotal: 1m 4s\tremaining: 11h 34m 40s\n",
      "154:\tlearn: 0.6773830\ttest: 0.9895897\tbest: 0.9894048 (153)\ttotal: 1m 4s\tremaining: 11h 34m 30s\n",
      "155:\tlearn: 0.6763811\ttest: 0.9892372\tbest: 0.9892372 (155)\ttotal: 1m 5s\tremaining: 11h 34m 31s\n",
      "156:\tlearn: 0.6750873\ttest: 0.9889603\tbest: 0.9889603 (156)\ttotal: 1m 5s\tremaining: 11h 34m 17s\n",
      "157:\tlearn: 0.6737591\ttest: 0.9887547\tbest: 0.9887547 (157)\ttotal: 1m 5s\tremaining: 11h 34m 13s\n",
      "158:\tlearn: 0.6718596\ttest: 0.9888436\tbest: 0.9887547 (157)\ttotal: 1m 6s\tremaining: 11h 34m 19s\n",
      "159:\tlearn: 0.6704930\ttest: 0.9883226\tbest: 0.9883226 (159)\ttotal: 1m 6s\tremaining: 11h 34m 9s\n",
      "160:\tlearn: 0.6691159\ttest: 0.9878514\tbest: 0.9878514 (160)\ttotal: 1m 7s\tremaining: 11h 34m 8s\n",
      "161:\tlearn: 0.6678293\ttest: 0.9876697\tbest: 0.9876697 (161)\ttotal: 1m 7s\tremaining: 11h 34m 5s\n",
      "162:\tlearn: 0.6664779\ttest: 0.9864672\tbest: 0.9864672 (162)\ttotal: 1m 7s\tremaining: 11h 34m\n",
      "163:\tlearn: 0.6649581\ttest: 0.9856193\tbest: 0.9856193 (163)\ttotal: 1m 8s\tremaining: 11h 33m 54s\n",
      "164:\tlearn: 0.6637336\ttest: 0.9856606\tbest: 0.9856193 (163)\ttotal: 1m 8s\tremaining: 11h 33m 48s\n",
      "165:\tlearn: 0.6622738\ttest: 0.9852728\tbest: 0.9852728 (165)\ttotal: 1m 9s\tremaining: 11h 33m 40s\n",
      "166:\tlearn: 0.6608528\ttest: 0.9846704\tbest: 0.9846704 (166)\ttotal: 1m 9s\tremaining: 11h 33m 41s\n",
      "167:\tlearn: 0.6597295\ttest: 0.9846277\tbest: 0.9846277 (167)\ttotal: 1m 10s\tremaining: 11h 33m 40s\n",
      "168:\tlearn: 0.6584665\ttest: 0.9846789\tbest: 0.9846277 (167)\ttotal: 1m 10s\tremaining: 11h 33m 33s\n",
      "169:\tlearn: 0.6571152\ttest: 0.9842524\tbest: 0.9842524 (169)\ttotal: 1m 10s\tremaining: 11h 33m 28s\n",
      "170:\tlearn: 0.6560324\ttest: 0.9837279\tbest: 0.9837279 (170)\ttotal: 1m 11s\tremaining: 11h 33m 22s\n",
      "171:\tlearn: 0.6547806\ttest: 0.9830870\tbest: 0.9830870 (171)\ttotal: 1m 11s\tremaining: 11h 33m 15s\n",
      "172:\tlearn: 0.6534667\ttest: 0.9830796\tbest: 0.9830796 (172)\ttotal: 1m 12s\tremaining: 11h 33m 15s\n",
      "173:\tlearn: 0.6522635\ttest: 0.9829635\tbest: 0.9829635 (173)\ttotal: 1m 12s\tremaining: 11h 33m 12s\n",
      "174:\tlearn: 0.6508517\ttest: 0.9824776\tbest: 0.9824776 (174)\ttotal: 1m 12s\tremaining: 11h 33m 7s\n",
      "175:\tlearn: 0.6494912\ttest: 0.9821070\tbest: 0.9821070 (175)\ttotal: 1m 13s\tremaining: 11h 33m 1s\n",
      "176:\tlearn: 0.6483173\ttest: 0.9815052\tbest: 0.9815052 (176)\ttotal: 1m 13s\tremaining: 11h 32m 56s\n",
      "177:\tlearn: 0.6466902\ttest: 0.9803719\tbest: 0.9803719 (177)\ttotal: 1m 14s\tremaining: 11h 33m 1s\n",
      "178:\tlearn: 0.6454543\ttest: 0.9799629\tbest: 0.9799629 (178)\ttotal: 1m 14s\tremaining: 11h 33m 4s\n",
      "179:\tlearn: 0.6443089\ttest: 0.9802105\tbest: 0.9799629 (178)\ttotal: 1m 14s\tremaining: 11h 33m 7s\n",
      "180:\tlearn: 0.6430879\ttest: 0.9802449\tbest: 0.9799629 (178)\ttotal: 1m 15s\tremaining: 11h 33m 12s\n",
      "181:\tlearn: 0.6419993\ttest: 0.9801685\tbest: 0.9799629 (178)\ttotal: 1m 15s\tremaining: 11h 33m 10s\n",
      "182:\tlearn: 0.6408921\ttest: 0.9791909\tbest: 0.9791909 (182)\ttotal: 1m 16s\tremaining: 11h 33m 13s\n",
      "183:\tlearn: 0.6396943\ttest: 0.9784805\tbest: 0.9784805 (183)\ttotal: 1m 16s\tremaining: 11h 33m 10s\n",
      "184:\tlearn: 0.6384779\ttest: 0.9781452\tbest: 0.9781452 (184)\ttotal: 1m 17s\tremaining: 11h 33m 14s\n",
      "185:\tlearn: 0.6372137\ttest: 0.9777525\tbest: 0.9777525 (185)\ttotal: 1m 17s\tremaining: 11h 33m 17s\n",
      "186:\tlearn: 0.6360736\ttest: 0.9773017\tbest: 0.9773017 (186)\ttotal: 1m 17s\tremaining: 11h 33m 11s\n",
      "187:\tlearn: 0.6348102\ttest: 0.9765352\tbest: 0.9765352 (187)\ttotal: 1m 18s\tremaining: 11h 33m 7s\n",
      "188:\tlearn: 0.6335857\ttest: 0.9759655\tbest: 0.9759655 (188)\ttotal: 1m 18s\tremaining: 11h 33m 1s\n",
      "189:\tlearn: 0.6323724\ttest: 0.9761309\tbest: 0.9759655 (188)\ttotal: 1m 19s\tremaining: 11h 32m 59s\n",
      "190:\tlearn: 0.6313269\ttest: 0.9755287\tbest: 0.9755287 (190)\ttotal: 1m 19s\tremaining: 11h 33m 9s\n",
      "191:\tlearn: 0.6299343\ttest: 0.9755700\tbest: 0.9755287 (190)\ttotal: 1m 20s\tremaining: 11h 33m 8s\n",
      "192:\tlearn: 0.6286829\ttest: 0.9751150\tbest: 0.9751150 (192)\ttotal: 1m 20s\tremaining: 11h 33m 11s\n",
      "193:\tlearn: 0.6277259\ttest: 0.9747991\tbest: 0.9747991 (193)\ttotal: 1m 20s\tremaining: 11h 33m 11s\n",
      "194:\tlearn: 0.6264422\ttest: 0.9745219\tbest: 0.9745219 (194)\ttotal: 1m 21s\tremaining: 11h 33m 10s\n",
      "195:\tlearn: 0.6249975\ttest: 0.9732823\tbest: 0.9732823 (195)\ttotal: 1m 21s\tremaining: 11h 33m 9s\n",
      "196:\tlearn: 0.6238300\ttest: 0.9729523\tbest: 0.9729523 (196)\ttotal: 1m 22s\tremaining: 11h 33m 8s\n",
      "197:\tlearn: 0.6226530\ttest: 0.9724467\tbest: 0.9724467 (197)\ttotal: 1m 22s\tremaining: 11h 33m 5s\n",
      "198:\tlearn: 0.6213570\ttest: 0.9712055\tbest: 0.9712055 (198)\ttotal: 1m 22s\tremaining: 11h 33m 5s\n",
      "199:\tlearn: 0.6202016\ttest: 0.9707097\tbest: 0.9707097 (199)\ttotal: 1m 23s\tremaining: 11h 33m 3s\n",
      "200:\tlearn: 0.6191290\ttest: 0.9703306\tbest: 0.9703306 (200)\ttotal: 1m 23s\tremaining: 11h 33m 1s\n",
      "201:\tlearn: 0.6181186\ttest: 0.9703218\tbest: 0.9703218 (201)\ttotal: 1m 24s\tremaining: 11h 33m 11s\n",
      "202:\tlearn: 0.6170216\ttest: 0.9700646\tbest: 0.9700646 (202)\ttotal: 1m 24s\tremaining: 11h 33m 12s\n",
      "203:\tlearn: 0.6160666\ttest: 0.9698977\tbest: 0.9698977 (203)\ttotal: 1m 25s\tremaining: 11h 33m 35s\n",
      "204:\tlearn: 0.6147536\ttest: 0.9700121\tbest: 0.9698977 (203)\ttotal: 1m 25s\tremaining: 11h 33m 35s\n",
      "205:\tlearn: 0.6138599\ttest: 0.9699232\tbest: 0.9698977 (203)\ttotal: 1m 25s\tremaining: 11h 33m 38s\n",
      "206:\tlearn: 0.6126877\ttest: 0.9696082\tbest: 0.9696082 (206)\ttotal: 1m 26s\tremaining: 11h 33m 34s\n",
      "207:\tlearn: 0.6114699\ttest: 0.9692061\tbest: 0.9692061 (207)\ttotal: 1m 26s\tremaining: 11h 33m 32s\n",
      "208:\tlearn: 0.6103373\ttest: 0.9686923\tbest: 0.9686923 (208)\ttotal: 1m 27s\tremaining: 11h 33m 36s\n",
      "209:\tlearn: 0.6093145\ttest: 0.9684602\tbest: 0.9684602 (209)\ttotal: 1m 27s\tremaining: 11h 33m 38s\n",
      "210:\tlearn: 0.6084072\ttest: 0.9684020\tbest: 0.9684020 (210)\ttotal: 1m 27s\tremaining: 11h 33m 35s\n",
      "211:\tlearn: 0.6071258\ttest: 0.9681637\tbest: 0.9681637 (211)\ttotal: 1m 28s\tremaining: 11h 33m 31s\n",
      "212:\tlearn: 0.6060714\ttest: 0.9682143\tbest: 0.9681637 (211)\ttotal: 1m 28s\tremaining: 11h 33m 26s\n",
      "213:\tlearn: 0.6051860\ttest: 0.9680462\tbest: 0.9680462 (213)\ttotal: 1m 29s\tremaining: 11h 33m 33s\n",
      "214:\tlearn: 0.6042692\ttest: 0.9678699\tbest: 0.9678699 (214)\ttotal: 1m 29s\tremaining: 11h 33m 35s\n",
      "215:\tlearn: 0.6032222\ttest: 0.9675091\tbest: 0.9675091 (215)\ttotal: 1m 30s\tremaining: 11h 33m 36s\n",
      "216:\tlearn: 0.6020759\ttest: 0.9666322\tbest: 0.9666322 (216)\ttotal: 1m 30s\tremaining: 11h 33m 31s\n",
      "217:\tlearn: 0.6009872\ttest: 0.9664133\tbest: 0.9664133 (217)\ttotal: 1m 30s\tremaining: 11h 33m 42s\n",
      "218:\tlearn: 0.5999473\ttest: 0.9662791\tbest: 0.9662791 (218)\ttotal: 1m 31s\tremaining: 11h 33m 40s\n",
      "219:\tlearn: 0.5990012\ttest: 0.9659223\tbest: 0.9659223 (219)\ttotal: 1m 31s\tremaining: 11h 33m 37s\n",
      "220:\tlearn: 0.5979023\ttest: 0.9657109\tbest: 0.9657109 (220)\ttotal: 1m 32s\tremaining: 11h 33m 40s\n",
      "221:\tlearn: 0.5967277\ttest: 0.9650519\tbest: 0.9650519 (221)\ttotal: 1m 32s\tremaining: 11h 33m 42s\n",
      "222:\tlearn: 0.5957573\ttest: 0.9648911\tbest: 0.9648911 (222)\ttotal: 1m 33s\tremaining: 11h 33m 39s\n",
      "223:\tlearn: 0.5946450\ttest: 0.9645455\tbest: 0.9645455 (223)\ttotal: 1m 33s\tremaining: 11h 33m 35s\n",
      "224:\tlearn: 0.5937005\ttest: 0.9639930\tbest: 0.9639930 (224)\ttotal: 1m 33s\tremaining: 11h 33m 35s\n",
      "225:\tlearn: 0.5927258\ttest: 0.9637439\tbest: 0.9637439 (225)\ttotal: 1m 34s\tremaining: 11h 33m 36s\n",
      "226:\tlearn: 0.5917383\ttest: 0.9630034\tbest: 0.9630034 (226)\ttotal: 1m 34s\tremaining: 11h 33m 37s\n",
      "227:\tlearn: 0.5908597\ttest: 0.9629114\tbest: 0.9629114 (227)\ttotal: 1m 35s\tremaining: 11h 33m 38s\n",
      "228:\tlearn: 0.5901071\ttest: 0.9629115\tbest: 0.9629114 (227)\ttotal: 1m 35s\tremaining: 11h 33m 52s\n",
      "229:\tlearn: 0.5893657\ttest: 0.9627478\tbest: 0.9627478 (229)\ttotal: 1m 35s\tremaining: 11h 33m 56s\n",
      "230:\tlearn: 0.5884036\ttest: 0.9621540\tbest: 0.9621540 (230)\ttotal: 1m 36s\tremaining: 11h 34m 4s\n",
      "231:\tlearn: 0.5876212\ttest: 0.9622809\tbest: 0.9621540 (230)\ttotal: 1m 36s\tremaining: 11h 34m 3s\n",
      "232:\tlearn: 0.5866328\ttest: 0.9613032\tbest: 0.9613032 (232)\ttotal: 1m 37s\tremaining: 11h 34m 6s\n",
      "233:\tlearn: 0.5856628\ttest: 0.9608147\tbest: 0.9608147 (233)\ttotal: 1m 37s\tremaining: 11h 34m\n",
      "234:\tlearn: 0.5849226\ttest: 0.9605458\tbest: 0.9605458 (234)\ttotal: 1m 38s\tremaining: 11h 33m 58s\n",
      "235:\tlearn: 0.5841337\ttest: 0.9602308\tbest: 0.9602308 (235)\ttotal: 1m 38s\tremaining: 11h 33m 53s\n",
      "236:\tlearn: 0.5829782\ttest: 0.9599829\tbest: 0.9599829 (236)\ttotal: 1m 38s\tremaining: 11h 33m 49s\n",
      "237:\tlearn: 0.5819536\ttest: 0.9597929\tbest: 0.9597929 (237)\ttotal: 1m 39s\tremaining: 11h 33m 47s\n",
      "238:\tlearn: 0.5811616\ttest: 0.9593470\tbest: 0.9593470 (238)\ttotal: 1m 39s\tremaining: 11h 33m 46s\n",
      "239:\tlearn: 0.5803064\ttest: 0.9593806\tbest: 0.9593470 (238)\ttotal: 1m 40s\tremaining: 11h 33m 45s\n",
      "240:\tlearn: 0.5795770\ttest: 0.9590715\tbest: 0.9590715 (240)\ttotal: 1m 40s\tremaining: 11h 33m 47s\n",
      "241:\tlearn: 0.5785877\ttest: 0.9586125\tbest: 0.9586125 (241)\ttotal: 1m 41s\tremaining: 11h 33m 55s\n",
      "242:\tlearn: 0.5777770\ttest: 0.9583649\tbest: 0.9583649 (242)\ttotal: 1m 41s\tremaining: 11h 33m 54s\n",
      "243:\tlearn: 0.5768955\ttest: 0.9581407\tbest: 0.9581407 (243)\ttotal: 1m 41s\tremaining: 11h 33m 54s\n",
      "244:\tlearn: 0.5758715\ttest: 0.9571506\tbest: 0.9571506 (244)\ttotal: 1m 42s\tremaining: 11h 33m 56s\n",
      "245:\tlearn: 0.5748350\ttest: 0.9569174\tbest: 0.9569174 (245)\ttotal: 1m 42s\tremaining: 11h 33m 54s\n",
      "246:\tlearn: 0.5737560\ttest: 0.9569298\tbest: 0.9569174 (245)\ttotal: 1m 43s\tremaining: 11h 33m 53s\n",
      "247:\tlearn: 0.5728962\ttest: 0.9568314\tbest: 0.9568314 (247)\ttotal: 1m 43s\tremaining: 11h 33m 50s\n",
      "248:\tlearn: 0.5720659\ttest: 0.9564987\tbest: 0.9564987 (248)\ttotal: 1m 43s\tremaining: 11h 33m 44s\n",
      "249:\tlearn: 0.5713481\ttest: 0.9565181\tbest: 0.9564987 (248)\ttotal: 1m 44s\tremaining: 11h 33m 44s\n",
      "250:\tlearn: 0.5705525\ttest: 0.9559270\tbest: 0.9559270 (250)\ttotal: 1m 44s\tremaining: 11h 33m 45s\n",
      "251:\tlearn: 0.5695931\ttest: 0.9556311\tbest: 0.9556311 (251)\ttotal: 1m 45s\tremaining: 11h 33m 45s\n",
      "252:\tlearn: 0.5687811\ttest: 0.9549284\tbest: 0.9549284 (252)\ttotal: 1m 45s\tremaining: 11h 33m 46s\n",
      "253:\tlearn: 0.5680361\ttest: 0.9545435\tbest: 0.9545435 (253)\ttotal: 1m 46s\tremaining: 11h 33m 45s\n",
      "254:\tlearn: 0.5672909\ttest: 0.9540736\tbest: 0.9540736 (254)\ttotal: 1m 46s\tremaining: 11h 33m 45s\n",
      "255:\tlearn: 0.5662131\ttest: 0.9531016\tbest: 0.9531016 (255)\ttotal: 1m 46s\tremaining: 11h 33m 48s\n",
      "256:\tlearn: 0.5653543\ttest: 0.9528649\tbest: 0.9528649 (256)\ttotal: 1m 47s\tremaining: 11h 33m 48s\n",
      "257:\tlearn: 0.5645661\ttest: 0.9524737\tbest: 0.9524737 (257)\ttotal: 1m 47s\tremaining: 11h 33m 49s\n",
      "258:\tlearn: 0.5637091\ttest: 0.9515559\tbest: 0.9515559 (258)\ttotal: 1m 48s\tremaining: 11h 33m 48s\n",
      "259:\tlearn: 0.5627680\ttest: 0.9515656\tbest: 0.9515559 (258)\ttotal: 1m 48s\tremaining: 11h 33m 46s\n",
      "260:\tlearn: 0.5619517\ttest: 0.9512939\tbest: 0.9512939 (260)\ttotal: 1m 48s\tremaining: 11h 33m 44s\n",
      "261:\tlearn: 0.5611325\ttest: 0.9501543\tbest: 0.9501543 (261)\ttotal: 1m 49s\tremaining: 11h 33m 43s\n",
      "262:\tlearn: 0.5603617\ttest: 0.9500642\tbest: 0.9500642 (262)\ttotal: 1m 49s\tremaining: 11h 33m 42s\n",
      "263:\tlearn: 0.5594947\ttest: 0.9501490\tbest: 0.9500642 (262)\ttotal: 1m 50s\tremaining: 11h 33m 43s\n",
      "264:\tlearn: 0.5586729\ttest: 0.9500889\tbest: 0.9500642 (262)\ttotal: 1m 50s\tremaining: 11h 33m 43s\n",
      "265:\tlearn: 0.5579101\ttest: 0.9498599\tbest: 0.9498599 (265)\ttotal: 1m 51s\tremaining: 11h 33m 38s\n",
      "266:\tlearn: 0.5569477\ttest: 0.9498993\tbest: 0.9498599 (265)\ttotal: 1m 51s\tremaining: 11h 33m 35s\n",
      "267:\tlearn: 0.5561351\ttest: 0.9495454\tbest: 0.9495454 (267)\ttotal: 1m 51s\tremaining: 11h 33m 34s\n",
      "268:\tlearn: 0.5553568\ttest: 0.9496113\tbest: 0.9495454 (267)\ttotal: 1m 52s\tremaining: 11h 33m 33s\n",
      "269:\tlearn: 0.5546544\ttest: 0.9491926\tbest: 0.9491926 (269)\ttotal: 1m 52s\tremaining: 11h 33m 33s\n",
      "270:\tlearn: 0.5539287\ttest: 0.9493127\tbest: 0.9491926 (269)\ttotal: 1m 53s\tremaining: 11h 33m 30s\n",
      "271:\tlearn: 0.5532228\ttest: 0.9490951\tbest: 0.9490951 (271)\ttotal: 1m 53s\tremaining: 11h 33m 27s\n",
      "272:\tlearn: 0.5525543\ttest: 0.9491874\tbest: 0.9490951 (271)\ttotal: 1m 53s\tremaining: 11h 33m 23s\n",
      "273:\tlearn: 0.5518662\ttest: 0.9488141\tbest: 0.9488141 (273)\ttotal: 1m 54s\tremaining: 11h 33m 22s\n",
      "274:\tlearn: 0.5510851\ttest: 0.9483969\tbest: 0.9483969 (274)\ttotal: 1m 54s\tremaining: 11h 33m 21s\n",
      "275:\tlearn: 0.5504286\ttest: 0.9480897\tbest: 0.9480897 (275)\ttotal: 1m 55s\tremaining: 11h 33m 25s\n",
      "276:\tlearn: 0.5496676\ttest: 0.9477373\tbest: 0.9477373 (276)\ttotal: 1m 55s\tremaining: 11h 33m 24s\n",
      "277:\tlearn: 0.5488413\ttest: 0.9476292\tbest: 0.9476292 (277)\ttotal: 1m 55s\tremaining: 11h 33m 27s\n",
      "278:\tlearn: 0.5480705\ttest: 0.9476927\tbest: 0.9476292 (277)\ttotal: 1m 56s\tremaining: 11h 33m 28s\n",
      "279:\tlearn: 0.5473514\ttest: 0.9472435\tbest: 0.9472435 (279)\ttotal: 1m 56s\tremaining: 11h 33m 26s\n",
      "280:\tlearn: 0.5465891\ttest: 0.9471287\tbest: 0.9471287 (280)\ttotal: 1m 57s\tremaining: 11h 33m 27s\n",
      "281:\tlearn: 0.5459698\ttest: 0.9471917\tbest: 0.9471287 (280)\ttotal: 1m 57s\tremaining: 11h 33m 26s\n",
      "282:\tlearn: 0.5452925\ttest: 0.9469940\tbest: 0.9469940 (282)\ttotal: 1m 58s\tremaining: 11h 33m 32s\n",
      "283:\tlearn: 0.5446052\ttest: 0.9465666\tbest: 0.9465666 (283)\ttotal: 1m 58s\tremaining: 11h 33m 26s\n",
      "284:\tlearn: 0.5440189\ttest: 0.9465734\tbest: 0.9465666 (283)\ttotal: 1m 58s\tremaining: 11h 33m 22s\n",
      "285:\tlearn: 0.5432490\ttest: 0.9464032\tbest: 0.9464032 (285)\ttotal: 1m 59s\tremaining: 11h 33m 22s\n",
      "286:\tlearn: 0.5425428\ttest: 0.9464023\tbest: 0.9464023 (286)\ttotal: 1m 59s\tremaining: 11h 33m 22s\n",
      "287:\tlearn: 0.5417965\ttest: 0.9463922\tbest: 0.9463922 (287)\ttotal: 2m\tremaining: 11h 33m 22s\n",
      "288:\tlearn: 0.5411540\ttest: 0.9464206\tbest: 0.9463922 (287)\ttotal: 2m\tremaining: 11h 33m 20s\n",
      "289:\tlearn: 0.5403900\ttest: 0.9463450\tbest: 0.9463450 (289)\ttotal: 2m 1s\tremaining: 11h 33m 27s\n",
      "290:\tlearn: 0.5396769\ttest: 0.9464014\tbest: 0.9463450 (289)\ttotal: 2m 1s\tremaining: 11h 33m 29s\n",
      "291:\tlearn: 0.5389650\ttest: 0.9455794\tbest: 0.9455794 (291)\ttotal: 2m 1s\tremaining: 11h 33m 28s\n",
      "292:\tlearn: 0.5383661\ttest: 0.9454781\tbest: 0.9454781 (292)\ttotal: 2m 2s\tremaining: 11h 33m 29s\n",
      "293:\tlearn: 0.5374579\ttest: 0.9448609\tbest: 0.9448609 (293)\ttotal: 2m 2s\tremaining: 11h 33m 27s\n",
      "294:\tlearn: 0.5366791\ttest: 0.9443428\tbest: 0.9443428 (294)\ttotal: 2m 3s\tremaining: 11h 33m 34s\n",
      "295:\tlearn: 0.5358081\ttest: 0.9443626\tbest: 0.9443428 (294)\ttotal: 2m 3s\tremaining: 11h 33m 31s\n",
      "296:\tlearn: 0.5352136\ttest: 0.9443389\tbest: 0.9443389 (296)\ttotal: 2m 3s\tremaining: 11h 33m 26s\n",
      "297:\tlearn: 0.5345906\ttest: 0.9443528\tbest: 0.9443389 (296)\ttotal: 2m 4s\tremaining: 11h 33m 25s\n",
      "298:\tlearn: 0.5339531\ttest: 0.9439058\tbest: 0.9439058 (298)\ttotal: 2m 4s\tremaining: 11h 33m 26s\n",
      "299:\tlearn: 0.5333587\ttest: 0.9439550\tbest: 0.9439058 (298)\ttotal: 2m 5s\tremaining: 11h 33m 23s\n",
      "300:\tlearn: 0.5326590\ttest: 0.9436757\tbest: 0.9436757 (300)\ttotal: 2m 5s\tremaining: 11h 33m 22s\n",
      "301:\tlearn: 0.5320588\ttest: 0.9436999\tbest: 0.9436757 (300)\ttotal: 2m 6s\tremaining: 11h 33m 22s\n",
      "302:\tlearn: 0.5314251\ttest: 0.9434418\tbest: 0.9434418 (302)\ttotal: 2m 6s\tremaining: 11h 33m 23s\n",
      "303:\tlearn: 0.5307486\ttest: 0.9433662\tbest: 0.9433662 (303)\ttotal: 2m 6s\tremaining: 11h 33m 23s\n",
      "304:\tlearn: 0.5300216\ttest: 0.9429481\tbest: 0.9429481 (304)\ttotal: 2m 7s\tremaining: 11h 33m 20s\n",
      "305:\tlearn: 0.5293445\ttest: 0.9430041\tbest: 0.9429481 (304)\ttotal: 2m 7s\tremaining: 11h 33m 18s\n",
      "306:\tlearn: 0.5286520\ttest: 0.9429986\tbest: 0.9429481 (304)\ttotal: 2m 8s\tremaining: 11h 33m 21s\n",
      "307:\tlearn: 0.5279605\ttest: 0.9425886\tbest: 0.9425886 (307)\ttotal: 2m 8s\tremaining: 11h 33m 18s\n",
      "308:\tlearn: 0.5273223\ttest: 0.9423032\tbest: 0.9423032 (308)\ttotal: 2m 8s\tremaining: 11h 33m 15s\n",
      "309:\tlearn: 0.5266473\ttest: 0.9418797\tbest: 0.9418797 (309)\ttotal: 2m 9s\tremaining: 11h 33m 12s\n",
      "310:\tlearn: 0.5260531\ttest: 0.9416943\tbest: 0.9416943 (310)\ttotal: 2m 9s\tremaining: 11h 33m 14s\n",
      "311:\tlearn: 0.5253744\ttest: 0.9417037\tbest: 0.9416943 (310)\ttotal: 2m 10s\tremaining: 11h 33m 14s\n",
      "312:\tlearn: 0.5247047\ttest: 0.9416854\tbest: 0.9416854 (312)\ttotal: 2m 10s\tremaining: 11h 33m 15s\n",
      "313:\tlearn: 0.5240650\ttest: 0.9415678\tbest: 0.9415678 (313)\ttotal: 2m 11s\tremaining: 11h 33m 15s\n",
      "314:\tlearn: 0.5234687\ttest: 0.9414774\tbest: 0.9414774 (314)\ttotal: 2m 11s\tremaining: 11h 33m 17s\n",
      "315:\tlearn: 0.5227967\ttest: 0.9414069\tbest: 0.9414069 (315)\ttotal: 2m 11s\tremaining: 11h 33m 20s\n",
      "316:\tlearn: 0.5222171\ttest: 0.9414658\tbest: 0.9414069 (315)\ttotal: 2m 12s\tremaining: 11h 33m 21s\n",
      "317:\tlearn: 0.5216330\ttest: 0.9410070\tbest: 0.9410070 (317)\ttotal: 2m 12s\tremaining: 11h 33m 22s\n",
      "318:\tlearn: 0.5208115\ttest: 0.9401785\tbest: 0.9401785 (318)\ttotal: 2m 13s\tremaining: 11h 33m 25s\n",
      "319:\tlearn: 0.5200773\ttest: 0.9402309\tbest: 0.9401785 (318)\ttotal: 2m 13s\tremaining: 11h 33m 25s\n",
      "320:\tlearn: 0.5195028\ttest: 0.9404540\tbest: 0.9401785 (318)\ttotal: 2m 13s\tremaining: 11h 33m 21s\n",
      "321:\tlearn: 0.5187054\ttest: 0.9400634\tbest: 0.9400634 (321)\ttotal: 2m 14s\tremaining: 11h 33m 19s\n",
      "322:\tlearn: 0.5182036\ttest: 0.9399346\tbest: 0.9399346 (322)\ttotal: 2m 14s\tremaining: 11h 33m 19s\n",
      "323:\tlearn: 0.5174577\ttest: 0.9398863\tbest: 0.9398863 (323)\ttotal: 2m 15s\tremaining: 11h 33m 23s\n",
      "324:\tlearn: 0.5168573\ttest: 0.9395472\tbest: 0.9395472 (324)\ttotal: 2m 15s\tremaining: 11h 33m 20s\n",
      "325:\tlearn: 0.5162958\ttest: 0.9393500\tbest: 0.9393500 (325)\ttotal: 2m 16s\tremaining: 11h 33m 21s\n",
      "326:\tlearn: 0.5155796\ttest: 0.9391881\tbest: 0.9391881 (326)\ttotal: 2m 16s\tremaining: 11h 33m 19s\n",
      "327:\tlearn: 0.5149537\ttest: 0.9390723\tbest: 0.9390723 (327)\ttotal: 2m 16s\tremaining: 11h 33m 17s\n",
      "328:\tlearn: 0.5143326\ttest: 0.9390493\tbest: 0.9390493 (328)\ttotal: 2m 17s\tremaining: 11h 33m 13s\n",
      "329:\tlearn: 0.5135962\ttest: 0.9388748\tbest: 0.9388748 (329)\ttotal: 2m 17s\tremaining: 11h 33m 9s\n",
      "330:\tlearn: 0.5129710\ttest: 0.9385688\tbest: 0.9385688 (330)\ttotal: 2m 18s\tremaining: 11h 33m 7s\n",
      "331:\tlearn: 0.5123809\ttest: 0.9379455\tbest: 0.9379455 (331)\ttotal: 2m 18s\tremaining: 11h 33m 4s\n",
      "332:\tlearn: 0.5117519\ttest: 0.9375143\tbest: 0.9375143 (332)\ttotal: 2m 18s\tremaining: 11h 33m 3s\n",
      "333:\tlearn: 0.5111615\ttest: 0.9373536\tbest: 0.9373536 (333)\ttotal: 2m 19s\tremaining: 11h 33m 4s\n",
      "334:\tlearn: 0.5104836\ttest: 0.9371583\tbest: 0.9371583 (334)\ttotal: 2m 19s\tremaining: 11h 33m 2s\n",
      "335:\tlearn: 0.5099604\ttest: 0.9370094\tbest: 0.9370094 (335)\ttotal: 2m 20s\tremaining: 11h 32m 59s\n",
      "336:\tlearn: 0.5094005\ttest: 0.9368567\tbest: 0.9368567 (336)\ttotal: 2m 20s\tremaining: 11h 32m 57s\n",
      "337:\tlearn: 0.5087713\ttest: 0.9369856\tbest: 0.9368567 (336)\ttotal: 2m 21s\tremaining: 11h 32m 57s\n",
      "338:\tlearn: 0.5081436\ttest: 0.9365330\tbest: 0.9365330 (338)\ttotal: 2m 21s\tremaining: 11h 32m 52s\n",
      "339:\tlearn: 0.5076084\ttest: 0.9366821\tbest: 0.9365330 (338)\ttotal: 2m 21s\tremaining: 11h 32m 51s\n",
      "340:\tlearn: 0.5070101\ttest: 0.9366735\tbest: 0.9365330 (338)\ttotal: 2m 22s\tremaining: 11h 32m 50s\n",
      "341:\tlearn: 0.5064923\ttest: 0.9366690\tbest: 0.9365330 (338)\ttotal: 2m 22s\tremaining: 11h 32m 47s\n",
      "342:\tlearn: 0.5058293\ttest: 0.9364262\tbest: 0.9364262 (342)\ttotal: 2m 23s\tremaining: 11h 32m 46s\n",
      "343:\tlearn: 0.5052602\ttest: 0.9364277\tbest: 0.9364262 (342)\ttotal: 2m 23s\tremaining: 11h 32m 41s\n",
      "344:\tlearn: 0.5046816\ttest: 0.9363685\tbest: 0.9363685 (344)\ttotal: 2m 23s\tremaining: 11h 32m 37s\n",
      "345:\tlearn: 0.5040113\ttest: 0.9361585\tbest: 0.9361585 (345)\ttotal: 2m 24s\tremaining: 11h 32m 34s\n",
      "346:\tlearn: 0.5034859\ttest: 0.9361328\tbest: 0.9361328 (346)\ttotal: 2m 24s\tremaining: 11h 32m 32s\n",
      "347:\tlearn: 0.5029626\ttest: 0.9361353\tbest: 0.9361328 (346)\ttotal: 2m 25s\tremaining: 11h 32m 31s\n",
      "348:\tlearn: 0.5024583\ttest: 0.9361538\tbest: 0.9361328 (346)\ttotal: 2m 25s\tremaining: 11h 32m 31s\n",
      "349:\tlearn: 0.5019728\ttest: 0.9360781\tbest: 0.9360781 (349)\ttotal: 2m 25s\tremaining: 11h 32m 33s\n",
      "350:\tlearn: 0.5013112\ttest: 0.9356833\tbest: 0.9356833 (350)\ttotal: 2m 26s\tremaining: 11h 32m 34s\n",
      "351:\tlearn: 0.5006825\ttest: 0.9354308\tbest: 0.9354308 (351)\ttotal: 2m 26s\tremaining: 11h 32m 32s\n",
      "352:\tlearn: 0.5000433\ttest: 0.9353404\tbest: 0.9353404 (352)\ttotal: 2m 27s\tremaining: 11h 32m 32s\n",
      "353:\tlearn: 0.4995899\ttest: 0.9352749\tbest: 0.9352749 (353)\ttotal: 2m 27s\tremaining: 11h 32m 33s\n",
      "354:\tlearn: 0.4989987\ttest: 0.9351082\tbest: 0.9351082 (354)\ttotal: 2m 28s\tremaining: 11h 32m 37s\n",
      "355:\tlearn: 0.4985251\ttest: 0.9345987\tbest: 0.9345987 (355)\ttotal: 2m 28s\tremaining: 11h 32m 36s\n",
      "356:\tlearn: 0.4981115\ttest: 0.9344430\tbest: 0.9344430 (356)\ttotal: 2m 28s\tremaining: 11h 32m 35s\n",
      "357:\tlearn: 0.4975921\ttest: 0.9344199\tbest: 0.9344199 (357)\ttotal: 2m 29s\tremaining: 11h 32m 31s\n",
      "358:\tlearn: 0.4969953\ttest: 0.9338897\tbest: 0.9338897 (358)\ttotal: 2m 29s\tremaining: 11h 32m 29s\n",
      "359:\tlearn: 0.4964129\ttest: 0.9336603\tbest: 0.9336603 (359)\ttotal: 2m 30s\tremaining: 11h 32m 28s\n",
      "360:\tlearn: 0.4958306\ttest: 0.9333757\tbest: 0.9333757 (360)\ttotal: 2m 30s\tremaining: 11h 32m 29s\n",
      "361:\tlearn: 0.4953187\ttest: 0.9333300\tbest: 0.9333300 (361)\ttotal: 2m 30s\tremaining: 11h 32m 27s\n",
      "362:\tlearn: 0.4948131\ttest: 0.9332728\tbest: 0.9332728 (362)\ttotal: 2m 31s\tremaining: 11h 32m 33s\n",
      "363:\tlearn: 0.4942765\ttest: 0.9333007\tbest: 0.9332728 (362)\ttotal: 2m 31s\tremaining: 11h 32m 32s\n",
      "364:\tlearn: 0.4937664\ttest: 0.9330360\tbest: 0.9330360 (364)\ttotal: 2m 32s\tremaining: 11h 32m 35s\n",
      "365:\tlearn: 0.4932414\ttest: 0.9330892\tbest: 0.9330360 (364)\ttotal: 2m 32s\tremaining: 11h 32m 36s\n",
      "366:\tlearn: 0.4927549\ttest: 0.9331526\tbest: 0.9330360 (364)\ttotal: 2m 33s\tremaining: 11h 32m 36s\n",
      "367:\tlearn: 0.4921537\ttest: 0.9331767\tbest: 0.9330360 (364)\ttotal: 2m 33s\tremaining: 11h 32m 35s\n",
      "368:\tlearn: 0.4916609\ttest: 0.9331676\tbest: 0.9330360 (364)\ttotal: 2m 33s\tremaining: 11h 32m 34s\n",
      "369:\tlearn: 0.4911228\ttest: 0.9331533\tbest: 0.9330360 (364)\ttotal: 2m 34s\tremaining: 11h 32m 32s\n",
      "370:\tlearn: 0.4905257\ttest: 0.9329798\tbest: 0.9329798 (370)\ttotal: 2m 34s\tremaining: 11h 32m 30s\n",
      "371:\tlearn: 0.4900864\ttest: 0.9329216\tbest: 0.9329216 (371)\ttotal: 2m 35s\tremaining: 11h 32m 30s\n",
      "372:\tlearn: 0.4896316\ttest: 0.9328528\tbest: 0.9328528 (372)\ttotal: 2m 35s\tremaining: 11h 32m 33s\n",
      "373:\tlearn: 0.4891081\ttest: 0.9329798\tbest: 0.9328528 (372)\ttotal: 2m 35s\tremaining: 11h 32m 29s\n",
      "374:\tlearn: 0.4885915\ttest: 0.9328665\tbest: 0.9328528 (372)\ttotal: 2m 36s\tremaining: 11h 32m 27s\n",
      "375:\tlearn: 0.4879756\ttest: 0.9325362\tbest: 0.9325362 (375)\ttotal: 2m 36s\tremaining: 11h 32m 25s\n",
      "376:\tlearn: 0.4874047\ttest: 0.9322862\tbest: 0.9322862 (376)\ttotal: 2m 37s\tremaining: 11h 32m 22s\n",
      "377:\tlearn: 0.4867305\ttest: 0.9319645\tbest: 0.9319645 (377)\ttotal: 2m 37s\tremaining: 11h 32m 19s\n",
      "378:\tlearn: 0.4863155\ttest: 0.9319022\tbest: 0.9319022 (378)\ttotal: 2m 38s\tremaining: 11h 32m 16s\n",
      "379:\tlearn: 0.4856993\ttest: 0.9316545\tbest: 0.9316545 (379)\ttotal: 2m 38s\tremaining: 11h 32m 15s\n",
      "380:\tlearn: 0.4853085\ttest: 0.9315339\tbest: 0.9315339 (380)\ttotal: 2m 38s\tremaining: 11h 32m 14s\n",
      "381:\tlearn: 0.4848263\ttest: 0.9315527\tbest: 0.9315339 (380)\ttotal: 2m 39s\tremaining: 11h 32m 12s\n",
      "382:\tlearn: 0.4843776\ttest: 0.9316754\tbest: 0.9315339 (380)\ttotal: 2m 39s\tremaining: 11h 32m 9s\n",
      "383:\tlearn: 0.4839032\ttest: 0.9317183\tbest: 0.9315339 (380)\ttotal: 2m 40s\tremaining: 11h 32m 11s\n",
      "384:\tlearn: 0.4832876\ttest: 0.9315955\tbest: 0.9315339 (380)\ttotal: 2m 40s\tremaining: 11h 32m 10s\n",
      "385:\tlearn: 0.4827982\ttest: 0.9314326\tbest: 0.9314326 (385)\ttotal: 2m 40s\tremaining: 11h 32m 9s\n",
      "386:\tlearn: 0.4823588\ttest: 0.9312957\tbest: 0.9312957 (386)\ttotal: 2m 41s\tremaining: 11h 32m 10s\n",
      "387:\tlearn: 0.4818364\ttest: 0.9316339\tbest: 0.9312957 (386)\ttotal: 2m 41s\tremaining: 11h 32m 12s\n",
      "388:\tlearn: 0.4814012\ttest: 0.9315892\tbest: 0.9312957 (386)\ttotal: 2m 42s\tremaining: 11h 32m 11s\n",
      "389:\tlearn: 0.4809130\ttest: 0.9314213\tbest: 0.9312957 (386)\ttotal: 2m 42s\tremaining: 11h 32m 11s\n",
      "390:\tlearn: 0.4804424\ttest: 0.9309949\tbest: 0.9309949 (390)\ttotal: 2m 43s\tremaining: 11h 32m 11s\n",
      "391:\tlearn: 0.4800029\ttest: 0.9308805\tbest: 0.9308805 (391)\ttotal: 2m 43s\tremaining: 11h 32m 13s\n",
      "392:\tlearn: 0.4794325\ttest: 0.9306193\tbest: 0.9306193 (392)\ttotal: 2m 43s\tremaining: 11h 32m 11s\n",
      "393:\tlearn: 0.4788831\ttest: 0.9304914\tbest: 0.9304914 (393)\ttotal: 2m 44s\tremaining: 11h 32m 10s\n",
      "394:\tlearn: 0.4784651\ttest: 0.9302959\tbest: 0.9302959 (394)\ttotal: 2m 44s\tremaining: 11h 32m 8s\n",
      "395:\tlearn: 0.4780480\ttest: 0.9302925\tbest: 0.9302925 (395)\ttotal: 2m 45s\tremaining: 11h 32m 9s\n",
      "396:\tlearn: 0.4776018\ttest: 0.9301261\tbest: 0.9301261 (396)\ttotal: 2m 45s\tremaining: 11h 32m 10s\n",
      "397:\tlearn: 0.4771070\ttest: 0.9297768\tbest: 0.9297768 (397)\ttotal: 2m 45s\tremaining: 11h 32m 10s\n",
      "398:\tlearn: 0.4766561\ttest: 0.9297872\tbest: 0.9297768 (397)\ttotal: 2m 46s\tremaining: 11h 32m 11s\n",
      "399:\tlearn: 0.4762402\ttest: 0.9296969\tbest: 0.9296969 (399)\ttotal: 2m 46s\tremaining: 11h 32m 15s\n",
      "400:\tlearn: 0.4757273\ttest: 0.9296543\tbest: 0.9296543 (400)\ttotal: 2m 47s\tremaining: 11h 32m 13s\n",
      "401:\tlearn: 0.4752429\ttest: 0.9296580\tbest: 0.9296543 (400)\ttotal: 2m 47s\tremaining: 11h 32m 10s\n",
      "402:\tlearn: 0.4747739\ttest: 0.9300076\tbest: 0.9296543 (400)\ttotal: 2m 48s\tremaining: 11h 32m 9s\n",
      "403:\tlearn: 0.4743421\ttest: 0.9299139\tbest: 0.9296543 (400)\ttotal: 2m 48s\tremaining: 11h 32m 11s\n",
      "404:\tlearn: 0.4739018\ttest: 0.9298380\tbest: 0.9296543 (400)\ttotal: 2m 48s\tremaining: 11h 32m 9s\n",
      "405:\tlearn: 0.4735000\ttest: 0.9297990\tbest: 0.9296543 (400)\ttotal: 2m 49s\tremaining: 11h 32m 8s\n",
      "406:\tlearn: 0.4730712\ttest: 0.9295201\tbest: 0.9295201 (406)\ttotal: 2m 49s\tremaining: 11h 32m 5s\n",
      "407:\tlearn: 0.4725039\ttest: 0.9294859\tbest: 0.9294859 (407)\ttotal: 2m 50s\tremaining: 11h 32m 5s\n",
      "408:\tlearn: 0.4719862\ttest: 0.9294537\tbest: 0.9294537 (408)\ttotal: 2m 50s\tremaining: 11h 32m 4s\n",
      "409:\tlearn: 0.4715552\ttest: 0.9293933\tbest: 0.9293933 (409)\ttotal: 2m 50s\tremaining: 11h 32m 6s\n",
      "410:\tlearn: 0.4710288\ttest: 0.9292344\tbest: 0.9292344 (410)\ttotal: 2m 51s\tremaining: 11h 32m 8s\n",
      "411:\tlearn: 0.4705591\ttest: 0.9286734\tbest: 0.9286734 (411)\ttotal: 2m 51s\tremaining: 11h 32m 18s\n",
      "412:\tlearn: 0.4701606\ttest: 0.9284721\tbest: 0.9284721 (412)\ttotal: 2m 52s\tremaining: 11h 32m 16s\n",
      "413:\tlearn: 0.4697381\ttest: 0.9284935\tbest: 0.9284721 (412)\ttotal: 2m 52s\tremaining: 11h 32m 15s\n",
      "414:\tlearn: 0.4692808\ttest: 0.9285614\tbest: 0.9284721 (412)\ttotal: 2m 53s\tremaining: 11h 32m 20s\n",
      "415:\tlearn: 0.4688184\ttest: 0.9285391\tbest: 0.9284721 (412)\ttotal: 2m 53s\tremaining: 11h 32m 20s\n",
      "416:\tlearn: 0.4682779\ttest: 0.9284735\tbest: 0.9284721 (412)\ttotal: 2m 53s\tremaining: 11h 32m 18s\n",
      "417:\tlearn: 0.4678698\ttest: 0.9284471\tbest: 0.9284471 (417)\ttotal: 2m 54s\tremaining: 11h 32m 16s\n",
      "418:\tlearn: 0.4673925\ttest: 0.9281864\tbest: 0.9281864 (418)\ttotal: 2m 54s\tremaining: 11h 32m 14s\n",
      "419:\tlearn: 0.4667934\ttest: 0.9278862\tbest: 0.9278862 (419)\ttotal: 2m 55s\tremaining: 11h 32m 14s\n",
      "420:\tlearn: 0.4663812\ttest: 0.9277749\tbest: 0.9277749 (420)\ttotal: 2m 55s\tremaining: 11h 32m 14s\n",
      "421:\tlearn: 0.4659095\ttest: 0.9275332\tbest: 0.9275332 (421)\ttotal: 2m 56s\tremaining: 11h 32m 15s\n",
      "422:\tlearn: 0.4655131\ttest: 0.9273875\tbest: 0.9273875 (422)\ttotal: 2m 56s\tremaining: 11h 32m 16s\n",
      "423:\tlearn: 0.4651136\ttest: 0.9273154\tbest: 0.9273154 (423)\ttotal: 2m 56s\tremaining: 11h 32m 15s\n",
      "424:\tlearn: 0.4647097\ttest: 0.9272664\tbest: 0.9272664 (424)\ttotal: 2m 57s\tremaining: 11h 32m 13s\n",
      "425:\tlearn: 0.4642413\ttest: 0.9269289\tbest: 0.9269289 (425)\ttotal: 2m 57s\tremaining: 11h 32m 12s\n",
      "426:\tlearn: 0.4638568\ttest: 0.9269132\tbest: 0.9269132 (426)\ttotal: 2m 58s\tremaining: 11h 32m 19s\n",
      "427:\tlearn: 0.4634539\ttest: 0.9269061\tbest: 0.9269061 (427)\ttotal: 2m 58s\tremaining: 11h 32m 21s\n",
      "428:\tlearn: 0.4630081\ttest: 0.9268935\tbest: 0.9268935 (428)\ttotal: 2m 58s\tremaining: 11h 32m 19s\n",
      "429:\tlearn: 0.4625827\ttest: 0.9269949\tbest: 0.9268935 (428)\ttotal: 2m 59s\tremaining: 11h 32m 18s\n",
      "430:\tlearn: 0.4620766\ttest: 0.9266731\tbest: 0.9266731 (430)\ttotal: 2m 59s\tremaining: 11h 32m 14s\n",
      "431:\tlearn: 0.4616518\ttest: 0.9267075\tbest: 0.9266731 (430)\ttotal: 3m\tremaining: 11h 32m 13s\n",
      "432:\tlearn: 0.4611770\ttest: 0.9270420\tbest: 0.9266731 (430)\ttotal: 3m\tremaining: 11h 32m 16s\n",
      "433:\tlearn: 0.4608156\ttest: 0.9272541\tbest: 0.9266731 (430)\ttotal: 3m 1s\tremaining: 11h 32m 17s\n",
      "434:\tlearn: 0.4604321\ttest: 0.9274603\tbest: 0.9266731 (430)\ttotal: 3m 1s\tremaining: 11h 32m 22s\n",
      "435:\tlearn: 0.4600296\ttest: 0.9272983\tbest: 0.9266731 (430)\ttotal: 3m 1s\tremaining: 11h 32m 24s\n",
      "436:\tlearn: 0.4595592\ttest: 0.9271315\tbest: 0.9266731 (430)\ttotal: 3m 2s\tremaining: 11h 32m 23s\n",
      "437:\tlearn: 0.4591801\ttest: 0.9270985\tbest: 0.9266731 (430)\ttotal: 3m 2s\tremaining: 11h 32m 23s\n",
      "438:\tlearn: 0.4588244\ttest: 0.9271362\tbest: 0.9266731 (430)\ttotal: 3m 3s\tremaining: 11h 32m 24s\n",
      "439:\tlearn: 0.4583081\ttest: 0.9270963\tbest: 0.9266731 (430)\ttotal: 3m 3s\tremaining: 11h 32m 23s\n",
      "440:\tlearn: 0.4578403\ttest: 0.9270845\tbest: 0.9266731 (430)\ttotal: 3m 4s\tremaining: 11h 32m 26s\n",
      "441:\tlearn: 0.4574221\ttest: 0.9269964\tbest: 0.9266731 (430)\ttotal: 3m 4s\tremaining: 11h 32m 23s\n",
      "442:\tlearn: 0.4570094\ttest: 0.9269128\tbest: 0.9266731 (430)\ttotal: 3m 4s\tremaining: 11h 32m 22s\n",
      "443:\tlearn: 0.4566649\ttest: 0.9269543\tbest: 0.9266731 (430)\ttotal: 3m 5s\tremaining: 11h 32m 23s\n",
      "444:\tlearn: 0.4561719\ttest: 0.9269499\tbest: 0.9266731 (430)\ttotal: 3m 5s\tremaining: 11h 32m 20s\n",
      "445:\tlearn: 0.4557382\ttest: 0.9267879\tbest: 0.9266731 (430)\ttotal: 3m 6s\tremaining: 11h 32m 20s\n",
      "446:\tlearn: 0.4553758\ttest: 0.9267599\tbest: 0.9266731 (430)\ttotal: 3m 6s\tremaining: 11h 32m 22s\n",
      "447:\tlearn: 0.4549073\ttest: 0.9263200\tbest: 0.9263200 (447)\ttotal: 3m 6s\tremaining: 11h 32m 21s\n",
      "448:\tlearn: 0.4543920\ttest: 0.9260181\tbest: 0.9260181 (448)\ttotal: 3m 7s\tremaining: 11h 32m 21s\n",
      "449:\tlearn: 0.4539960\ttest: 0.9257986\tbest: 0.9257986 (449)\ttotal: 3m 7s\tremaining: 11h 32m 21s\n",
      "450:\tlearn: 0.4534792\ttest: 0.9257036\tbest: 0.9257036 (450)\ttotal: 3m 8s\tremaining: 11h 32m 21s\n",
      "451:\tlearn: 0.4530860\ttest: 0.9256700\tbest: 0.9256700 (451)\ttotal: 3m 8s\tremaining: 11h 32m 21s\n",
      "452:\tlearn: 0.4527102\ttest: 0.9256882\tbest: 0.9256700 (451)\ttotal: 3m 9s\tremaining: 11h 32m 23s\n",
      "453:\tlearn: 0.4522244\ttest: 0.9252827\tbest: 0.9252827 (453)\ttotal: 3m 9s\tremaining: 11h 32m 22s\n",
      "454:\tlearn: 0.4518173\ttest: 0.9251203\tbest: 0.9251203 (454)\ttotal: 3m 9s\tremaining: 11h 32m 19s\n",
      "455:\tlearn: 0.4514524\ttest: 0.9250784\tbest: 0.9250784 (455)\ttotal: 3m 10s\tremaining: 11h 32m 17s\n",
      "456:\tlearn: 0.4509991\ttest: 0.9251224\tbest: 0.9250784 (455)\ttotal: 3m 10s\tremaining: 11h 32m 15s\n",
      "457:\tlearn: 0.4506213\ttest: 0.9249623\tbest: 0.9249623 (457)\ttotal: 3m 11s\tremaining: 11h 32m 19s\n",
      "458:\tlearn: 0.4502428\ttest: 0.9248504\tbest: 0.9248504 (458)\ttotal: 3m 11s\tremaining: 11h 32m 23s\n",
      "459:\tlearn: 0.4497697\ttest: 0.9248754\tbest: 0.9248504 (458)\ttotal: 3m 11s\tremaining: 11h 32m 22s\n",
      "460:\tlearn: 0.4494417\ttest: 0.9249004\tbest: 0.9248504 (458)\ttotal: 3m 12s\tremaining: 11h 32m 22s\n",
      "461:\tlearn: 0.4490628\ttest: 0.9247888\tbest: 0.9247888 (461)\ttotal: 3m 12s\tremaining: 11h 32m 23s\n",
      "462:\tlearn: 0.4486549\ttest: 0.9246019\tbest: 0.9246019 (462)\ttotal: 3m 13s\tremaining: 11h 32m 21s\n",
      "463:\tlearn: 0.4482791\ttest: 0.9246517\tbest: 0.9246019 (462)\ttotal: 3m 13s\tremaining: 11h 32m 21s\n",
      "464:\tlearn: 0.4478377\ttest: 0.9246197\tbest: 0.9246019 (462)\ttotal: 3m 14s\tremaining: 11h 32m 23s\n",
      "465:\tlearn: 0.4474077\ttest: 0.9244709\tbest: 0.9244709 (465)\ttotal: 3m 14s\tremaining: 11h 32m 21s\n",
      "466:\tlearn: 0.4469993\ttest: 0.9244136\tbest: 0.9244136 (466)\ttotal: 3m 14s\tremaining: 11h 32m 18s\n",
      "467:\tlearn: 0.4466295\ttest: 0.9245602\tbest: 0.9244136 (466)\ttotal: 3m 15s\tremaining: 11h 32m 18s\n",
      "468:\tlearn: 0.4461886\ttest: 0.9243445\tbest: 0.9243445 (468)\ttotal: 3m 15s\tremaining: 11h 32m 18s\n",
      "469:\tlearn: 0.4458508\ttest: 0.9241704\tbest: 0.9241704 (469)\ttotal: 3m 16s\tremaining: 11h 32m 19s\n",
      "470:\tlearn: 0.4455172\ttest: 0.9240195\tbest: 0.9240195 (470)\ttotal: 3m 16s\tremaining: 11h 32m 20s\n",
      "471:\tlearn: 0.4450377\ttest: 0.9239909\tbest: 0.9239909 (471)\ttotal: 3m 17s\tremaining: 11h 32m 20s\n",
      "472:\tlearn: 0.4445691\ttest: 0.9236211\tbest: 0.9236211 (472)\ttotal: 3m 17s\tremaining: 11h 32m 22s\n",
      "473:\tlearn: 0.4442114\ttest: 0.9235178\tbest: 0.9235178 (473)\ttotal: 3m 17s\tremaining: 11h 32m 25s\n",
      "474:\tlearn: 0.4438204\ttest: 0.9235474\tbest: 0.9235178 (473)\ttotal: 3m 18s\tremaining: 11h 32m 25s\n",
      "475:\tlearn: 0.4434267\ttest: 0.9234878\tbest: 0.9234878 (475)\ttotal: 3m 18s\tremaining: 11h 32m 25s\n",
      "476:\tlearn: 0.4429926\ttest: 0.9235347\tbest: 0.9234878 (475)\ttotal: 3m 19s\tremaining: 11h 32m 26s\n",
      "477:\tlearn: 0.4426422\ttest: 0.9236051\tbest: 0.9234878 (475)\ttotal: 3m 19s\tremaining: 11h 32m 25s\n",
      "478:\tlearn: 0.4423033\ttest: 0.9237524\tbest: 0.9234878 (475)\ttotal: 3m 19s\tremaining: 11h 32m 23s\n",
      "479:\tlearn: 0.4419345\ttest: 0.9234476\tbest: 0.9234476 (479)\ttotal: 3m 20s\tremaining: 11h 32m 21s\n",
      "480:\tlearn: 0.4415306\ttest: 0.9234165\tbest: 0.9234165 (480)\ttotal: 3m 20s\tremaining: 11h 32m 20s\n",
      "481:\tlearn: 0.4411073\ttest: 0.9232850\tbest: 0.9232850 (481)\ttotal: 3m 21s\tremaining: 11h 32m 26s\n",
      "482:\tlearn: 0.4407871\ttest: 0.9233075\tbest: 0.9232850 (481)\ttotal: 3m 21s\tremaining: 11h 32m 25s\n",
      "483:\tlearn: 0.4403702\ttest: 0.9232851\tbest: 0.9232850 (481)\ttotal: 3m 22s\tremaining: 11h 32m 24s\n",
      "484:\tlearn: 0.4400334\ttest: 0.9232318\tbest: 0.9232318 (484)\ttotal: 3m 22s\tremaining: 11h 32m 24s\n",
      "485:\tlearn: 0.4397347\ttest: 0.9230991\tbest: 0.9230991 (485)\ttotal: 3m 22s\tremaining: 11h 32m 24s\n",
      "486:\tlearn: 0.4393179\ttest: 0.9227708\tbest: 0.9227708 (486)\ttotal: 3m 23s\tremaining: 11h 32m 23s\n",
      "487:\tlearn: 0.4389573\ttest: 0.9227892\tbest: 0.9227708 (486)\ttotal: 3m 23s\tremaining: 11h 32m 24s\n",
      "488:\tlearn: 0.4386049\ttest: 0.9226853\tbest: 0.9226853 (488)\ttotal: 3m 24s\tremaining: 11h 32m 26s\n",
      "489:\tlearn: 0.4382195\ttest: 0.9228110\tbest: 0.9226853 (488)\ttotal: 3m 24s\tremaining: 11h 32m 24s\n",
      "490:\tlearn: 0.4378596\ttest: 0.9226702\tbest: 0.9226702 (490)\ttotal: 3m 24s\tremaining: 11h 32m 22s\n",
      "491:\tlearn: 0.4375081\ttest: 0.9226557\tbest: 0.9226557 (491)\ttotal: 3m 25s\tremaining: 11h 32m 19s\n",
      "492:\tlearn: 0.4371672\ttest: 0.9225846\tbest: 0.9225846 (492)\ttotal: 3m 25s\tremaining: 11h 32m 19s\n",
      "493:\tlearn: 0.4367986\ttest: 0.9225885\tbest: 0.9225846 (492)\ttotal: 3m 26s\tremaining: 11h 32m 21s\n",
      "494:\tlearn: 0.4363485\ttest: 0.9223585\tbest: 0.9223585 (494)\ttotal: 3m 26s\tremaining: 11h 32m 20s\n",
      "495:\tlearn: 0.4360101\ttest: 0.9224496\tbest: 0.9223585 (494)\ttotal: 3m 27s\tremaining: 11h 32m 19s\n",
      "496:\tlearn: 0.4355317\ttest: 0.9224029\tbest: 0.9223585 (494)\ttotal: 3m 27s\tremaining: 11h 32m 20s\n",
      "497:\tlearn: 0.4352188\ttest: 0.9223638\tbest: 0.9223585 (494)\ttotal: 3m 27s\tremaining: 11h 32m 19s\n",
      "498:\tlearn: 0.4348262\ttest: 0.9223958\tbest: 0.9223585 (494)\ttotal: 3m 28s\tremaining: 11h 32m 21s\n",
      "499:\tlearn: 0.4344749\ttest: 0.9221716\tbest: 0.9221716 (499)\ttotal: 3m 28s\tremaining: 11h 32m 21s\n",
      "500:\tlearn: 0.4341045\ttest: 0.9220135\tbest: 0.9220135 (500)\ttotal: 3m 29s\tremaining: 11h 32m 25s\n",
      "501:\tlearn: 0.4337203\ttest: 0.9221227\tbest: 0.9220135 (500)\ttotal: 3m 29s\tremaining: 11h 32m 26s\n",
      "502:\tlearn: 0.4333523\ttest: 0.9216277\tbest: 0.9216277 (502)\ttotal: 3m 30s\tremaining: 11h 32m 24s\n",
      "503:\tlearn: 0.4330255\ttest: 0.9215501\tbest: 0.9215501 (503)\ttotal: 3m 30s\tremaining: 11h 32m 23s\n",
      "504:\tlearn: 0.4326894\ttest: 0.9214083\tbest: 0.9214083 (504)\ttotal: 3m 30s\tremaining: 11h 32m 25s\n",
      "505:\tlearn: 0.4323653\ttest: 0.9215611\tbest: 0.9214083 (504)\ttotal: 3m 31s\tremaining: 11h 32m 24s\n",
      "506:\tlearn: 0.4320728\ttest: 0.9212372\tbest: 0.9212372 (506)\ttotal: 3m 31s\tremaining: 11h 32m 24s\n",
      "507:\tlearn: 0.4317539\ttest: 0.9210338\tbest: 0.9210338 (507)\ttotal: 3m 32s\tremaining: 11h 32m 25s\n",
      "508:\tlearn: 0.4314895\ttest: 0.9209695\tbest: 0.9209695 (508)\ttotal: 3m 32s\tremaining: 11h 32m 23s\n",
      "509:\tlearn: 0.4312061\ttest: 0.9208116\tbest: 0.9208116 (509)\ttotal: 3m 32s\tremaining: 11h 32m 22s\n",
      "510:\tlearn: 0.4307466\ttest: 0.9207142\tbest: 0.9207142 (510)\ttotal: 3m 33s\tremaining: 11h 32m 20s\n",
      "511:\tlearn: 0.4303460\ttest: 0.9204933\tbest: 0.9204933 (511)\ttotal: 3m 33s\tremaining: 11h 32m 18s\n",
      "512:\tlearn: 0.4300172\ttest: 0.9203686\tbest: 0.9203686 (512)\ttotal: 3m 34s\tremaining: 11h 32m 15s\n",
      "513:\tlearn: 0.4296629\ttest: 0.9202723\tbest: 0.9202723 (513)\ttotal: 3m 34s\tremaining: 11h 32m 12s\n",
      "514:\tlearn: 0.4293682\ttest: 0.9203532\tbest: 0.9202723 (513)\ttotal: 3m 34s\tremaining: 11h 32m 11s\n",
      "515:\tlearn: 0.4290144\ttest: 0.9204436\tbest: 0.9202723 (513)\ttotal: 3m 35s\tremaining: 11h 32m 11s\n",
      "516:\tlearn: 0.4287010\ttest: 0.9206708\tbest: 0.9202723 (513)\ttotal: 3m 35s\tremaining: 11h 32m 10s\n",
      "517:\tlearn: 0.4283092\ttest: 0.9206223\tbest: 0.9202723 (513)\ttotal: 3m 36s\tremaining: 11h 32m 7s\n",
      "518:\tlearn: 0.4279569\ttest: 0.9202632\tbest: 0.9202632 (518)\ttotal: 3m 36s\tremaining: 11h 32m 5s\n",
      "519:\tlearn: 0.4276694\ttest: 0.9205155\tbest: 0.9202632 (518)\ttotal: 3m 37s\tremaining: 11h 32m 5s\n",
      "520:\tlearn: 0.4273197\ttest: 0.9203530\tbest: 0.9202632 (518)\ttotal: 3m 37s\tremaining: 11h 32m\n",
      "521:\tlearn: 0.4269595\ttest: 0.9202878\tbest: 0.9202632 (518)\ttotal: 3m 37s\tremaining: 11h 31m 58s\n",
      "522:\tlearn: 0.4266386\ttest: 0.9201889\tbest: 0.9201889 (522)\ttotal: 3m 38s\tremaining: 11h 31m 54s\n",
      "523:\tlearn: 0.4262799\ttest: 0.9201027\tbest: 0.9201027 (523)\ttotal: 3m 38s\tremaining: 11h 31m 52s\n",
      "524:\tlearn: 0.4259741\ttest: 0.9201195\tbest: 0.9201027 (523)\ttotal: 3m 39s\tremaining: 11h 31m 48s\n",
      "525:\tlearn: 0.4256519\ttest: 0.9201038\tbest: 0.9201027 (523)\ttotal: 3m 39s\tremaining: 11h 31m 47s\n",
      "526:\tlearn: 0.4253244\ttest: 0.9198389\tbest: 0.9198389 (526)\ttotal: 3m 39s\tremaining: 11h 31m 44s\n",
      "527:\tlearn: 0.4249682\ttest: 0.9197709\tbest: 0.9197709 (527)\ttotal: 3m 40s\tremaining: 11h 31m 44s\n",
      "528:\tlearn: 0.4246057\ttest: 0.9195785\tbest: 0.9195785 (528)\ttotal: 3m 40s\tremaining: 11h 31m 43s\n",
      "529:\tlearn: 0.4243580\ttest: 0.9194711\tbest: 0.9194711 (529)\ttotal: 3m 41s\tremaining: 11h 31m 43s\n",
      "530:\tlearn: 0.4239911\ttest: 0.9190823\tbest: 0.9190823 (530)\ttotal: 3m 41s\tremaining: 11h 31m 43s\n",
      "531:\tlearn: 0.4236349\ttest: 0.9191014\tbest: 0.9190823 (530)\ttotal: 3m 41s\tremaining: 11h 31m 44s\n",
      "532:\tlearn: 0.4232655\ttest: 0.9191055\tbest: 0.9190823 (530)\ttotal: 3m 42s\tremaining: 11h 31m 45s\n",
      "533:\tlearn: 0.4229228\ttest: 0.9189451\tbest: 0.9189451 (533)\ttotal: 3m 42s\tremaining: 11h 31m 45s\n",
      "534:\tlearn: 0.4225406\ttest: 0.9188880\tbest: 0.9188880 (534)\ttotal: 3m 43s\tremaining: 11h 31m 46s\n",
      "535:\tlearn: 0.4223114\ttest: 0.9189921\tbest: 0.9188880 (534)\ttotal: 3m 43s\tremaining: 11h 31m 46s\n",
      "536:\tlearn: 0.4220295\ttest: 0.9190517\tbest: 0.9188880 (534)\ttotal: 3m 44s\tremaining: 11h 31m 46s\n",
      "537:\tlearn: 0.4216475\ttest: 0.9188653\tbest: 0.9188653 (537)\ttotal: 3m 44s\tremaining: 11h 31m 47s\n",
      "538:\tlearn: 0.4212764\ttest: 0.9188900\tbest: 0.9188653 (537)\ttotal: 3m 44s\tremaining: 11h 31m 44s\n",
      "539:\tlearn: 0.4209457\ttest: 0.9186277\tbest: 0.9186277 (539)\ttotal: 3m 45s\tremaining: 11h 31m 43s\n",
      "540:\tlearn: 0.4205487\ttest: 0.9186925\tbest: 0.9186277 (539)\ttotal: 3m 45s\tremaining: 11h 31m 43s\n",
      "541:\tlearn: 0.4202457\ttest: 0.9188385\tbest: 0.9186277 (539)\ttotal: 3m 46s\tremaining: 11h 31m 42s\n",
      "542:\tlearn: 0.4199193\ttest: 0.9187520\tbest: 0.9186277 (539)\ttotal: 3m 46s\tremaining: 11h 31m 41s\n",
      "543:\tlearn: 0.4195578\ttest: 0.9185932\tbest: 0.9185932 (543)\ttotal: 3m 47s\tremaining: 11h 31m 43s\n",
      "544:\tlearn: 0.4192604\ttest: 0.9187023\tbest: 0.9185932 (543)\ttotal: 3m 47s\tremaining: 11h 31m 44s\n",
      "545:\tlearn: 0.4189313\ttest: 0.9187279\tbest: 0.9185932 (543)\ttotal: 3m 47s\tremaining: 11h 31m 44s\n",
      "546:\tlearn: 0.4185726\ttest: 0.9187087\tbest: 0.9185932 (543)\ttotal: 3m 48s\tremaining: 11h 31m 45s\n",
      "547:\tlearn: 0.4182403\ttest: 0.9186096\tbest: 0.9185932 (543)\ttotal: 3m 48s\tremaining: 11h 31m 51s\n",
      "548:\tlearn: 0.4179098\ttest: 0.9186356\tbest: 0.9185932 (543)\ttotal: 3m 49s\tremaining: 11h 31m 50s\n",
      "549:\tlearn: 0.4175685\ttest: 0.9183423\tbest: 0.9183423 (549)\ttotal: 3m 49s\tremaining: 11h 31m 49s\n",
      "550:\tlearn: 0.4172167\ttest: 0.9180798\tbest: 0.9180798 (550)\ttotal: 3m 49s\tremaining: 11h 31m 49s\n",
      "551:\tlearn: 0.4168932\ttest: 0.9179798\tbest: 0.9179798 (551)\ttotal: 3m 50s\tremaining: 11h 31m 48s\n",
      "552:\tlearn: 0.4166542\ttest: 0.9178892\tbest: 0.9178892 (552)\ttotal: 3m 50s\tremaining: 11h 31m 48s\n",
      "553:\tlearn: 0.4163746\ttest: 0.9177722\tbest: 0.9177722 (553)\ttotal: 3m 51s\tremaining: 11h 31m 48s\n",
      "554:\tlearn: 0.4160548\ttest: 0.9175330\tbest: 0.9175330 (554)\ttotal: 3m 51s\tremaining: 11h 31m 47s\n",
      "555:\tlearn: 0.4157271\ttest: 0.9175219\tbest: 0.9175219 (555)\ttotal: 3m 52s\tremaining: 11h 31m 46s\n",
      "556:\tlearn: 0.4154647\ttest: 0.9175709\tbest: 0.9175219 (555)\ttotal: 3m 52s\tremaining: 11h 31m 43s\n",
      "557:\tlearn: 0.4151661\ttest: 0.9173231\tbest: 0.9173231 (557)\ttotal: 3m 52s\tremaining: 11h 31m 41s\n",
      "558:\tlearn: 0.4148673\ttest: 0.9173022\tbest: 0.9173022 (558)\ttotal: 3m 53s\tremaining: 11h 31m 39s\n",
      "559:\tlearn: 0.4146440\ttest: 0.9172939\tbest: 0.9172939 (559)\ttotal: 3m 53s\tremaining: 11h 31m 35s\n",
      "560:\tlearn: 0.4143632\ttest: 0.9172635\tbest: 0.9172635 (560)\ttotal: 3m 54s\tremaining: 11h 31m 33s\n",
      "561:\tlearn: 0.4140684\ttest: 0.9172934\tbest: 0.9172635 (560)\ttotal: 3m 54s\tremaining: 11h 31m 31s\n",
      "562:\tlearn: 0.4136794\ttest: 0.9170218\tbest: 0.9170218 (562)\ttotal: 3m 54s\tremaining: 11h 31m 29s\n",
      "563:\tlearn: 0.4133758\ttest: 0.9167315\tbest: 0.9167315 (563)\ttotal: 3m 55s\tremaining: 11h 31m 29s\n",
      "564:\tlearn: 0.4130586\ttest: 0.9168528\tbest: 0.9167315 (563)\ttotal: 3m 55s\tremaining: 11h 31m 27s\n",
      "565:\tlearn: 0.4127957\ttest: 0.9167537\tbest: 0.9167315 (563)\ttotal: 3m 56s\tremaining: 11h 31m 28s\n",
      "566:\tlearn: 0.4124928\ttest: 0.9167040\tbest: 0.9167040 (566)\ttotal: 3m 56s\tremaining: 11h 31m 28s\n",
      "567:\tlearn: 0.4122121\ttest: 0.9167599\tbest: 0.9167040 (566)\ttotal: 3m 57s\tremaining: 11h 31m 29s\n",
      "568:\tlearn: 0.4119144\ttest: 0.9167797\tbest: 0.9167040 (566)\ttotal: 3m 57s\tremaining: 11h 31m 27s\n",
      "569:\tlearn: 0.4116341\ttest: 0.9166251\tbest: 0.9166251 (569)\ttotal: 3m 57s\tremaining: 11h 31m 26s\n",
      "570:\tlearn: 0.4112820\ttest: 0.9165323\tbest: 0.9165323 (570)\ttotal: 3m 58s\tremaining: 11h 31m 27s\n",
      "571:\tlearn: 0.4110267\ttest: 0.9166235\tbest: 0.9165323 (570)\ttotal: 3m 58s\tremaining: 11h 31m 27s\n",
      "572:\tlearn: 0.4107401\ttest: 0.9164798\tbest: 0.9164798 (572)\ttotal: 3m 59s\tremaining: 11h 31m 27s\n",
      "573:\tlearn: 0.4104029\ttest: 0.9159488\tbest: 0.9159488 (573)\ttotal: 3m 59s\tremaining: 11h 31m 29s\n",
      "574:\tlearn: 0.4101451\ttest: 0.9157117\tbest: 0.9157117 (574)\ttotal: 3m 59s\tremaining: 11h 31m 30s\n",
      "575:\tlearn: 0.4098622\ttest: 0.9157001\tbest: 0.9157001 (575)\ttotal: 4m\tremaining: 11h 31m 28s\n",
      "576:\tlearn: 0.4095257\ttest: 0.9157020\tbest: 0.9157001 (575)\ttotal: 4m\tremaining: 11h 31m 24s\n",
      "577:\tlearn: 0.4093020\ttest: 0.9156347\tbest: 0.9156347 (577)\ttotal: 4m 1s\tremaining: 11h 31m 23s\n",
      "578:\tlearn: 0.4090482\ttest: 0.9156744\tbest: 0.9156347 (577)\ttotal: 4m 1s\tremaining: 11h 31m 22s\n",
      "579:\tlearn: 0.4087710\ttest: 0.9154571\tbest: 0.9154571 (579)\ttotal: 4m 2s\tremaining: 11h 31m 22s\n",
      "580:\tlearn: 0.4084318\ttest: 0.9154989\tbest: 0.9154571 (579)\ttotal: 4m 2s\tremaining: 11h 31m 21s\n",
      "581:\tlearn: 0.4081482\ttest: 0.9153491\tbest: 0.9153491 (581)\ttotal: 4m 2s\tremaining: 11h 31m 20s\n",
      "582:\tlearn: 0.4078581\ttest: 0.9152990\tbest: 0.9152990 (582)\ttotal: 4m 3s\tremaining: 11h 31m 22s\n",
      "583:\tlearn: 0.4075719\ttest: 0.9150705\tbest: 0.9150705 (583)\ttotal: 4m 3s\tremaining: 11h 31m 22s\n",
      "584:\tlearn: 0.4073142\ttest: 0.9149812\tbest: 0.9149812 (584)\ttotal: 4m 4s\tremaining: 11h 31m 22s\n",
      "585:\tlearn: 0.4070516\ttest: 0.9148529\tbest: 0.9148529 (585)\ttotal: 4m 4s\tremaining: 11h 31m 22s\n",
      "586:\tlearn: 0.4068571\ttest: 0.9149029\tbest: 0.9148529 (585)\ttotal: 4m 4s\tremaining: 11h 31m 22s\n",
      "587:\tlearn: 0.4065407\ttest: 0.9146697\tbest: 0.9146697 (587)\ttotal: 4m 5s\tremaining: 11h 31m 20s\n",
      "588:\tlearn: 0.4062726\ttest: 0.9145577\tbest: 0.9145577 (588)\ttotal: 4m 5s\tremaining: 11h 31m 18s\n",
      "589:\tlearn: 0.4059867\ttest: 0.9145191\tbest: 0.9145191 (589)\ttotal: 4m 6s\tremaining: 11h 31m 19s\n",
      "590:\tlearn: 0.4056633\ttest: 0.9145834\tbest: 0.9145191 (589)\ttotal: 4m 6s\tremaining: 11h 31m 19s\n",
      "591:\tlearn: 0.4054032\ttest: 0.9145607\tbest: 0.9145191 (589)\ttotal: 4m 7s\tremaining: 11h 31m 19s\n",
      "592:\tlearn: 0.4051379\ttest: 0.9145222\tbest: 0.9145191 (589)\ttotal: 4m 7s\tremaining: 11h 31m 19s\n",
      "593:\tlearn: 0.4048514\ttest: 0.9145105\tbest: 0.9145105 (593)\ttotal: 4m 7s\tremaining: 11h 31m 22s\n",
      "594:\tlearn: 0.4045790\ttest: 0.9144922\tbest: 0.9144922 (594)\ttotal: 4m 8s\tremaining: 11h 31m 23s\n",
      "595:\tlearn: 0.4043377\ttest: 0.9145062\tbest: 0.9144922 (594)\ttotal: 4m 8s\tremaining: 11h 31m 22s\n",
      "596:\tlearn: 0.4040576\ttest: 0.9145054\tbest: 0.9144922 (594)\ttotal: 4m 9s\tremaining: 11h 31m 22s\n",
      "597:\tlearn: 0.4038027\ttest: 0.9144872\tbest: 0.9144872 (597)\ttotal: 4m 9s\tremaining: 11h 31m 23s\n",
      "598:\tlearn: 0.4034988\ttest: 0.9143091\tbest: 0.9143091 (598)\ttotal: 4m 9s\tremaining: 11h 31m 24s\n",
      "599:\tlearn: 0.4032155\ttest: 0.9141114\tbest: 0.9141114 (599)\ttotal: 4m 10s\tremaining: 11h 31m 23s\n",
      "600:\tlearn: 0.4029517\ttest: 0.9140762\tbest: 0.9140762 (600)\ttotal: 4m 10s\tremaining: 11h 31m 21s\n",
      "601:\tlearn: 0.4026626\ttest: 0.9139866\tbest: 0.9139866 (601)\ttotal: 4m 11s\tremaining: 11h 31m 21s\n",
      "602:\tlearn: 0.4023925\ttest: 0.9138500\tbest: 0.9138500 (602)\ttotal: 4m 11s\tremaining: 11h 31m 21s\n",
      "603:\tlearn: 0.4020935\ttest: 0.9138032\tbest: 0.9138032 (603)\ttotal: 4m 12s\tremaining: 11h 31m 22s\n",
      "604:\tlearn: 0.4018672\ttest: 0.9137390\tbest: 0.9137390 (604)\ttotal: 4m 12s\tremaining: 11h 31m 22s\n",
      "605:\tlearn: 0.4016086\ttest: 0.9135281\tbest: 0.9135281 (605)\ttotal: 4m 12s\tremaining: 11h 31m 23s\n",
      "606:\tlearn: 0.4013468\ttest: 0.9135907\tbest: 0.9135281 (605)\ttotal: 4m 13s\tremaining: 11h 31m 25s\n",
      "607:\tlearn: 0.4011170\ttest: 0.9136977\tbest: 0.9135281 (605)\ttotal: 4m 13s\tremaining: 11h 31m 24s\n",
      "608:\tlearn: 0.4008925\ttest: 0.9135720\tbest: 0.9135281 (605)\ttotal: 4m 14s\tremaining: 11h 31m 26s\n",
      "609:\tlearn: 0.4006692\ttest: 0.9135738\tbest: 0.9135281 (605)\ttotal: 4m 14s\tremaining: 11h 31m 26s\n",
      "610:\tlearn: 0.4003552\ttest: 0.9135619\tbest: 0.9135281 (605)\ttotal: 4m 15s\tremaining: 11h 31m 27s\n",
      "611:\tlearn: 0.4000875\ttest: 0.9135186\tbest: 0.9135186 (611)\ttotal: 4m 15s\tremaining: 11h 31m 26s\n",
      "612:\tlearn: 0.3998188\ttest: 0.9133692\tbest: 0.9133692 (612)\ttotal: 4m 15s\tremaining: 11h 31m 23s\n",
      "613:\tlearn: 0.3994824\ttest: 0.9130149\tbest: 0.9130149 (613)\ttotal: 4m 16s\tremaining: 11h 31m 24s\n",
      "614:\tlearn: 0.3991146\ttest: 0.9129856\tbest: 0.9129856 (614)\ttotal: 4m 16s\tremaining: 11h 31m 25s\n",
      "615:\tlearn: 0.3988671\ttest: 0.9130196\tbest: 0.9129856 (614)\ttotal: 4m 17s\tremaining: 11h 31m 24s\n",
      "616:\tlearn: 0.3985869\ttest: 0.9126539\tbest: 0.9126539 (616)\ttotal: 4m 17s\tremaining: 11h 31m 24s\n",
      "617:\tlearn: 0.3982859\ttest: 0.9124784\tbest: 0.9124784 (617)\ttotal: 4m 17s\tremaining: 11h 31m 26s\n",
      "618:\tlearn: 0.3979642\ttest: 0.9122624\tbest: 0.9122624 (618)\ttotal: 4m 18s\tremaining: 11h 31m 25s\n",
      "619:\tlearn: 0.3976834\ttest: 0.9123196\tbest: 0.9122624 (618)\ttotal: 4m 18s\tremaining: 11h 31m 24s\n",
      "620:\tlearn: 0.3974315\ttest: 0.9121594\tbest: 0.9121594 (620)\ttotal: 4m 19s\tremaining: 11h 31m 26s\n",
      "621:\tlearn: 0.3972015\ttest: 0.9121012\tbest: 0.9121012 (621)\ttotal: 4m 19s\tremaining: 11h 31m 24s\n",
      "622:\tlearn: 0.3969431\ttest: 0.9119889\tbest: 0.9119889 (622)\ttotal: 4m 20s\tremaining: 11h 31m 24s\n",
      "623:\tlearn: 0.3966559\ttest: 0.9118247\tbest: 0.9118247 (623)\ttotal: 4m 20s\tremaining: 11h 31m 23s\n",
      "624:\tlearn: 0.3963823\ttest: 0.9118061\tbest: 0.9118061 (624)\ttotal: 4m 20s\tremaining: 11h 31m 22s\n",
      "625:\tlearn: 0.3961299\ttest: 0.9118154\tbest: 0.9118061 (624)\ttotal: 4m 21s\tremaining: 11h 31m 21s\n",
      "626:\tlearn: 0.3958300\ttest: 0.9117156\tbest: 0.9117156 (626)\ttotal: 4m 21s\tremaining: 11h 31m 22s\n",
      "627:\tlearn: 0.3955392\ttest: 0.9117404\tbest: 0.9117156 (626)\ttotal: 4m 22s\tremaining: 11h 31m 27s\n",
      "628:\tlearn: 0.3952594\ttest: 0.9117383\tbest: 0.9117156 (626)\ttotal: 4m 22s\tremaining: 11h 31m 31s\n",
      "629:\tlearn: 0.3950136\ttest: 0.9116727\tbest: 0.9116727 (629)\ttotal: 4m 23s\tremaining: 11h 31m 33s\n",
      "630:\tlearn: 0.3947355\ttest: 0.9117409\tbest: 0.9116727 (629)\ttotal: 4m 23s\tremaining: 11h 31m 32s\n",
      "631:\tlearn: 0.3945590\ttest: 0.9116742\tbest: 0.9116727 (629)\ttotal: 4m 23s\tremaining: 11h 31m 32s\n",
      "632:\tlearn: 0.3942995\ttest: 0.9115635\tbest: 0.9115635 (632)\ttotal: 4m 24s\tremaining: 11h 31m 37s\n",
      "633:\tlearn: 0.3940312\ttest: 0.9114225\tbest: 0.9114225 (633)\ttotal: 4m 24s\tremaining: 11h 31m 36s\n",
      "634:\tlearn: 0.3937967\ttest: 0.9114770\tbest: 0.9114225 (633)\ttotal: 4m 25s\tremaining: 11h 31m 38s\n",
      "635:\tlearn: 0.3935536\ttest: 0.9113150\tbest: 0.9113150 (635)\ttotal: 4m 25s\tremaining: 11h 31m 36s\n",
      "636:\tlearn: 0.3932838\ttest: 0.9114798\tbest: 0.9113150 (635)\ttotal: 4m 26s\tremaining: 11h 31m 34s\n",
      "637:\tlearn: 0.3929690\ttest: 0.9115030\tbest: 0.9113150 (635)\ttotal: 4m 26s\tremaining: 11h 31m 34s\n",
      "638:\tlearn: 0.3926920\ttest: 0.9115981\tbest: 0.9113150 (635)\ttotal: 4m 26s\tremaining: 11h 31m 34s\n",
      "639:\tlearn: 0.3923428\ttest: 0.9116009\tbest: 0.9113150 (635)\ttotal: 4m 27s\tremaining: 11h 31m 34s\n",
      "640:\tlearn: 0.3920556\ttest: 0.9114757\tbest: 0.9113150 (635)\ttotal: 4m 27s\tremaining: 11h 31m 34s\n",
      "641:\tlearn: 0.3918091\ttest: 0.9115590\tbest: 0.9113150 (635)\ttotal: 4m 28s\tremaining: 11h 31m 33s\n",
      "642:\tlearn: 0.3914965\ttest: 0.9116258\tbest: 0.9113150 (635)\ttotal: 4m 28s\tremaining: 11h 31m 33s\n",
      "643:\tlearn: 0.3912835\ttest: 0.9114036\tbest: 0.9113150 (635)\ttotal: 4m 28s\tremaining: 11h 31m 33s\n",
      "644:\tlearn: 0.3910582\ttest: 0.9113625\tbest: 0.9113150 (635)\ttotal: 4m 29s\tremaining: 11h 31m 32s\n",
      "645:\tlearn: 0.3907560\ttest: 0.9113815\tbest: 0.9113150 (635)\ttotal: 4m 29s\tremaining: 11h 31m 32s\n",
      "646:\tlearn: 0.3905088\ttest: 0.9113353\tbest: 0.9113150 (635)\ttotal: 4m 30s\tremaining: 11h 31m 33s\n",
      "647:\tlearn: 0.3903301\ttest: 0.9114641\tbest: 0.9113150 (635)\ttotal: 4m 30s\tremaining: 11h 31m 32s\n",
      "648:\tlearn: 0.3900772\ttest: 0.9114516\tbest: 0.9113150 (635)\ttotal: 4m 31s\tremaining: 11h 31m 29s\n",
      "649:\tlearn: 0.3898361\ttest: 0.9113773\tbest: 0.9113150 (635)\ttotal: 4m 31s\tremaining: 11h 31m 29s\n",
      "650:\tlearn: 0.3895684\ttest: 0.9114644\tbest: 0.9113150 (635)\ttotal: 4m 31s\tremaining: 11h 31m 28s\n",
      "651:\tlearn: 0.3893491\ttest: 0.9115051\tbest: 0.9113150 (635)\ttotal: 4m 32s\tremaining: 11h 31m 30s\n",
      "652:\tlearn: 0.3891330\ttest: 0.9115589\tbest: 0.9113150 (635)\ttotal: 4m 32s\tremaining: 11h 31m 31s\n",
      "653:\tlearn: 0.3888453\ttest: 0.9112386\tbest: 0.9112386 (653)\ttotal: 4m 33s\tremaining: 11h 31m 32s\n",
      "654:\tlearn: 0.3885883\ttest: 0.9111992\tbest: 0.9111992 (654)\ttotal: 4m 33s\tremaining: 11h 31m 30s\n",
      "655:\tlearn: 0.3883752\ttest: 0.9111256\tbest: 0.9111256 (655)\ttotal: 4m 33s\tremaining: 11h 31m 31s\n",
      "656:\tlearn: 0.3881133\ttest: 0.9109297\tbest: 0.9109297 (656)\ttotal: 4m 34s\tremaining: 11h 31m 29s\n",
      "657:\tlearn: 0.3878716\ttest: 0.9109521\tbest: 0.9109297 (656)\ttotal: 4m 34s\tremaining: 11h 31m 29s\n",
      "658:\tlearn: 0.3876829\ttest: 0.9108612\tbest: 0.9108612 (658)\ttotal: 4m 35s\tremaining: 11h 31m 28s\n",
      "659:\tlearn: 0.3874099\ttest: 0.9107501\tbest: 0.9107501 (659)\ttotal: 4m 35s\tremaining: 11h 31m 27s\n",
      "660:\tlearn: 0.3871556\ttest: 0.9107058\tbest: 0.9107058 (660)\ttotal: 4m 36s\tremaining: 11h 31m 26s\n",
      "661:\tlearn: 0.3869209\ttest: 0.9106957\tbest: 0.9106957 (661)\ttotal: 4m 36s\tremaining: 11h 31m 25s\n",
      "662:\tlearn: 0.3867059\ttest: 0.9107280\tbest: 0.9106957 (661)\ttotal: 4m 36s\tremaining: 11h 31m 24s\n",
      "663:\tlearn: 0.3864852\ttest: 0.9107035\tbest: 0.9106957 (661)\ttotal: 4m 37s\tremaining: 11h 31m 25s\n",
      "664:\tlearn: 0.3862658\ttest: 0.9106130\tbest: 0.9106130 (664)\ttotal: 4m 37s\tremaining: 11h 31m 25s\n",
      "665:\tlearn: 0.3860092\ttest: 0.9108692\tbest: 0.9106130 (664)\ttotal: 4m 38s\tremaining: 11h 31m 25s\n",
      "666:\tlearn: 0.3857962\ttest: 0.9109469\tbest: 0.9106130 (664)\ttotal: 4m 38s\tremaining: 11h 31m 26s\n",
      "667:\tlearn: 0.3855719\ttest: 0.9108617\tbest: 0.9106130 (664)\ttotal: 4m 38s\tremaining: 11h 31m 26s\n",
      "668:\tlearn: 0.3853316\ttest: 0.9107678\tbest: 0.9106130 (664)\ttotal: 4m 39s\tremaining: 11h 31m 24s\n",
      "669:\tlearn: 0.3851035\ttest: 0.9106798\tbest: 0.9106130 (664)\ttotal: 4m 39s\tremaining: 11h 31m 23s\n",
      "670:\tlearn: 0.3848408\ttest: 0.9105321\tbest: 0.9105321 (670)\ttotal: 4m 40s\tremaining: 11h 31m 26s\n",
      "671:\tlearn: 0.3845799\ttest: 0.9104523\tbest: 0.9104523 (671)\ttotal: 4m 40s\tremaining: 11h 31m 24s\n",
      "672:\tlearn: 0.3843541\ttest: 0.9103450\tbest: 0.9103450 (672)\ttotal: 4m 41s\tremaining: 11h 31m 23s\n",
      "673:\tlearn: 0.3841300\ttest: 0.9103094\tbest: 0.9103094 (673)\ttotal: 4m 41s\tremaining: 11h 31m 22s\n",
      "674:\tlearn: 0.3838933\ttest: 0.9103660\tbest: 0.9103094 (673)\ttotal: 4m 41s\tremaining: 11h 31m 21s\n",
      "675:\tlearn: 0.3836119\ttest: 0.9105084\tbest: 0.9103094 (673)\ttotal: 4m 42s\tremaining: 11h 31m 20s\n",
      "676:\tlearn: 0.3833865\ttest: 0.9105074\tbest: 0.9103094 (673)\ttotal: 4m 42s\tremaining: 11h 31m 19s\n",
      "677:\tlearn: 0.3831682\ttest: 0.9103414\tbest: 0.9103094 (673)\ttotal: 4m 43s\tremaining: 11h 31m 20s\n",
      "678:\tlearn: 0.3828997\ttest: 0.9102272\tbest: 0.9102272 (678)\ttotal: 4m 43s\tremaining: 11h 31m 20s\n",
      "679:\tlearn: 0.3826904\ttest: 0.9103509\tbest: 0.9102272 (678)\ttotal: 4m 43s\tremaining: 11h 31m 19s\n",
      "680:\tlearn: 0.3824605\ttest: 0.9103459\tbest: 0.9102272 (678)\ttotal: 4m 44s\tremaining: 11h 31m 18s\n",
      "681:\tlearn: 0.3822574\ttest: 0.9105490\tbest: 0.9102272 (678)\ttotal: 4m 44s\tremaining: 11h 31m 18s\n",
      "682:\tlearn: 0.3820685\ttest: 0.9103884\tbest: 0.9102272 (678)\ttotal: 4m 45s\tremaining: 11h 31m 20s\n",
      "683:\tlearn: 0.3818604\ttest: 0.9101940\tbest: 0.9101940 (683)\ttotal: 4m 45s\tremaining: 11h 31m 23s\n",
      "684:\tlearn: 0.3816755\ttest: 0.9102557\tbest: 0.9101940 (683)\ttotal: 4m 46s\tremaining: 11h 31m 23s\n",
      "685:\tlearn: 0.3813676\ttest: 0.9102473\tbest: 0.9101940 (683)\ttotal: 4m 46s\tremaining: 11h 31m 21s\n",
      "686:\tlearn: 0.3811658\ttest: 0.9101773\tbest: 0.9101773 (686)\ttotal: 4m 46s\tremaining: 11h 31m 21s\n",
      "687:\tlearn: 0.3809277\ttest: 0.9102431\tbest: 0.9101773 (686)\ttotal: 4m 47s\tremaining: 11h 31m 20s\n",
      "688:\tlearn: 0.3807083\ttest: 0.9102564\tbest: 0.9101773 (686)\ttotal: 4m 47s\tremaining: 11h 31m 22s\n",
      "689:\tlearn: 0.3805133\ttest: 0.9103838\tbest: 0.9101773 (686)\ttotal: 4m 48s\tremaining: 11h 31m 19s\n",
      "690:\tlearn: 0.3802717\ttest: 0.9103976\tbest: 0.9101773 (686)\ttotal: 4m 48s\tremaining: 11h 31m 17s\n",
      "691:\tlearn: 0.3800492\ttest: 0.9102159\tbest: 0.9101773 (686)\ttotal: 4m 49s\tremaining: 11h 31m 14s\n",
      "692:\tlearn: 0.3798436\ttest: 0.9102432\tbest: 0.9101773 (686)\ttotal: 4m 49s\tremaining: 11h 31m 12s\n",
      "693:\tlearn: 0.3795967\ttest: 0.9102222\tbest: 0.9101773 (686)\ttotal: 4m 49s\tremaining: 11h 31m 10s\n",
      "694:\tlearn: 0.3793772\ttest: 0.9103081\tbest: 0.9101773 (686)\ttotal: 4m 50s\tremaining: 11h 31m 9s\n",
      "695:\tlearn: 0.3791231\ttest: 0.9102747\tbest: 0.9101773 (686)\ttotal: 4m 50s\tremaining: 11h 31m 7s\n",
      "696:\tlearn: 0.3789011\ttest: 0.9102832\tbest: 0.9101773 (686)\ttotal: 4m 51s\tremaining: 11h 31m 6s\n",
      "697:\tlearn: 0.3786533\ttest: 0.9104008\tbest: 0.9101773 (686)\ttotal: 4m 51s\tremaining: 11h 31m 5s\n",
      "698:\tlearn: 0.3784642\ttest: 0.9102894\tbest: 0.9101773 (686)\ttotal: 4m 51s\tremaining: 11h 31m 4s\n",
      "699:\tlearn: 0.3782544\ttest: 0.9104163\tbest: 0.9101773 (686)\ttotal: 4m 52s\tremaining: 11h 31m 3s\n",
      "700:\tlearn: 0.3780398\ttest: 0.9104666\tbest: 0.9101773 (686)\ttotal: 4m 52s\tremaining: 11h 31m 3s\n",
      "701:\tlearn: 0.3777922\ttest: 0.9104512\tbest: 0.9101773 (686)\ttotal: 4m 53s\tremaining: 11h 31m 4s\n",
      "702:\tlearn: 0.3775373\ttest: 0.9104649\tbest: 0.9101773 (686)\ttotal: 4m 53s\tremaining: 11h 31m 5s\n",
      "703:\tlearn: 0.3772803\ttest: 0.9106941\tbest: 0.9101773 (686)\ttotal: 4m 54s\tremaining: 11h 31m 7s\n",
      "704:\tlearn: 0.3770417\ttest: 0.9108039\tbest: 0.9101773 (686)\ttotal: 4m 54s\tremaining: 11h 31m 7s\n",
      "705:\tlearn: 0.3768376\ttest: 0.9108381\tbest: 0.9101773 (686)\ttotal: 4m 54s\tremaining: 11h 31m 10s\n",
      "706:\tlearn: 0.3766414\ttest: 0.9108965\tbest: 0.9101773 (686)\ttotal: 4m 55s\tremaining: 11h 31m 9s\n",
      "707:\tlearn: 0.3764504\ttest: 0.9107971\tbest: 0.9101773 (686)\ttotal: 4m 55s\tremaining: 11h 31m 9s\n",
      "708:\tlearn: 0.3762518\ttest: 0.9107732\tbest: 0.9101773 (686)\ttotal: 4m 56s\tremaining: 11h 31m 7s\n",
      "709:\tlearn: 0.3760062\ttest: 0.9105657\tbest: 0.9101773 (686)\ttotal: 4m 56s\tremaining: 11h 31m 5s\n",
      "710:\tlearn: 0.3757778\ttest: 0.9105603\tbest: 0.9101773 (686)\ttotal: 4m 56s\tremaining: 11h 31m 5s\n",
      "711:\tlearn: 0.3755087\ttest: 0.9105541\tbest: 0.9101773 (686)\ttotal: 4m 57s\tremaining: 11h 31m 5s\n",
      "712:\tlearn: 0.3752759\ttest: 0.9106112\tbest: 0.9101773 (686)\ttotal: 4m 57s\tremaining: 11h 31m 5s\n",
      "713:\tlearn: 0.3750302\ttest: 0.9106516\tbest: 0.9101773 (686)\ttotal: 4m 58s\tremaining: 11h 31m 7s\n",
      "714:\tlearn: 0.3748769\ttest: 0.9106155\tbest: 0.9101773 (686)\ttotal: 4m 58s\tremaining: 11h 31m 6s\n",
      "715:\tlearn: 0.3746559\ttest: 0.9104395\tbest: 0.9101773 (686)\ttotal: 4m 59s\tremaining: 11h 31m 6s\n",
      "716:\tlearn: 0.3744399\ttest: 0.9104588\tbest: 0.9101773 (686)\ttotal: 4m 59s\tremaining: 11h 31m 5s\n",
      "717:\tlearn: 0.3742387\ttest: 0.9105326\tbest: 0.9101773 (686)\ttotal: 4m 59s\tremaining: 11h 31m 5s\n",
      "718:\tlearn: 0.3739693\ttest: 0.9104286\tbest: 0.9101773 (686)\ttotal: 5m\tremaining: 11h 31m 6s\n",
      "719:\tlearn: 0.3736040\ttest: 0.9103783\tbest: 0.9101773 (686)\ttotal: 5m\tremaining: 11h 31m 3s\n",
      "720:\tlearn: 0.3734442\ttest: 0.9102666\tbest: 0.9101773 (686)\ttotal: 5m 1s\tremaining: 11h 31m 1s\n",
      "721:\tlearn: 0.3732553\ttest: 0.9101818\tbest: 0.9101773 (686)\ttotal: 5m 1s\tremaining: 11h 31m\n",
      "722:\tlearn: 0.3730497\ttest: 0.9102447\tbest: 0.9101773 (686)\ttotal: 5m 1s\tremaining: 11h 31m 1s\n",
      "723:\tlearn: 0.3728562\ttest: 0.9102445\tbest: 0.9101773 (686)\ttotal: 5m 2s\tremaining: 11h 31m 1s\n",
      "724:\tlearn: 0.3726658\ttest: 0.9102138\tbest: 0.9101773 (686)\ttotal: 5m 2s\tremaining: 11h 31m\n",
      "725:\tlearn: 0.3724762\ttest: 0.9101580\tbest: 0.9101580 (725)\ttotal: 5m 3s\tremaining: 11h 30m 58s\n",
      "726:\tlearn: 0.3722748\ttest: 0.9102399\tbest: 0.9101580 (725)\ttotal: 5m 3s\tremaining: 11h 30m 56s\n",
      "727:\tlearn: 0.3720266\ttest: 0.9100189\tbest: 0.9100189 (727)\ttotal: 5m 3s\tremaining: 11h 30m 53s\n",
      "728:\tlearn: 0.3718032\ttest: 0.9099442\tbest: 0.9099442 (728)\ttotal: 5m 4s\tremaining: 11h 30m 52s\n",
      "729:\tlearn: 0.3715999\ttest: 0.9099726\tbest: 0.9099442 (728)\ttotal: 5m 4s\tremaining: 11h 30m 50s\n",
      "730:\tlearn: 0.3713851\ttest: 0.9097378\tbest: 0.9097378 (730)\ttotal: 5m 5s\tremaining: 11h 30m 48s\n",
      "731:\tlearn: 0.3711489\ttest: 0.9096676\tbest: 0.9096676 (731)\ttotal: 5m 5s\tremaining: 11h 30m 46s\n",
      "732:\tlearn: 0.3709247\ttest: 0.9095530\tbest: 0.9095530 (732)\ttotal: 5m 6s\tremaining: 11h 30m 46s\n",
      "733:\tlearn: 0.3706892\ttest: 0.9095987\tbest: 0.9095530 (732)\ttotal: 5m 6s\tremaining: 11h 30m 44s\n",
      "734:\tlearn: 0.3704786\ttest: 0.9096407\tbest: 0.9095530 (732)\ttotal: 5m 6s\tremaining: 11h 30m 43s\n",
      "735:\tlearn: 0.3702656\ttest: 0.9096283\tbest: 0.9095530 (732)\ttotal: 5m 7s\tremaining: 11h 30m 41s\n",
      "736:\tlearn: 0.3700529\ttest: 0.9096268\tbest: 0.9095530 (732)\ttotal: 5m 7s\tremaining: 11h 30m 42s\n",
      "737:\tlearn: 0.3698579\ttest: 0.9092797\tbest: 0.9092797 (737)\ttotal: 5m 8s\tremaining: 11h 30m 43s\n",
      "738:\tlearn: 0.3696257\ttest: 0.9092299\tbest: 0.9092299 (738)\ttotal: 5m 8s\tremaining: 11h 30m 44s\n",
      "739:\tlearn: 0.3693540\ttest: 0.9090782\tbest: 0.9090782 (739)\ttotal: 5m 8s\tremaining: 11h 30m 43s\n",
      "740:\tlearn: 0.3691047\ttest: 0.9089357\tbest: 0.9089357 (740)\ttotal: 5m 9s\tremaining: 11h 30m 44s\n",
      "741:\tlearn: 0.3689032\ttest: 0.9088678\tbest: 0.9088678 (741)\ttotal: 5m 9s\tremaining: 11h 30m 43s\n",
      "742:\tlearn: 0.3686712\ttest: 0.9088585\tbest: 0.9088585 (742)\ttotal: 5m 10s\tremaining: 11h 30m 42s\n",
      "743:\tlearn: 0.3684554\ttest: 0.9088897\tbest: 0.9088585 (742)\ttotal: 5m 10s\tremaining: 11h 30m 41s\n",
      "744:\tlearn: 0.3682445\ttest: 0.9089854\tbest: 0.9088585 (742)\ttotal: 5m 11s\tremaining: 11h 30m 40s\n",
      "745:\tlearn: 0.3680271\ttest: 0.9089550\tbest: 0.9088585 (742)\ttotal: 5m 11s\tremaining: 11h 30m 39s\n",
      "746:\tlearn: 0.3678030\ttest: 0.9089208\tbest: 0.9088585 (742)\ttotal: 5m 11s\tremaining: 11h 30m 39s\n",
      "747:\tlearn: 0.3676217\ttest: 0.9088146\tbest: 0.9088146 (747)\ttotal: 5m 12s\tremaining: 11h 30m 39s\n",
      "748:\tlearn: 0.3674051\ttest: 0.9089494\tbest: 0.9088146 (747)\ttotal: 5m 12s\tremaining: 11h 30m 38s\n",
      "749:\tlearn: 0.3671522\ttest: 0.9089544\tbest: 0.9088146 (747)\ttotal: 5m 13s\tremaining: 11h 30m 40s\n",
      "750:\tlearn: 0.3669936\ttest: 0.9091277\tbest: 0.9088146 (747)\ttotal: 5m 13s\tremaining: 11h 30m 42s\n",
      "751:\tlearn: 0.3668382\ttest: 0.9090420\tbest: 0.9088146 (747)\ttotal: 5m 14s\tremaining: 11h 30m 44s\n",
      "752:\tlearn: 0.3665966\ttest: 0.9091109\tbest: 0.9088146 (747)\ttotal: 5m 14s\tremaining: 11h 30m 46s\n",
      "753:\tlearn: 0.3663933\ttest: 0.9090727\tbest: 0.9088146 (747)\ttotal: 5m 14s\tremaining: 11h 30m 48s\n",
      "754:\tlearn: 0.3662048\ttest: 0.9090417\tbest: 0.9088146 (747)\ttotal: 5m 15s\tremaining: 11h 30m 50s\n",
      "755:\tlearn: 0.3659927\ttest: 0.9089706\tbest: 0.9088146 (747)\ttotal: 5m 15s\tremaining: 11h 30m 53s\n",
      "756:\tlearn: 0.3658426\ttest: 0.9089783\tbest: 0.9088146 (747)\ttotal: 5m 16s\tremaining: 11h 30m 51s\n",
      "757:\tlearn: 0.3656361\ttest: 0.9091015\tbest: 0.9088146 (747)\ttotal: 5m 16s\tremaining: 11h 30m 50s\n",
      "758:\tlearn: 0.3654540\ttest: 0.9091621\tbest: 0.9088146 (747)\ttotal: 5m 17s\tremaining: 11h 30m 49s\n",
      "759:\tlearn: 0.3652538\ttest: 0.9091352\tbest: 0.9088146 (747)\ttotal: 5m 17s\tremaining: 11h 30m 51s\n",
      "760:\tlearn: 0.3650394\ttest: 0.9091566\tbest: 0.9088146 (747)\ttotal: 5m 17s\tremaining: 11h 30m 49s\n",
      "761:\tlearn: 0.3647773\ttest: 0.9091781\tbest: 0.9088146 (747)\ttotal: 5m 18s\tremaining: 11h 30m 53s\n",
      "762:\tlearn: 0.3645146\ttest: 0.9090801\tbest: 0.9088146 (747)\ttotal: 5m 18s\tremaining: 11h 30m 50s\n",
      "763:\tlearn: 0.3643304\ttest: 0.9090565\tbest: 0.9088146 (747)\ttotal: 5m 19s\tremaining: 11h 30m 50s\n",
      "764:\tlearn: 0.3641419\ttest: 0.9090843\tbest: 0.9088146 (747)\ttotal: 5m 19s\tremaining: 11h 30m 48s\n",
      "765:\tlearn: 0.3639600\ttest: 0.9091264\tbest: 0.9088146 (747)\ttotal: 5m 19s\tremaining: 11h 30m 46s\n",
      "766:\tlearn: 0.3637425\ttest: 0.9090807\tbest: 0.9088146 (747)\ttotal: 5m 20s\tremaining: 11h 30m 45s\n",
      "767:\tlearn: 0.3635438\ttest: 0.9090552\tbest: 0.9088146 (747)\ttotal: 5m 20s\tremaining: 11h 30m 44s\n",
      "768:\tlearn: 0.3633632\ttest: 0.9087500\tbest: 0.9087500 (768)\ttotal: 5m 21s\tremaining: 11h 30m 43s\n",
      "769:\tlearn: 0.3631886\ttest: 0.9086384\tbest: 0.9086384 (769)\ttotal: 5m 21s\tremaining: 11h 30m 41s\n",
      "770:\tlearn: 0.3630077\ttest: 0.9086436\tbest: 0.9086384 (769)\ttotal: 5m 21s\tremaining: 11h 30m 39s\n",
      "771:\tlearn: 0.3628118\ttest: 0.9085900\tbest: 0.9085900 (771)\ttotal: 5m 22s\tremaining: 11h 30m 38s\n",
      "772:\tlearn: 0.3626705\ttest: 0.9087329\tbest: 0.9085900 (771)\ttotal: 5m 22s\tremaining: 11h 30m 38s\n",
      "773:\tlearn: 0.3624741\ttest: 0.9086778\tbest: 0.9085900 (771)\ttotal: 5m 23s\tremaining: 11h 30m 37s\n",
      "774:\tlearn: 0.3622933\ttest: 0.9086108\tbest: 0.9085900 (771)\ttotal: 5m 23s\tremaining: 11h 30m 35s\n",
      "775:\tlearn: 0.3621311\ttest: 0.9084432\tbest: 0.9084432 (775)\ttotal: 5m 24s\tremaining: 11h 30m 35s\n",
      "776:\tlearn: 0.3619298\ttest: 0.9084502\tbest: 0.9084432 (775)\ttotal: 5m 24s\tremaining: 11h 30m 33s\n",
      "777:\tlearn: 0.3617331\ttest: 0.9083612\tbest: 0.9083612 (777)\ttotal: 5m 24s\tremaining: 11h 30m 31s\n",
      "778:\tlearn: 0.3615289\ttest: 0.9084375\tbest: 0.9083612 (777)\ttotal: 5m 25s\tremaining: 11h 30m 29s\n",
      "779:\tlearn: 0.3613167\ttest: 0.9084874\tbest: 0.9083612 (777)\ttotal: 5m 25s\tremaining: 11h 30m 28s\n",
      "780:\tlearn: 0.3611046\ttest: 0.9084853\tbest: 0.9083612 (777)\ttotal: 5m 26s\tremaining: 11h 30m 27s\n",
      "781:\tlearn: 0.3608982\ttest: 0.9084529\tbest: 0.9083612 (777)\ttotal: 5m 26s\tremaining: 11h 30m 26s\n",
      "782:\tlearn: 0.3607224\ttest: 0.9083780\tbest: 0.9083612 (777)\ttotal: 5m 26s\tremaining: 11h 30m 25s\n",
      "783:\tlearn: 0.3605521\ttest: 0.9082835\tbest: 0.9082835 (783)\ttotal: 5m 27s\tremaining: 11h 30m 24s\n",
      "784:\tlearn: 0.3603540\ttest: 0.9081516\tbest: 0.9081516 (784)\ttotal: 5m 27s\tremaining: 11h 30m 27s\n",
      "785:\tlearn: 0.3601814\ttest: 0.9081941\tbest: 0.9081516 (784)\ttotal: 5m 28s\tremaining: 11h 30m 26s\n",
      "786:\tlearn: 0.3600162\ttest: 0.9082015\tbest: 0.9081516 (784)\ttotal: 5m 28s\tremaining: 11h 30m 25s\n",
      "787:\tlearn: 0.3598485\ttest: 0.9081938\tbest: 0.9081516 (784)\ttotal: 5m 29s\tremaining: 11h 30m 26s\n",
      "788:\tlearn: 0.3596660\ttest: 0.9081768\tbest: 0.9081516 (784)\ttotal: 5m 29s\tremaining: 11h 30m 24s\n",
      "789:\tlearn: 0.3594683\ttest: 0.9082369\tbest: 0.9081516 (784)\ttotal: 5m 29s\tremaining: 11h 30m 23s\n",
      "790:\tlearn: 0.3593047\ttest: 0.9082836\tbest: 0.9081516 (784)\ttotal: 5m 30s\tremaining: 11h 30m 23s\n",
      "791:\tlearn: 0.3591210\ttest: 0.9082320\tbest: 0.9081516 (784)\ttotal: 5m 30s\tremaining: 11h 30m 24s\n",
      "792:\tlearn: 0.3589650\ttest: 0.9081676\tbest: 0.9081516 (784)\ttotal: 5m 31s\tremaining: 11h 30m 24s\n",
      "793:\tlearn: 0.3587873\ttest: 0.9080825\tbest: 0.9080825 (793)\ttotal: 5m 31s\tremaining: 11h 30m 22s\n",
      "794:\tlearn: 0.3586082\ttest: 0.9080440\tbest: 0.9080440 (794)\ttotal: 5m 31s\tremaining: 11h 30m 20s\n",
      "795:\tlearn: 0.3584482\ttest: 0.9080677\tbest: 0.9080440 (794)\ttotal: 5m 32s\tremaining: 11h 30m 19s\n",
      "796:\tlearn: 0.3582648\ttest: 0.9081934\tbest: 0.9080440 (794)\ttotal: 5m 32s\tremaining: 11h 30m 20s\n",
      "797:\tlearn: 0.3580780\ttest: 0.9081451\tbest: 0.9080440 (794)\ttotal: 5m 33s\tremaining: 11h 30m 20s\n",
      "798:\tlearn: 0.3578776\ttest: 0.9081448\tbest: 0.9080440 (794)\ttotal: 5m 33s\tremaining: 11h 30m 18s\n",
      "799:\tlearn: 0.3576843\ttest: 0.9082178\tbest: 0.9080440 (794)\ttotal: 5m 34s\tremaining: 11h 30m 18s\n",
      "800:\tlearn: 0.3574354\ttest: 0.9085314\tbest: 0.9080440 (794)\ttotal: 5m 34s\tremaining: 11h 30m 17s\n",
      "801:\tlearn: 0.3572250\ttest: 0.9084737\tbest: 0.9080440 (794)\ttotal: 5m 34s\tremaining: 11h 30m 15s\n",
      "802:\tlearn: 0.3570519\ttest: 0.9084449\tbest: 0.9080440 (794)\ttotal: 5m 35s\tremaining: 11h 30m 16s\n",
      "803:\tlearn: 0.3568689\ttest: 0.9082107\tbest: 0.9080440 (794)\ttotal: 5m 35s\tremaining: 11h 30m 16s\n",
      "804:\tlearn: 0.3567085\ttest: 0.9082335\tbest: 0.9080440 (794)\ttotal: 5m 36s\tremaining: 11h 30m 17s\n",
      "805:\tlearn: 0.3565450\ttest: 0.9082843\tbest: 0.9080440 (794)\ttotal: 5m 36s\tremaining: 11h 30m 14s\n",
      "806:\tlearn: 0.3563573\ttest: 0.9081798\tbest: 0.9080440 (794)\ttotal: 5m 36s\tremaining: 11h 30m 13s\n",
      "807:\tlearn: 0.3561237\ttest: 0.9081928\tbest: 0.9080440 (794)\ttotal: 5m 37s\tremaining: 11h 30m 12s\n",
      "808:\tlearn: 0.3559714\ttest: 0.9080239\tbest: 0.9080239 (808)\ttotal: 5m 37s\tremaining: 11h 30m 12s\n",
      "809:\tlearn: 0.3558100\ttest: 0.9079559\tbest: 0.9079559 (809)\ttotal: 5m 38s\tremaining: 11h 30m 12s\n",
      "810:\tlearn: 0.3556237\ttest: 0.9079172\tbest: 0.9079172 (810)\ttotal: 5m 38s\tremaining: 11h 30m 11s\n",
      "811:\tlearn: 0.3554185\ttest: 0.9079219\tbest: 0.9079172 (810)\ttotal: 5m 39s\tremaining: 11h 30m 11s\n",
      "812:\tlearn: 0.3552402\ttest: 0.9078885\tbest: 0.9078885 (812)\ttotal: 5m 39s\tremaining: 11h 30m 9s\n",
      "813:\tlearn: 0.3550265\ttest: 0.9079399\tbest: 0.9078885 (812)\ttotal: 5m 39s\tremaining: 11h 30m 7s\n",
      "814:\tlearn: 0.3548016\ttest: 0.9080530\tbest: 0.9078885 (812)\ttotal: 5m 40s\tremaining: 11h 30m 7s\n",
      "815:\tlearn: 0.3546002\ttest: 0.9082137\tbest: 0.9078885 (812)\ttotal: 5m 40s\tremaining: 11h 30m 5s\n",
      "816:\tlearn: 0.3544368\ttest: 0.9081688\tbest: 0.9078885 (812)\ttotal: 5m 41s\tremaining: 11h 30m 3s\n",
      "817:\tlearn: 0.3542253\ttest: 0.9081550\tbest: 0.9078885 (812)\ttotal: 5m 41s\tremaining: 11h 30m 1s\n",
      "818:\tlearn: 0.3540232\ttest: 0.9077982\tbest: 0.9077982 (818)\ttotal: 5m 41s\tremaining: 11h 30m\n",
      "819:\tlearn: 0.3538514\ttest: 0.9077531\tbest: 0.9077531 (819)\ttotal: 5m 42s\tremaining: 11h 30m\n",
      "820:\tlearn: 0.3537252\ttest: 0.9077065\tbest: 0.9077065 (820)\ttotal: 5m 42s\tremaining: 11h 30m\n",
      "821:\tlearn: 0.3535493\ttest: 0.9076205\tbest: 0.9076205 (821)\ttotal: 5m 43s\tremaining: 11h 29m 59s\n",
      "822:\tlearn: 0.3533333\ttest: 0.9076660\tbest: 0.9076205 (821)\ttotal: 5m 43s\tremaining: 11h 29m 59s\n",
      "823:\tlearn: 0.3531555\ttest: 0.9075886\tbest: 0.9075886 (823)\ttotal: 5m 43s\tremaining: 11h 29m 59s\n",
      "824:\tlearn: 0.3529665\ttest: 0.9076126\tbest: 0.9075886 (823)\ttotal: 5m 44s\tremaining: 11h 29m 58s\n",
      "825:\tlearn: 0.3528024\ttest: 0.9076300\tbest: 0.9075886 (823)\ttotal: 5m 44s\tremaining: 11h 29m 57s\n",
      "826:\tlearn: 0.3526193\ttest: 0.9077167\tbest: 0.9075886 (823)\ttotal: 5m 45s\tremaining: 11h 29m 59s\n",
      "827:\tlearn: 0.3524385\ttest: 0.9078319\tbest: 0.9075886 (823)\ttotal: 5m 45s\tremaining: 11h 29m 59s\n",
      "828:\tlearn: 0.3522425\ttest: 0.9078126\tbest: 0.9075886 (823)\ttotal: 5m 46s\tremaining: 11h 29m 59s\n",
      "829:\tlearn: 0.3520709\ttest: 0.9077678\tbest: 0.9075886 (823)\ttotal: 5m 46s\tremaining: 11h 29m 58s\n",
      "830:\tlearn: 0.3519161\ttest: 0.9078323\tbest: 0.9075886 (823)\ttotal: 5m 46s\tremaining: 11h 29m 57s\n",
      "831:\tlearn: 0.3517656\ttest: 0.9079003\tbest: 0.9075886 (823)\ttotal: 5m 47s\tremaining: 11h 29m 56s\n",
      "832:\tlearn: 0.3516153\ttest: 0.9079056\tbest: 0.9075886 (823)\ttotal: 5m 47s\tremaining: 11h 29m 56s\n",
      "833:\tlearn: 0.3514440\ttest: 0.9079400\tbest: 0.9075886 (823)\ttotal: 5m 48s\tremaining: 11h 29m 54s\n",
      "834:\tlearn: 0.3512655\ttest: 0.9078602\tbest: 0.9075886 (823)\ttotal: 5m 48s\tremaining: 11h 29m 57s\n",
      "835:\tlearn: 0.3510984\ttest: 0.9079347\tbest: 0.9075886 (823)\ttotal: 5m 49s\tremaining: 11h 29m 57s\n",
      "836:\tlearn: 0.3509257\ttest: 0.9078335\tbest: 0.9075886 (823)\ttotal: 5m 49s\tremaining: 11h 29m 58s\n",
      "837:\tlearn: 0.3507378\ttest: 0.9077102\tbest: 0.9075886 (823)\ttotal: 5m 49s\tremaining: 11h 29m 57s\n",
      "838:\tlearn: 0.3505548\ttest: 0.9077273\tbest: 0.9075886 (823)\ttotal: 5m 50s\tremaining: 11h 29m 58s\n",
      "839:\tlearn: 0.3503819\ttest: 0.9075764\tbest: 0.9075764 (839)\ttotal: 5m 50s\tremaining: 11h 29m 57s\n",
      "840:\tlearn: 0.3501814\ttest: 0.9079106\tbest: 0.9075764 (839)\ttotal: 5m 51s\tremaining: 11h 30m 3s\n",
      "841:\tlearn: 0.3500023\ttest: 0.9079244\tbest: 0.9075764 (839)\ttotal: 5m 51s\tremaining: 11h 30m 1s\n",
      "842:\tlearn: 0.3498136\ttest: 0.9078838\tbest: 0.9075764 (839)\ttotal: 5m 51s\tremaining: 11h 30m\n",
      "843:\tlearn: 0.3496545\ttest: 0.9078305\tbest: 0.9075764 (839)\ttotal: 5m 52s\tremaining: 11h 29m 59s\n",
      "844:\tlearn: 0.3494850\ttest: 0.9078365\tbest: 0.9075764 (839)\ttotal: 5m 52s\tremaining: 11h 30m\n",
      "845:\tlearn: 0.3493069\ttest: 0.9077028\tbest: 0.9075764 (839)\ttotal: 5m 53s\tremaining: 11h 30m 1s\n",
      "846:\tlearn: 0.3491198\ttest: 0.9075554\tbest: 0.9075554 (846)\ttotal: 5m 53s\tremaining: 11h 30m 2s\n",
      "847:\tlearn: 0.3489558\ttest: 0.9074837\tbest: 0.9074837 (847)\ttotal: 5m 54s\tremaining: 11h 30m 4s\n",
      "848:\tlearn: 0.3487552\ttest: 0.9075337\tbest: 0.9074837 (847)\ttotal: 5m 54s\tremaining: 11h 30m 1s\n",
      "849:\tlearn: 0.3485889\ttest: 0.9075432\tbest: 0.9074837 (847)\ttotal: 5m 54s\tremaining: 11h 30m 2s\n",
      "850:\tlearn: 0.3484113\ttest: 0.9075066\tbest: 0.9074837 (847)\ttotal: 5m 55s\tremaining: 11h 29m 59s\n",
      "851:\tlearn: 0.3482588\ttest: 0.9074796\tbest: 0.9074796 (851)\ttotal: 5m 55s\tremaining: 11h 29m 57s\n",
      "852:\tlearn: 0.3481078\ttest: 0.9074854\tbest: 0.9074796 (851)\ttotal: 5m 56s\tremaining: 11h 29m 56s\n",
      "853:\tlearn: 0.3479354\ttest: 0.9073075\tbest: 0.9073075 (853)\ttotal: 5m 56s\tremaining: 11h 29m 55s\n",
      "854:\tlearn: 0.3477527\ttest: 0.9072897\tbest: 0.9072897 (854)\ttotal: 5m 56s\tremaining: 11h 29m 53s\n",
      "855:\tlearn: 0.3475799\ttest: 0.9072290\tbest: 0.9072290 (855)\ttotal: 5m 57s\tremaining: 11h 29m 50s\n",
      "856:\tlearn: 0.3473586\ttest: 0.9073070\tbest: 0.9072290 (855)\ttotal: 5m 57s\tremaining: 11h 29m 50s\n",
      "857:\tlearn: 0.3472048\ttest: 0.9071702\tbest: 0.9071702 (857)\ttotal: 5m 58s\tremaining: 11h 29m 52s\n",
      "858:\tlearn: 0.3470497\ttest: 0.9071301\tbest: 0.9071301 (858)\ttotal: 5m 58s\tremaining: 11h 29m 53s\n",
      "859:\tlearn: 0.3468883\ttest: 0.9071252\tbest: 0.9071252 (859)\ttotal: 5m 59s\tremaining: 11h 29m 52s\n",
      "860:\tlearn: 0.3466673\ttest: 0.9070884\tbest: 0.9070884 (860)\ttotal: 5m 59s\tremaining: 11h 29m 50s\n",
      "861:\tlearn: 0.3464868\ttest: 0.9070335\tbest: 0.9070335 (861)\ttotal: 5m 59s\tremaining: 11h 29m 50s\n",
      "862:\tlearn: 0.3463052\ttest: 0.9069914\tbest: 0.9069914 (862)\ttotal: 6m\tremaining: 11h 29m 49s\n",
      "863:\tlearn: 0.3461700\ttest: 0.9068902\tbest: 0.9068902 (863)\ttotal: 6m\tremaining: 11h 29m 48s\n",
      "864:\tlearn: 0.3459849\ttest: 0.9068453\tbest: 0.9068453 (864)\ttotal: 6m 1s\tremaining: 11h 29m 48s\n",
      "865:\tlearn: 0.3457953\ttest: 0.9068268\tbest: 0.9068268 (865)\ttotal: 6m 1s\tremaining: 11h 29m 47s\n",
      "866:\tlearn: 0.3456497\ttest: 0.9067960\tbest: 0.9067960 (866)\ttotal: 6m 1s\tremaining: 11h 29m 46s\n",
      "867:\tlearn: 0.3455082\ttest: 0.9068157\tbest: 0.9067960 (866)\ttotal: 6m 2s\tremaining: 11h 29m 45s\n",
      "868:\tlearn: 0.3453389\ttest: 0.9068365\tbest: 0.9067960 (866)\ttotal: 6m 2s\tremaining: 11h 29m 44s\n",
      "869:\tlearn: 0.3451762\ttest: 0.9068674\tbest: 0.9067960 (866)\ttotal: 6m 3s\tremaining: 11h 29m 44s\n",
      "870:\tlearn: 0.3450361\ttest: 0.9068179\tbest: 0.9067960 (866)\ttotal: 6m 3s\tremaining: 11h 29m 42s\n",
      "871:\tlearn: 0.3448615\ttest: 0.9067972\tbest: 0.9067960 (866)\ttotal: 6m 4s\tremaining: 11h 29m 41s\n",
      "872:\tlearn: 0.3446955\ttest: 0.9068579\tbest: 0.9067960 (866)\ttotal: 6m 4s\tremaining: 11h 29m 41s\n",
      "873:\tlearn: 0.3445202\ttest: 0.9068757\tbest: 0.9067960 (866)\ttotal: 6m 4s\tremaining: 11h 29m 39s\n",
      "874:\tlearn: 0.3443213\ttest: 0.9068693\tbest: 0.9067960 (866)\ttotal: 6m 5s\tremaining: 11h 29m 38s\n",
      "875:\tlearn: 0.3441495\ttest: 0.9069454\tbest: 0.9067960 (866)\ttotal: 6m 5s\tremaining: 11h 29m 36s\n",
      "876:\tlearn: 0.3439611\ttest: 0.9069133\tbest: 0.9067960 (866)\ttotal: 6m 6s\tremaining: 11h 29m 34s\n",
      "877:\tlearn: 0.3437778\ttest: 0.9068889\tbest: 0.9067960 (866)\ttotal: 6m 6s\tremaining: 11h 29m 32s\n",
      "878:\tlearn: 0.3436127\ttest: 0.9068257\tbest: 0.9067960 (866)\ttotal: 6m 6s\tremaining: 11h 29m 32s\n",
      "879:\tlearn: 0.3434063\ttest: 0.9066189\tbest: 0.9066189 (879)\ttotal: 6m 7s\tremaining: 11h 29m 30s\n",
      "880:\tlearn: 0.3432529\ttest: 0.9064697\tbest: 0.9064697 (880)\ttotal: 6m 7s\tremaining: 11h 29m 28s\n",
      "881:\tlearn: 0.3430898\ttest: 0.9064284\tbest: 0.9064284 (881)\ttotal: 6m 8s\tremaining: 11h 29m 29s\n",
      "882:\tlearn: 0.3428657\ttest: 0.9064469\tbest: 0.9064284 (881)\ttotal: 6m 8s\tremaining: 11h 29m 28s\n",
      "883:\tlearn: 0.3426734\ttest: 0.9063577\tbest: 0.9063577 (883)\ttotal: 6m 8s\tremaining: 11h 29m 27s\n",
      "884:\tlearn: 0.3425195\ttest: 0.9062184\tbest: 0.9062184 (884)\ttotal: 6m 9s\tremaining: 11h 29m 28s\n",
      "885:\tlearn: 0.3423469\ttest: 0.9062331\tbest: 0.9062184 (884)\ttotal: 6m 9s\tremaining: 11h 29m 27s\n",
      "886:\tlearn: 0.3421955\ttest: 0.9059827\tbest: 0.9059827 (886)\ttotal: 6m 10s\tremaining: 11h 29m 27s\n",
      "887:\tlearn: 0.3420345\ttest: 0.9057992\tbest: 0.9057992 (887)\ttotal: 6m 10s\tremaining: 11h 29m 27s\n",
      "888:\tlearn: 0.3418547\ttest: 0.9057633\tbest: 0.9057633 (888)\ttotal: 6m 11s\tremaining: 11h 29m 27s\n",
      "889:\tlearn: 0.3416957\ttest: 0.9057358\tbest: 0.9057358 (889)\ttotal: 6m 11s\tremaining: 11h 29m 26s\n",
      "890:\tlearn: 0.3415116\ttest: 0.9056884\tbest: 0.9056884 (890)\ttotal: 6m 11s\tremaining: 11h 29m 26s\n",
      "891:\tlearn: 0.3413562\ttest: 0.9056933\tbest: 0.9056884 (890)\ttotal: 6m 12s\tremaining: 11h 29m 25s\n",
      "892:\tlearn: 0.3412265\ttest: 0.9056725\tbest: 0.9056725 (892)\ttotal: 6m 12s\tremaining: 11h 29m 23s\n",
      "893:\tlearn: 0.3411003\ttest: 0.9057871\tbest: 0.9056725 (892)\ttotal: 6m 13s\tremaining: 11h 29m 23s\n",
      "894:\tlearn: 0.3409099\ttest: 0.9056649\tbest: 0.9056649 (894)\ttotal: 6m 13s\tremaining: 11h 29m 22s\n",
      "895:\tlearn: 0.3407630\ttest: 0.9056034\tbest: 0.9056034 (895)\ttotal: 6m 13s\tremaining: 11h 29m 23s\n",
      "896:\tlearn: 0.3405972\ttest: 0.9056226\tbest: 0.9056034 (895)\ttotal: 6m 14s\tremaining: 11h 29m 25s\n",
      "897:\tlearn: 0.3404589\ttest: 0.9057390\tbest: 0.9056034 (895)\ttotal: 6m 14s\tremaining: 11h 29m 26s\n",
      "898:\tlearn: 0.3402857\ttest: 0.9057571\tbest: 0.9056034 (895)\ttotal: 6m 15s\tremaining: 11h 29m 26s\n",
      "899:\tlearn: 0.3401111\ttest: 0.9058816\tbest: 0.9056034 (895)\ttotal: 6m 15s\tremaining: 11h 29m 26s\n",
      "900:\tlearn: 0.3399375\ttest: 0.9058451\tbest: 0.9056034 (895)\ttotal: 6m 16s\tremaining: 11h 29m 27s\n",
      "901:\tlearn: 0.3397800\ttest: 0.9058394\tbest: 0.9056034 (895)\ttotal: 6m 16s\tremaining: 11h 29m 29s\n",
      "902:\tlearn: 0.3396278\ttest: 0.9059119\tbest: 0.9056034 (895)\ttotal: 6m 16s\tremaining: 11h 29m 28s\n",
      "903:\tlearn: 0.3394628\ttest: 0.9058120\tbest: 0.9056034 (895)\ttotal: 6m 17s\tremaining: 11h 29m 27s\n",
      "904:\tlearn: 0.3393035\ttest: 0.9058126\tbest: 0.9056034 (895)\ttotal: 6m 17s\tremaining: 11h 29m 27s\n",
      "905:\tlearn: 0.3391027\ttest: 0.9058146\tbest: 0.9056034 (895)\ttotal: 6m 18s\tremaining: 11h 29m 28s\n",
      "906:\tlearn: 0.3389541\ttest: 0.9057379\tbest: 0.9056034 (895)\ttotal: 6m 18s\tremaining: 11h 29m 28s\n",
      "907:\tlearn: 0.3388095\ttest: 0.9058006\tbest: 0.9056034 (895)\ttotal: 6m 19s\tremaining: 11h 29m 27s\n",
      "908:\tlearn: 0.3386628\ttest: 0.9057089\tbest: 0.9056034 (895)\ttotal: 6m 19s\tremaining: 11h 29m 27s\n",
      "909:\tlearn: 0.3385028\ttest: 0.9058202\tbest: 0.9056034 (895)\ttotal: 6m 19s\tremaining: 11h 29m 27s\n",
      "910:\tlearn: 0.3383080\ttest: 0.9057052\tbest: 0.9056034 (895)\ttotal: 6m 20s\tremaining: 11h 29m 26s\n",
      "911:\tlearn: 0.3381671\ttest: 0.9057564\tbest: 0.9056034 (895)\ttotal: 6m 20s\tremaining: 11h 29m 26s\n",
      "912:\tlearn: 0.3380184\ttest: 0.9057975\tbest: 0.9056034 (895)\ttotal: 6m 21s\tremaining: 11h 29m 26s\n",
      "913:\tlearn: 0.3378757\ttest: 0.9057042\tbest: 0.9056034 (895)\ttotal: 6m 21s\tremaining: 11h 29m 25s\n",
      "914:\tlearn: 0.3376855\ttest: 0.9056285\tbest: 0.9056034 (895)\ttotal: 6m 21s\tremaining: 11h 29m 24s\n",
      "915:\tlearn: 0.3375601\ttest: 0.9057949\tbest: 0.9056034 (895)\ttotal: 6m 22s\tremaining: 11h 29m 23s\n",
      "916:\tlearn: 0.3374167\ttest: 0.9056449\tbest: 0.9056034 (895)\ttotal: 6m 22s\tremaining: 11h 29m 23s\n",
      "917:\tlearn: 0.3372637\ttest: 0.9056868\tbest: 0.9056034 (895)\ttotal: 6m 23s\tremaining: 11h 29m 23s\n",
      "918:\tlearn: 0.3371209\ttest: 0.9056964\tbest: 0.9056034 (895)\ttotal: 6m 23s\tremaining: 11h 29m 22s\n",
      "919:\tlearn: 0.3369301\ttest: 0.9058329\tbest: 0.9056034 (895)\ttotal: 6m 24s\tremaining: 11h 29m 22s\n",
      "920:\tlearn: 0.3367700\ttest: 0.9057700\tbest: 0.9056034 (895)\ttotal: 6m 24s\tremaining: 11h 29m 21s\n",
      "921:\tlearn: 0.3366431\ttest: 0.9057568\tbest: 0.9056034 (895)\ttotal: 6m 24s\tremaining: 11h 29m 20s\n",
      "922:\tlearn: 0.3365035\ttest: 0.9057906\tbest: 0.9056034 (895)\ttotal: 6m 25s\tremaining: 11h 29m 20s\n",
      "923:\tlearn: 0.3363651\ttest: 0.9057327\tbest: 0.9056034 (895)\ttotal: 6m 25s\tremaining: 11h 29m 19s\n",
      "924:\tlearn: 0.3362305\ttest: 0.9057470\tbest: 0.9056034 (895)\ttotal: 6m 26s\tremaining: 11h 29m 20s\n",
      "925:\tlearn: 0.3360646\ttest: 0.9057670\tbest: 0.9056034 (895)\ttotal: 6m 26s\tremaining: 11h 29m 20s\n",
      "926:\tlearn: 0.3358860\ttest: 0.9057693\tbest: 0.9056034 (895)\ttotal: 6m 26s\tremaining: 11h 29m 19s\n",
      "927:\tlearn: 0.3357472\ttest: 0.9057789\tbest: 0.9056034 (895)\ttotal: 6m 27s\tremaining: 11h 29m 18s\n",
      "928:\tlearn: 0.3356065\ttest: 0.9055926\tbest: 0.9055926 (928)\ttotal: 6m 27s\tremaining: 11h 29m 17s\n",
      "929:\tlearn: 0.3354550\ttest: 0.9055819\tbest: 0.9055819 (929)\ttotal: 6m 28s\tremaining: 11h 29m 16s\n",
      "930:\tlearn: 0.3353228\ttest: 0.9054303\tbest: 0.9054303 (930)\ttotal: 6m 28s\tremaining: 11h 29m 15s\n",
      "931:\tlearn: 0.3351305\ttest: 0.9054438\tbest: 0.9054303 (930)\ttotal: 6m 29s\tremaining: 11h 29m 14s\n",
      "932:\tlearn: 0.3349742\ttest: 0.9053742\tbest: 0.9053742 (932)\ttotal: 6m 29s\tremaining: 11h 29m 13s\n",
      "933:\tlearn: 0.3348274\ttest: 0.9053637\tbest: 0.9053637 (933)\ttotal: 6m 29s\tremaining: 11h 29m 13s\n",
      "934:\tlearn: 0.3346718\ttest: 0.9053225\tbest: 0.9053225 (934)\ttotal: 6m 30s\tremaining: 11h 29m 13s\n",
      "935:\tlearn: 0.3345475\ttest: 0.9054131\tbest: 0.9053225 (934)\ttotal: 6m 30s\tremaining: 11h 29m 15s\n",
      "936:\tlearn: 0.3343870\ttest: 0.9053377\tbest: 0.9053225 (934)\ttotal: 6m 31s\tremaining: 11h 29m 13s\n",
      "937:\tlearn: 0.3342321\ttest: 0.9052273\tbest: 0.9052273 (937)\ttotal: 6m 31s\tremaining: 11h 29m 11s\n",
      "938:\tlearn: 0.3340911\ttest: 0.9052788\tbest: 0.9052273 (937)\ttotal: 6m 31s\tremaining: 11h 29m 10s\n",
      "939:\tlearn: 0.3339486\ttest: 0.9053614\tbest: 0.9052273 (937)\ttotal: 6m 32s\tremaining: 11h 29m 8s\n",
      "940:\tlearn: 0.3337995\ttest: 0.9051654\tbest: 0.9051654 (940)\ttotal: 6m 32s\tremaining: 11h 29m 7s\n",
      "941:\tlearn: 0.3336485\ttest: 0.9053121\tbest: 0.9051654 (940)\ttotal: 6m 33s\tremaining: 11h 29m 6s\n",
      "942:\tlearn: 0.3335070\ttest: 0.9053239\tbest: 0.9051654 (940)\ttotal: 6m 33s\tremaining: 11h 29m 6s\n",
      "943:\tlearn: 0.3333581\ttest: 0.9052931\tbest: 0.9051654 (940)\ttotal: 6m 34s\tremaining: 11h 29m 6s\n",
      "944:\tlearn: 0.3332037\ttest: 0.9054173\tbest: 0.9051654 (940)\ttotal: 6m 34s\tremaining: 11h 29m 4s\n",
      "945:\tlearn: 0.3330021\ttest: 0.9055291\tbest: 0.9051654 (940)\ttotal: 6m 34s\tremaining: 11h 29m 3s\n",
      "946:\tlearn: 0.3328674\ttest: 0.9055255\tbest: 0.9051654 (940)\ttotal: 6m 35s\tremaining: 11h 29m 2s\n",
      "947:\tlearn: 0.3327247\ttest: 0.9056422\tbest: 0.9051654 (940)\ttotal: 6m 35s\tremaining: 11h 29m 1s\n",
      "948:\tlearn: 0.3325839\ttest: 0.9057108\tbest: 0.9051654 (940)\ttotal: 6m 36s\tremaining: 11h 29m\n",
      "949:\tlearn: 0.3324089\ttest: 0.9057576\tbest: 0.9051654 (940)\ttotal: 6m 36s\tremaining: 11h 28m 59s\n",
      "950:\tlearn: 0.3322402\ttest: 0.9056845\tbest: 0.9051654 (940)\ttotal: 6m 36s\tremaining: 11h 28m 57s\n",
      "951:\tlearn: 0.3320910\ttest: 0.9058011\tbest: 0.9051654 (940)\ttotal: 6m 37s\tremaining: 11h 28m 57s\n",
      "952:\tlearn: 0.3319435\ttest: 0.9057585\tbest: 0.9051654 (940)\ttotal: 6m 37s\tremaining: 11h 28m 56s\n",
      "953:\tlearn: 0.3317762\ttest: 0.9058009\tbest: 0.9051654 (940)\ttotal: 6m 38s\tremaining: 11h 28m 55s\n",
      "954:\tlearn: 0.3316224\ttest: 0.9059038\tbest: 0.9051654 (940)\ttotal: 6m 38s\tremaining: 11h 28m 53s\n",
      "955:\tlearn: 0.3314707\ttest: 0.9058713\tbest: 0.9051654 (940)\ttotal: 6m 38s\tremaining: 11h 28m 53s\n",
      "956:\tlearn: 0.3312770\ttest: 0.9059083\tbest: 0.9051654 (940)\ttotal: 6m 39s\tremaining: 11h 28m 51s\n",
      "957:\tlearn: 0.3311354\ttest: 0.9059415\tbest: 0.9051654 (940)\ttotal: 6m 39s\tremaining: 11h 28m 50s\n",
      "958:\tlearn: 0.3309699\ttest: 0.9057474\tbest: 0.9051654 (940)\ttotal: 6m 40s\tremaining: 11h 28m 49s\n",
      "959:\tlearn: 0.3308117\ttest: 0.9057765\tbest: 0.9051654 (940)\ttotal: 6m 40s\tremaining: 11h 28m 48s\n",
      "960:\tlearn: 0.3306874\ttest: 0.9056937\tbest: 0.9051654 (940)\ttotal: 6m 41s\tremaining: 11h 28m 46s\n",
      "961:\tlearn: 0.3305045\ttest: 0.9056910\tbest: 0.9051654 (940)\ttotal: 6m 41s\tremaining: 11h 28m 44s\n",
      "962:\tlearn: 0.3303830\ttest: 0.9057316\tbest: 0.9051654 (940)\ttotal: 6m 41s\tremaining: 11h 28m 43s\n",
      "963:\tlearn: 0.3302563\ttest: 0.9058822\tbest: 0.9051654 (940)\ttotal: 6m 42s\tremaining: 11h 28m 42s\n",
      "964:\tlearn: 0.3301430\ttest: 0.9059546\tbest: 0.9051654 (940)\ttotal: 6m 42s\tremaining: 11h 28m 40s\n",
      "965:\tlearn: 0.3300191\ttest: 0.9058444\tbest: 0.9051654 (940)\ttotal: 6m 43s\tremaining: 11h 28m 39s\n",
      "966:\tlearn: 0.3298934\ttest: 0.9058163\tbest: 0.9051654 (940)\ttotal: 6m 43s\tremaining: 11h 28m 38s\n",
      "967:\tlearn: 0.3297349\ttest: 0.9058319\tbest: 0.9051654 (940)\ttotal: 6m 43s\tremaining: 11h 28m 39s\n",
      "968:\tlearn: 0.3295906\ttest: 0.9057420\tbest: 0.9051654 (940)\ttotal: 6m 44s\tremaining: 11h 28m 38s\n",
      "969:\tlearn: 0.3294626\ttest: 0.9057913\tbest: 0.9051654 (940)\ttotal: 6m 44s\tremaining: 11h 28m 37s\n",
      "970:\tlearn: 0.3293110\ttest: 0.9058359\tbest: 0.9051654 (940)\ttotal: 6m 45s\tremaining: 11h 28m 37s\n",
      "971:\tlearn: 0.3291797\ttest: 0.9057398\tbest: 0.9051654 (940)\ttotal: 6m 45s\tremaining: 11h 28m 36s\n",
      "972:\tlearn: 0.3290355\ttest: 0.9057766\tbest: 0.9051654 (940)\ttotal: 6m 45s\tremaining: 11h 28m 36s\n",
      "973:\tlearn: 0.3288638\ttest: 0.9058143\tbest: 0.9051654 (940)\ttotal: 6m 46s\tremaining: 11h 28m 35s\n",
      "974:\tlearn: 0.3286953\ttest: 0.9058839\tbest: 0.9051654 (940)\ttotal: 6m 46s\tremaining: 11h 28m 35s\n",
      "975:\tlearn: 0.3285438\ttest: 0.9058238\tbest: 0.9051654 (940)\ttotal: 6m 47s\tremaining: 11h 28m 35s\n",
      "976:\tlearn: 0.3283703\ttest: 0.9060240\tbest: 0.9051654 (940)\ttotal: 6m 47s\tremaining: 11h 28m 33s\n",
      "977:\tlearn: 0.3282652\ttest: 0.9059958\tbest: 0.9051654 (940)\ttotal: 6m 48s\tremaining: 11h 28m 31s\n",
      "978:\tlearn: 0.3281338\ttest: 0.9060752\tbest: 0.9051654 (940)\ttotal: 6m 48s\tremaining: 11h 28m 32s\n",
      "979:\tlearn: 0.3280008\ttest: 0.9060453\tbest: 0.9051654 (940)\ttotal: 6m 48s\tremaining: 11h 28m 32s\n",
      "980:\tlearn: 0.3278490\ttest: 0.9060693\tbest: 0.9051654 (940)\ttotal: 6m 49s\tremaining: 11h 28m 32s\n",
      "981:\tlearn: 0.3277074\ttest: 0.9060579\tbest: 0.9051654 (940)\ttotal: 6m 49s\tremaining: 11h 28m 32s\n",
      "982:\tlearn: 0.3275615\ttest: 0.9060462\tbest: 0.9051654 (940)\ttotal: 6m 50s\tremaining: 11h 28m 34s\n",
      "983:\tlearn: 0.3274384\ttest: 0.9060101\tbest: 0.9051654 (940)\ttotal: 6m 50s\tremaining: 11h 28m 34s\n",
      "984:\tlearn: 0.3272886\ttest: 0.9059524\tbest: 0.9051654 (940)\ttotal: 6m 51s\tremaining: 11h 28m 35s\n",
      "985:\tlearn: 0.3271259\ttest: 0.9059707\tbest: 0.9051654 (940)\ttotal: 6m 51s\tremaining: 11h 28m 35s\n",
      "986:\tlearn: 0.3269646\ttest: 0.9059898\tbest: 0.9051654 (940)\ttotal: 6m 51s\tremaining: 11h 28m 36s\n",
      "987:\tlearn: 0.3268278\ttest: 0.9059704\tbest: 0.9051654 (940)\ttotal: 6m 52s\tremaining: 11h 28m 36s\n",
      "988:\tlearn: 0.3266665\ttest: 0.9059795\tbest: 0.9051654 (940)\ttotal: 6m 52s\tremaining: 11h 28m 34s\n",
      "989:\tlearn: 0.3264984\ttest: 0.9060368\tbest: 0.9051654 (940)\ttotal: 6m 53s\tremaining: 11h 28m 34s\n",
      "990:\tlearn: 0.3263789\ttest: 0.9060247\tbest: 0.9051654 (940)\ttotal: 6m 53s\tremaining: 11h 28m 34s\n",
      "991:\tlearn: 0.3262174\ttest: 0.9060587\tbest: 0.9051654 (940)\ttotal: 6m 53s\tremaining: 11h 28m 36s\n",
      "992:\tlearn: 0.3260655\ttest: 0.9060694\tbest: 0.9051654 (940)\ttotal: 6m 54s\tremaining: 11h 28m 35s\n",
      "993:\tlearn: 0.3259384\ttest: 0.9060459\tbest: 0.9051654 (940)\ttotal: 6m 54s\tremaining: 11h 28m 35s\n",
      "994:\tlearn: 0.3258190\ttest: 0.9060594\tbest: 0.9051654 (940)\ttotal: 6m 55s\tremaining: 11h 28m 34s\n",
      "995:\tlearn: 0.3256877\ttest: 0.9060375\tbest: 0.9051654 (940)\ttotal: 6m 55s\tremaining: 11h 28m 33s\n",
      "996:\tlearn: 0.3255447\ttest: 0.9060347\tbest: 0.9051654 (940)\ttotal: 6m 56s\tremaining: 11h 28m 34s\n",
      "997:\tlearn: 0.3254030\ttest: 0.9059182\tbest: 0.9051654 (940)\ttotal: 6m 56s\tremaining: 11h 28m 33s\n",
      "998:\tlearn: 0.3252821\ttest: 0.9058660\tbest: 0.9051654 (940)\ttotal: 6m 56s\tremaining: 11h 28m 33s\n",
      "999:\tlearn: 0.3250929\ttest: 0.9056508\tbest: 0.9051654 (940)\ttotal: 6m 57s\tremaining: 11h 28m 32s\n",
      "1000:\tlearn: 0.3249628\ttest: 0.9057217\tbest: 0.9051654 (940)\ttotal: 6m 57s\tremaining: 11h 28m 30s\n",
      "1001:\tlearn: 0.3248483\ttest: 0.9057461\tbest: 0.9051654 (940)\ttotal: 6m 58s\tremaining: 11h 28m 29s\n",
      "1002:\tlearn: 0.3246858\ttest: 0.9057913\tbest: 0.9051654 (940)\ttotal: 6m 58s\tremaining: 11h 28m 28s\n",
      "1003:\tlearn: 0.3245427\ttest: 0.9057339\tbest: 0.9051654 (940)\ttotal: 6m 58s\tremaining: 11h 28m 28s\n",
      "1004:\tlearn: 0.3244137\ttest: 0.9056803\tbest: 0.9051654 (940)\ttotal: 6m 59s\tremaining: 11h 28m 27s\n",
      "1005:\tlearn: 0.3242674\ttest: 0.9056736\tbest: 0.9051654 (940)\ttotal: 6m 59s\tremaining: 11h 28m 27s\n",
      "1006:\tlearn: 0.3241471\ttest: 0.9055958\tbest: 0.9051654 (940)\ttotal: 7m\tremaining: 11h 28m 27s\n",
      "1007:\tlearn: 0.3240025\ttest: 0.9055126\tbest: 0.9051654 (940)\ttotal: 7m\tremaining: 11h 28m 26s\n",
      "1008:\tlearn: 0.3238453\ttest: 0.9056277\tbest: 0.9051654 (940)\ttotal: 7m 1s\tremaining: 11h 28m 28s\n",
      "1009:\tlearn: 0.3237177\ttest: 0.9055764\tbest: 0.9051654 (940)\ttotal: 7m 1s\tremaining: 11h 28m 29s\n",
      "1010:\tlearn: 0.3235708\ttest: 0.9054824\tbest: 0.9051654 (940)\ttotal: 7m 1s\tremaining: 11h 28m 31s\n",
      "1011:\tlearn: 0.3234397\ttest: 0.9055727\tbest: 0.9051654 (940)\ttotal: 7m 2s\tremaining: 11h 28m 32s\n",
      "1012:\tlearn: 0.3233244\ttest: 0.9057164\tbest: 0.9051654 (940)\ttotal: 7m 2s\tremaining: 11h 28m 30s\n",
      "1013:\tlearn: 0.3232004\ttest: 0.9057918\tbest: 0.9051654 (940)\ttotal: 7m 3s\tremaining: 11h 28m 29s\n",
      "1014:\tlearn: 0.3230748\ttest: 0.9058785\tbest: 0.9051654 (940)\ttotal: 7m 3s\tremaining: 11h 28m 27s\n",
      "1015:\tlearn: 0.3229586\ttest: 0.9060353\tbest: 0.9051654 (940)\ttotal: 7m 3s\tremaining: 11h 28m 28s\n",
      "1016:\tlearn: 0.3228368\ttest: 0.9059831\tbest: 0.9051654 (940)\ttotal: 7m 4s\tremaining: 11h 28m 29s\n",
      "1017:\tlearn: 0.3227185\ttest: 0.9059344\tbest: 0.9051654 (940)\ttotal: 7m 4s\tremaining: 11h 28m 34s\n",
      "1018:\tlearn: 0.3225992\ttest: 0.9060071\tbest: 0.9051654 (940)\ttotal: 7m 5s\tremaining: 11h 28m 35s\n",
      "1019:\tlearn: 0.3224891\ttest: 0.9059864\tbest: 0.9051654 (940)\ttotal: 7m 5s\tremaining: 11h 28m 36s\n",
      "1020:\tlearn: 0.3223675\ttest: 0.9060138\tbest: 0.9051654 (940)\ttotal: 7m 6s\tremaining: 11h 28m 37s\n",
      "1021:\tlearn: 0.3222168\ttest: 0.9060549\tbest: 0.9051654 (940)\ttotal: 7m 6s\tremaining: 11h 28m 38s\n",
      "1022:\tlearn: 0.3220580\ttest: 0.9061113\tbest: 0.9051654 (940)\ttotal: 7m 7s\tremaining: 11h 28m 41s\n",
      "1023:\tlearn: 0.3219127\ttest: 0.9062680\tbest: 0.9051654 (940)\ttotal: 7m 7s\tremaining: 11h 28m 41s\n",
      "1024:\tlearn: 0.3217881\ttest: 0.9062044\tbest: 0.9051654 (940)\ttotal: 7m 7s\tremaining: 11h 28m 39s\n",
      "1025:\tlearn: 0.3216303\ttest: 0.9062695\tbest: 0.9051654 (940)\ttotal: 7m 8s\tremaining: 11h 28m 38s\n",
      "1026:\tlearn: 0.3215041\ttest: 0.9062720\tbest: 0.9051654 (940)\ttotal: 7m 8s\tremaining: 11h 28m 37s\n",
      "1027:\tlearn: 0.3213630\ttest: 0.9063209\tbest: 0.9051654 (940)\ttotal: 7m 9s\tremaining: 11h 28m 39s\n",
      "1028:\tlearn: 0.3212179\ttest: 0.9064208\tbest: 0.9051654 (940)\ttotal: 7m 9s\tremaining: 11h 28m 39s\n",
      "1029:\tlearn: 0.3210849\ttest: 0.9064191\tbest: 0.9051654 (940)\ttotal: 7m 10s\tremaining: 11h 28m 40s\n",
      "1030:\tlearn: 0.3209399\ttest: 0.9064858\tbest: 0.9051654 (940)\ttotal: 7m 10s\tremaining: 11h 28m 41s\n",
      "1031:\tlearn: 0.3207992\ttest: 0.9064040\tbest: 0.9051654 (940)\ttotal: 7m 10s\tremaining: 11h 28m 41s\n",
      "1032:\tlearn: 0.3206600\ttest: 0.9063591\tbest: 0.9051654 (940)\ttotal: 7m 11s\tremaining: 11h 28m 41s\n",
      "1033:\tlearn: 0.3205326\ttest: 0.9063529\tbest: 0.9051654 (940)\ttotal: 7m 11s\tremaining: 11h 28m 41s\n",
      "1034:\tlearn: 0.3204107\ttest: 0.9062797\tbest: 0.9051654 (940)\ttotal: 7m 12s\tremaining: 11h 28m 42s\n",
      "1035:\tlearn: 0.3202772\ttest: 0.9063304\tbest: 0.9051654 (940)\ttotal: 7m 12s\tremaining: 11h 28m 42s\n",
      "1036:\tlearn: 0.3201509\ttest: 0.9063088\tbest: 0.9051654 (940)\ttotal: 7m 13s\tremaining: 11h 28m 42s\n",
      "1037:\tlearn: 0.3200079\ttest: 0.9063234\tbest: 0.9051654 (940)\ttotal: 7m 13s\tremaining: 11h 28m 40s\n",
      "1038:\tlearn: 0.3198704\ttest: 0.9062976\tbest: 0.9051654 (940)\ttotal: 7m 13s\tremaining: 11h 28m 40s\n",
      "1039:\tlearn: 0.3197303\ttest: 0.9065251\tbest: 0.9051654 (940)\ttotal: 7m 14s\tremaining: 11h 28m 40s\n",
      "1040:\tlearn: 0.3195990\ttest: 0.9065198\tbest: 0.9051654 (940)\ttotal: 7m 14s\tremaining: 11h 28m 40s\n",
      "1041:\tlearn: 0.3194592\ttest: 0.9065371\tbest: 0.9051654 (940)\ttotal: 7m 15s\tremaining: 11h 28m 39s\n",
      "1042:\tlearn: 0.3193319\ttest: 0.9065700\tbest: 0.9051654 (940)\ttotal: 7m 15s\tremaining: 11h 28m 39s\n",
      "1043:\tlearn: 0.3192194\ttest: 0.9066534\tbest: 0.9051654 (940)\ttotal: 7m 15s\tremaining: 11h 28m 40s\n",
      "1044:\tlearn: 0.3191069\ttest: 0.9066493\tbest: 0.9051654 (940)\ttotal: 7m 16s\tremaining: 11h 28m 41s\n",
      "1045:\tlearn: 0.3189149\ttest: 0.9064850\tbest: 0.9051654 (940)\ttotal: 7m 16s\tremaining: 11h 28m 42s\n",
      "1046:\tlearn: 0.3187908\ttest: 0.9065399\tbest: 0.9051654 (940)\ttotal: 7m 17s\tremaining: 11h 28m 42s\n",
      "1047:\tlearn: 0.3186181\ttest: 0.9065348\tbest: 0.9051654 (940)\ttotal: 7m 17s\tremaining: 11h 28m 41s\n",
      "1048:\tlearn: 0.3184736\ttest: 0.9065449\tbest: 0.9051654 (940)\ttotal: 7m 18s\tremaining: 11h 28m 41s\n",
      "1049:\tlearn: 0.3183608\ttest: 0.9065375\tbest: 0.9051654 (940)\ttotal: 7m 18s\tremaining: 11h 28m 39s\n",
      "1050:\tlearn: 0.3182378\ttest: 0.9065604\tbest: 0.9051654 (940)\ttotal: 7m 18s\tremaining: 11h 28m 40s\n",
      "1051:\tlearn: 0.3180892\ttest: 0.9067530\tbest: 0.9051654 (940)\ttotal: 7m 19s\tremaining: 11h 28m 41s\n",
      "1052:\tlearn: 0.3179151\ttest: 0.9067114\tbest: 0.9051654 (940)\ttotal: 7m 19s\tremaining: 11h 28m 42s\n",
      "1053:\tlearn: 0.3177758\ttest: 0.9066070\tbest: 0.9051654 (940)\ttotal: 7m 20s\tremaining: 11h 28m 42s\n",
      "1054:\tlearn: 0.3176365\ttest: 0.9066040\tbest: 0.9051654 (940)\ttotal: 7m 20s\tremaining: 11h 28m 42s\n",
      "1055:\tlearn: 0.3175051\ttest: 0.9067121\tbest: 0.9051654 (940)\ttotal: 7m 21s\tremaining: 11h 28m 41s\n",
      "1056:\tlearn: 0.3173832\ttest: 0.9066825\tbest: 0.9051654 (940)\ttotal: 7m 21s\tremaining: 11h 28m 41s\n",
      "1057:\tlearn: 0.3172399\ttest: 0.9066938\tbest: 0.9051654 (940)\ttotal: 7m 21s\tremaining: 11h 28m 42s\n",
      "1058:\tlearn: 0.3171246\ttest: 0.9067521\tbest: 0.9051654 (940)\ttotal: 7m 22s\tremaining: 11h 28m 43s\n",
      "1059:\tlearn: 0.3169931\ttest: 0.9067460\tbest: 0.9051654 (940)\ttotal: 7m 22s\tremaining: 11h 28m 43s\n",
      "1060:\tlearn: 0.3168751\ttest: 0.9067420\tbest: 0.9051654 (940)\ttotal: 7m 23s\tremaining: 11h 28m 42s\n",
      "1061:\tlearn: 0.3167374\ttest: 0.9067051\tbest: 0.9051654 (940)\ttotal: 7m 23s\tremaining: 11h 28m 41s\n",
      "1062:\tlearn: 0.3166111\ttest: 0.9066398\tbest: 0.9051654 (940)\ttotal: 7m 23s\tremaining: 11h 28m 41s\n",
      "1063:\tlearn: 0.3164505\ttest: 0.9066500\tbest: 0.9051654 (940)\ttotal: 7m 24s\tremaining: 11h 28m 41s\n",
      "1064:\tlearn: 0.3163196\ttest: 0.9066705\tbest: 0.9051654 (940)\ttotal: 7m 24s\tremaining: 11h 28m 40s\n",
      "1065:\tlearn: 0.3162252\ttest: 0.9067238\tbest: 0.9051654 (940)\ttotal: 7m 25s\tremaining: 11h 28m 44s\n",
      "1066:\tlearn: 0.3160497\ttest: 0.9066372\tbest: 0.9051654 (940)\ttotal: 7m 25s\tremaining: 11h 28m 43s\n",
      "1067:\tlearn: 0.3159278\ttest: 0.9065997\tbest: 0.9051654 (940)\ttotal: 7m 26s\tremaining: 11h 28m 43s\n",
      "1068:\tlearn: 0.3158094\ttest: 0.9065258\tbest: 0.9051654 (940)\ttotal: 7m 26s\tremaining: 11h 28m 44s\n",
      "1069:\tlearn: 0.3157081\ttest: 0.9065351\tbest: 0.9051654 (940)\ttotal: 7m 26s\tremaining: 11h 28m 46s\n",
      "1070:\tlearn: 0.3155819\ttest: 0.9065493\tbest: 0.9051654 (940)\ttotal: 7m 27s\tremaining: 11h 28m 46s\n",
      "1071:\tlearn: 0.3154650\ttest: 0.9066882\tbest: 0.9051654 (940)\ttotal: 7m 27s\tremaining: 11h 28m 46s\n",
      "1072:\tlearn: 0.3153670\ttest: 0.9067767\tbest: 0.9051654 (940)\ttotal: 7m 28s\tremaining: 11h 28m 46s\n",
      "1073:\tlearn: 0.3152390\ttest: 0.9068224\tbest: 0.9051654 (940)\ttotal: 7m 28s\tremaining: 11h 28m 45s\n",
      "1074:\tlearn: 0.3151040\ttest: 0.9068209\tbest: 0.9051654 (940)\ttotal: 7m 29s\tremaining: 11h 28m 46s\n",
      "1075:\tlearn: 0.3149808\ttest: 0.9068110\tbest: 0.9051654 (940)\ttotal: 7m 29s\tremaining: 11h 28m 45s\n",
      "1076:\tlearn: 0.3148667\ttest: 0.9068369\tbest: 0.9051654 (940)\ttotal: 7m 29s\tremaining: 11h 28m 48s\n",
      "1077:\tlearn: 0.3147460\ttest: 0.9067944\tbest: 0.9051654 (940)\ttotal: 7m 30s\tremaining: 11h 28m 49s\n",
      "1078:\tlearn: 0.3146058\ttest: 0.9068110\tbest: 0.9051654 (940)\ttotal: 7m 30s\tremaining: 11h 28m 48s\n",
      "1079:\tlearn: 0.3144968\ttest: 0.9068723\tbest: 0.9051654 (940)\ttotal: 7m 31s\tremaining: 11h 28m 48s\n",
      "1080:\tlearn: 0.3143773\ttest: 0.9067910\tbest: 0.9051654 (940)\ttotal: 7m 31s\tremaining: 11h 28m 48s\n",
      "1081:\tlearn: 0.3142348\ttest: 0.9068881\tbest: 0.9051654 (940)\ttotal: 7m 32s\tremaining: 11h 28m 48s\n",
      "1082:\tlearn: 0.3141142\ttest: 0.9069790\tbest: 0.9051654 (940)\ttotal: 7m 32s\tremaining: 11h 28m 47s\n",
      "1083:\tlearn: 0.3140066\ttest: 0.9070834\tbest: 0.9051654 (940)\ttotal: 7m 32s\tremaining: 11h 28m 47s\n",
      "1084:\tlearn: 0.3138692\ttest: 0.9071114\tbest: 0.9051654 (940)\ttotal: 7m 33s\tremaining: 11h 28m 45s\n",
      "1085:\tlearn: 0.3137649\ttest: 0.9071410\tbest: 0.9051654 (940)\ttotal: 7m 33s\tremaining: 11h 28m 44s\n",
      "1086:\tlearn: 0.3136291\ttest: 0.9071978\tbest: 0.9051654 (940)\ttotal: 7m 34s\tremaining: 11h 28m 44s\n",
      "1087:\tlearn: 0.3135221\ttest: 0.9071391\tbest: 0.9051654 (940)\ttotal: 7m 34s\tremaining: 11h 28m 43s\n",
      "1088:\tlearn: 0.3134102\ttest: 0.9071149\tbest: 0.9051654 (940)\ttotal: 7m 34s\tremaining: 11h 28m 44s\n",
      "1089:\tlearn: 0.3132783\ttest: 0.9070704\tbest: 0.9051654 (940)\ttotal: 7m 35s\tremaining: 11h 28m 44s\n",
      "1090:\tlearn: 0.3131059\ttest: 0.9070978\tbest: 0.9051654 (940)\ttotal: 7m 35s\tremaining: 11h 28m 45s\n",
      "1091:\tlearn: 0.3129856\ttest: 0.9070255\tbest: 0.9051654 (940)\ttotal: 7m 36s\tremaining: 11h 28m 45s\n",
      "1092:\tlearn: 0.3128636\ttest: 0.9069889\tbest: 0.9051654 (940)\ttotal: 7m 36s\tremaining: 11h 28m 45s\n",
      "1093:\tlearn: 0.3127230\ttest: 0.9069899\tbest: 0.9051654 (940)\ttotal: 7m 37s\tremaining: 11h 28m 46s\n",
      "1094:\tlearn: 0.3126054\ttest: 0.9070440\tbest: 0.9051654 (940)\ttotal: 7m 37s\tremaining: 11h 28m 46s\n",
      "1095:\tlearn: 0.3124743\ttest: 0.9070968\tbest: 0.9051654 (940)\ttotal: 7m 37s\tremaining: 11h 28m 45s\n",
      "1096:\tlearn: 0.3123801\ttest: 0.9071859\tbest: 0.9051654 (940)\ttotal: 7m 38s\tremaining: 11h 28m 45s\n",
      "1097:\tlearn: 0.3122597\ttest: 0.9071197\tbest: 0.9051654 (940)\ttotal: 7m 38s\tremaining: 11h 28m 44s\n",
      "1098:\tlearn: 0.3121222\ttest: 0.9070809\tbest: 0.9051654 (940)\ttotal: 7m 39s\tremaining: 11h 28m 43s\n",
      "1099:\tlearn: 0.3120120\ttest: 0.9071342\tbest: 0.9051654 (940)\ttotal: 7m 39s\tremaining: 11h 28m 43s\n",
      "1100:\tlearn: 0.3118923\ttest: 0.9070297\tbest: 0.9051654 (940)\ttotal: 7m 40s\tremaining: 11h 28m 47s\n",
      "1101:\tlearn: 0.3117543\ttest: 0.9070854\tbest: 0.9051654 (940)\ttotal: 7m 40s\tremaining: 11h 28m 48s\n",
      "1102:\tlearn: 0.3116359\ttest: 0.9070230\tbest: 0.9051654 (940)\ttotal: 7m 40s\tremaining: 11h 28m 49s\n",
      "1103:\tlearn: 0.3115261\ttest: 0.9070838\tbest: 0.9051654 (940)\ttotal: 7m 41s\tremaining: 11h 28m 49s\n",
      "1104:\tlearn: 0.3113949\ttest: 0.9071470\tbest: 0.9051654 (940)\ttotal: 7m 41s\tremaining: 11h 28m 49s\n",
      "1105:\tlearn: 0.3112405\ttest: 0.9071080\tbest: 0.9051654 (940)\ttotal: 7m 42s\tremaining: 11h 28m 48s\n",
      "1106:\tlearn: 0.3111187\ttest: 0.9070216\tbest: 0.9051654 (940)\ttotal: 7m 42s\tremaining: 11h 28m 49s\n",
      "1107:\tlearn: 0.3110374\ttest: 0.9069278\tbest: 0.9051654 (940)\ttotal: 7m 43s\tremaining: 11h 28m 49s\n",
      "1108:\tlearn: 0.3109150\ttest: 0.9068763\tbest: 0.9051654 (940)\ttotal: 7m 43s\tremaining: 11h 28m 48s\n",
      "1109:\tlearn: 0.3107864\ttest: 0.9069370\tbest: 0.9051654 (940)\ttotal: 7m 43s\tremaining: 11h 28m 47s\n",
      "1110:\tlearn: 0.3106568\ttest: 0.9069879\tbest: 0.9051654 (940)\ttotal: 7m 44s\tremaining: 11h 28m 46s\n",
      "1111:\tlearn: 0.3105444\ttest: 0.9069134\tbest: 0.9051654 (940)\ttotal: 7m 44s\tremaining: 11h 28m 46s\n",
      "1112:\tlearn: 0.3104251\ttest: 0.9069224\tbest: 0.9051654 (940)\ttotal: 7m 45s\tremaining: 11h 28m 46s\n",
      "1113:\tlearn: 0.3103037\ttest: 0.9069648\tbest: 0.9051654 (940)\ttotal: 7m 45s\tremaining: 11h 28m 46s\n",
      "1114:\tlearn: 0.3101878\ttest: 0.9069601\tbest: 0.9051654 (940)\ttotal: 7m 45s\tremaining: 11h 28m 45s\n",
      "1115:\tlearn: 0.3100748\ttest: 0.9069303\tbest: 0.9051654 (940)\ttotal: 7m 46s\tremaining: 11h 28m 44s\n",
      "1116:\tlearn: 0.3099560\ttest: 0.9069371\tbest: 0.9051654 (940)\ttotal: 7m 46s\tremaining: 11h 28m 44s\n",
      "1117:\tlearn: 0.3098170\ttest: 0.9069327\tbest: 0.9051654 (940)\ttotal: 7m 47s\tremaining: 11h 28m 45s\n",
      "1118:\tlearn: 0.3097153\ttest: 0.9069605\tbest: 0.9051654 (940)\ttotal: 7m 47s\tremaining: 11h 28m 44s\n",
      "1119:\tlearn: 0.3096184\ttest: 0.9069175\tbest: 0.9051654 (940)\ttotal: 7m 48s\tremaining: 11h 28m 44s\n",
      "1120:\tlearn: 0.3094693\ttest: 0.9069500\tbest: 0.9051654 (940)\ttotal: 7m 48s\tremaining: 11h 28m 43s\n",
      "1121:\tlearn: 0.3093338\ttest: 0.9068105\tbest: 0.9051654 (940)\ttotal: 7m 48s\tremaining: 11h 28m 42s\n",
      "1122:\tlearn: 0.3092084\ttest: 0.9067031\tbest: 0.9051654 (940)\ttotal: 7m 49s\tremaining: 11h 28m 41s\n",
      "1123:\tlearn: 0.3091087\ttest: 0.9068040\tbest: 0.9051654 (940)\ttotal: 7m 49s\tremaining: 11h 28m 41s\n",
      "1124:\tlearn: 0.3090012\ttest: 0.9068106\tbest: 0.9051654 (940)\ttotal: 7m 50s\tremaining: 11h 28m 41s\n",
      "1125:\tlearn: 0.3088723\ttest: 0.9067896\tbest: 0.9051654 (940)\ttotal: 7m 50s\tremaining: 11h 28m 40s\n",
      "1126:\tlearn: 0.3087294\ttest: 0.9069506\tbest: 0.9051654 (940)\ttotal: 7m 51s\tremaining: 11h 28m 41s\n",
      "1127:\tlearn: 0.3085824\ttest: 0.9068772\tbest: 0.9051654 (940)\ttotal: 7m 51s\tremaining: 11h 28m 41s\n",
      "1128:\tlearn: 0.3084545\ttest: 0.9068634\tbest: 0.9051654 (940)\ttotal: 7m 51s\tremaining: 11h 28m 40s\n",
      "1129:\tlearn: 0.3083481\ttest: 0.9068640\tbest: 0.9051654 (940)\ttotal: 7m 52s\tremaining: 11h 28m 41s\n",
      "1130:\tlearn: 0.3082288\ttest: 0.9069745\tbest: 0.9051654 (940)\ttotal: 7m 52s\tremaining: 11h 28m 40s\n",
      "1131:\tlearn: 0.3080914\ttest: 0.9069138\tbest: 0.9051654 (940)\ttotal: 7m 53s\tremaining: 11h 28m 40s\n",
      "1132:\tlearn: 0.3079843\ttest: 0.9069053\tbest: 0.9051654 (940)\ttotal: 7m 53s\tremaining: 11h 28m 39s\n",
      "1133:\tlearn: 0.3078731\ttest: 0.9068562\tbest: 0.9051654 (940)\ttotal: 7m 53s\tremaining: 11h 28m 38s\n",
      "1134:\tlearn: 0.3077619\ttest: 0.9069351\tbest: 0.9051654 (940)\ttotal: 7m 54s\tremaining: 11h 28m 38s\n",
      "1135:\tlearn: 0.3076402\ttest: 0.9068003\tbest: 0.9051654 (940)\ttotal: 7m 54s\tremaining: 11h 28m 39s\n",
      "1136:\tlearn: 0.3075357\ttest: 0.9067935\tbest: 0.9051654 (940)\ttotal: 7m 55s\tremaining: 11h 28m 39s\n",
      "1137:\tlearn: 0.3073807\ttest: 0.9067364\tbest: 0.9051654 (940)\ttotal: 7m 55s\tremaining: 11h 28m 39s\n",
      "1138:\tlearn: 0.3072483\ttest: 0.9067139\tbest: 0.9051654 (940)\ttotal: 7m 56s\tremaining: 11h 28m 39s\n",
      "1139:\tlearn: 0.3070964\ttest: 0.9066117\tbest: 0.9051654 (940)\ttotal: 7m 56s\tremaining: 11h 28m 39s\n",
      "1140:\tlearn: 0.3070010\ttest: 0.9066796\tbest: 0.9051654 (940)\ttotal: 7m 56s\tremaining: 11h 28m 38s\n",
      "1141:\tlearn: 0.3068741\ttest: 0.9067184\tbest: 0.9051654 (940)\ttotal: 7m 57s\tremaining: 11h 28m 38s\n",
      "1142:\tlearn: 0.3067021\ttest: 0.9065984\tbest: 0.9051654 (940)\ttotal: 7m 57s\tremaining: 11h 28m 37s\n",
      "1143:\tlearn: 0.3065930\ttest: 0.9066877\tbest: 0.9051654 (940)\ttotal: 7m 58s\tremaining: 11h 28m 36s\n",
      "1144:\tlearn: 0.3065029\ttest: 0.9067348\tbest: 0.9051654 (940)\ttotal: 7m 58s\tremaining: 11h 28m 34s\n",
      "1145:\tlearn: 0.3063747\ttest: 0.9068244\tbest: 0.9051654 (940)\ttotal: 7m 58s\tremaining: 11h 28m 33s\n",
      "1146:\tlearn: 0.3062766\ttest: 0.9068323\tbest: 0.9051654 (940)\ttotal: 7m 59s\tremaining: 11h 28m 33s\n",
      "1147:\tlearn: 0.3061383\ttest: 0.9069187\tbest: 0.9051654 (940)\ttotal: 7m 59s\tremaining: 11h 28m 32s\n",
      "1148:\tlearn: 0.3060169\ttest: 0.9069170\tbest: 0.9051654 (940)\ttotal: 8m\tremaining: 11h 28m 32s\n",
      "1149:\tlearn: 0.3059124\ttest: 0.9069314\tbest: 0.9051654 (940)\ttotal: 8m\tremaining: 11h 28m 32s\n",
      "1150:\tlearn: 0.3057985\ttest: 0.9070175\tbest: 0.9051654 (940)\ttotal: 8m 1s\tremaining: 11h 28m 32s\n",
      "1151:\tlearn: 0.3056740\ttest: 0.9070826\tbest: 0.9051654 (940)\ttotal: 8m 1s\tremaining: 11h 28m 32s\n",
      "1152:\tlearn: 0.3055557\ttest: 0.9070198\tbest: 0.9051654 (940)\ttotal: 8m 1s\tremaining: 11h 28m 32s\n",
      "1153:\tlearn: 0.3054530\ttest: 0.9070184\tbest: 0.9051654 (940)\ttotal: 8m 2s\tremaining: 11h 28m 31s\n",
      "1154:\tlearn: 0.3053135\ttest: 0.9070529\tbest: 0.9051654 (940)\ttotal: 8m 2s\tremaining: 11h 28m 30s\n",
      "1155:\tlearn: 0.3052038\ttest: 0.9070609\tbest: 0.9051654 (940)\ttotal: 8m 3s\tremaining: 11h 28m 31s\n",
      "1156:\tlearn: 0.3051036\ttest: 0.9071173\tbest: 0.9051654 (940)\ttotal: 8m 3s\tremaining: 11h 28m 29s\n",
      "1157:\tlearn: 0.3049840\ttest: 0.9073849\tbest: 0.9051654 (940)\ttotal: 8m 3s\tremaining: 11h 28m 28s\n",
      "1158:\tlearn: 0.3048508\ttest: 0.9074896\tbest: 0.9051654 (940)\ttotal: 8m 4s\tremaining: 11h 28m 28s\n",
      "1159:\tlearn: 0.3047496\ttest: 0.9075376\tbest: 0.9051654 (940)\ttotal: 8m 4s\tremaining: 11h 28m 27s\n",
      "1160:\tlearn: 0.3046253\ttest: 0.9075551\tbest: 0.9051654 (940)\ttotal: 8m 5s\tremaining: 11h 28m 26s\n",
      "1161:\tlearn: 0.3045269\ttest: 0.9076491\tbest: 0.9051654 (940)\ttotal: 8m 5s\tremaining: 11h 28m 27s\n",
      "1162:\tlearn: 0.3044097\ttest: 0.9076181\tbest: 0.9051654 (940)\ttotal: 8m 6s\tremaining: 11h 28m 29s\n",
      "1163:\tlearn: 0.3042818\ttest: 0.9076402\tbest: 0.9051654 (940)\ttotal: 8m 6s\tremaining: 11h 28m 29s\n",
      "1164:\tlearn: 0.3041765\ttest: 0.9076073\tbest: 0.9051654 (940)\ttotal: 8m 6s\tremaining: 11h 28m 29s\n",
      "1165:\tlearn: 0.3040636\ttest: 0.9076081\tbest: 0.9051654 (940)\ttotal: 8m 7s\tremaining: 11h 28m 29s\n",
      "1166:\tlearn: 0.3039547\ttest: 0.9076631\tbest: 0.9051654 (940)\ttotal: 8m 7s\tremaining: 11h 28m 29s\n",
      "1167:\tlearn: 0.3038384\ttest: 0.9077311\tbest: 0.9051654 (940)\ttotal: 8m 8s\tremaining: 11h 28m 29s\n",
      "1168:\tlearn: 0.3037191\ttest: 0.9076632\tbest: 0.9051654 (940)\ttotal: 8m 8s\tremaining: 11h 28m 28s\n",
      "1169:\tlearn: 0.3036140\ttest: 0.9076721\tbest: 0.9051654 (940)\ttotal: 8m 9s\tremaining: 11h 28m 26s\n",
      "1170:\tlearn: 0.3035333\ttest: 0.9076719\tbest: 0.9051654 (940)\ttotal: 8m 9s\tremaining: 11h 28m 25s\n",
      "1171:\tlearn: 0.3034200\ttest: 0.9075644\tbest: 0.9051654 (940)\ttotal: 8m 9s\tremaining: 11h 28m 25s\n",
      "1172:\tlearn: 0.3033181\ttest: 0.9076350\tbest: 0.9051654 (940)\ttotal: 8m 10s\tremaining: 11h 28m 25s\n",
      "1173:\tlearn: 0.3032049\ttest: 0.9078236\tbest: 0.9051654 (940)\ttotal: 8m 10s\tremaining: 11h 28m 25s\n",
      "1174:\tlearn: 0.3030633\ttest: 0.9078087\tbest: 0.9051654 (940)\ttotal: 8m 11s\tremaining: 11h 28m 25s\n",
      "1175:\tlearn: 0.3029260\ttest: 0.9079330\tbest: 0.9051654 (940)\ttotal: 8m 11s\tremaining: 11h 28m 24s\n",
      "1176:\tlearn: 0.3028078\ttest: 0.9079502\tbest: 0.9051654 (940)\ttotal: 8m 11s\tremaining: 11h 28m 25s\n",
      "1177:\tlearn: 0.3026700\ttest: 0.9080657\tbest: 0.9051654 (940)\ttotal: 8m 12s\tremaining: 11h 28m 23s\n",
      "1178:\tlearn: 0.3025544\ttest: 0.9080485\tbest: 0.9051654 (940)\ttotal: 8m 12s\tremaining: 11h 28m 24s\n",
      "1179:\tlearn: 0.3024314\ttest: 0.9079781\tbest: 0.9051654 (940)\ttotal: 8m 13s\tremaining: 11h 28m 23s\n",
      "1180:\tlearn: 0.3023222\ttest: 0.9079749\tbest: 0.9051654 (940)\ttotal: 8m 13s\tremaining: 11h 28m 22s\n",
      "1181:\tlearn: 0.3022160\ttest: 0.9080046\tbest: 0.9051654 (940)\ttotal: 8m 14s\tremaining: 11h 28m 21s\n",
      "1182:\tlearn: 0.3020993\ttest: 0.9081118\tbest: 0.9051654 (940)\ttotal: 8m 14s\tremaining: 11h 28m 20s\n",
      "1183:\tlearn: 0.3019887\ttest: 0.9081423\tbest: 0.9051654 (940)\ttotal: 8m 14s\tremaining: 11h 28m 21s\n",
      "1184:\tlearn: 0.3018732\ttest: 0.9081605\tbest: 0.9051654 (940)\ttotal: 8m 15s\tremaining: 11h 28m 21s\n",
      "1185:\tlearn: 0.3017599\ttest: 0.9081357\tbest: 0.9051654 (940)\ttotal: 8m 15s\tremaining: 11h 28m 21s\n",
      "1186:\tlearn: 0.3016346\ttest: 0.9081680\tbest: 0.9051654 (940)\ttotal: 8m 16s\tremaining: 11h 28m 23s\n",
      "1187:\tlearn: 0.3015436\ttest: 0.9083683\tbest: 0.9051654 (940)\ttotal: 8m 16s\tremaining: 11h 28m 22s\n",
      "1188:\tlearn: 0.3014373\ttest: 0.9084151\tbest: 0.9051654 (940)\ttotal: 8m 16s\tremaining: 11h 28m 22s\n",
      "1189:\tlearn: 0.3013103\ttest: 0.9084084\tbest: 0.9051654 (940)\ttotal: 8m 17s\tremaining: 11h 28m 22s\n",
      "1190:\tlearn: 0.3011980\ttest: 0.9085050\tbest: 0.9051654 (940)\ttotal: 8m 17s\tremaining: 11h 28m 23s\n",
      "Stopped by overfitting detector  (250 iterations wait)\n",
      "\n",
      "bestTest = 0.9051654196\n",
      "bestIteration = 940\n",
      "\n",
      "Shrink model to first 941 iterations.\n",
      "CPU times: user 1h 48min 56s, sys: 25.8 s, total: 1h 49min 22s\n",
      "Wall time: 8min 18s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<catboost.core.CatBoostClassifier at 0x7f72dfb98880>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "N=5\n",
    "DIRECTION='is_turnpt'\n",
    "WEIGHTS_QUNATILE = 0.01\n",
    "\n",
    "model = CatBoostClassifier(iterations=10**5, # set very large number and set early stops\n",
    "                          depth=10, #fine-tune\n",
    "                          learning_rate=0.1, #max: 0.5\n",
    "                          loss_function='MultiClass', \n",
    "                          l2_leaf_reg = 202, #fine-tune\n",
    "                          od_type = 'Iter',\n",
    "                          od_wait = 250,\n",
    "#                           subsample = 0.3,\n",
    "                          rsm = 0.95,\n",
    "                           random_strength = 0.95,\n",
    "                           bagging_temperature = 19.95\n",
    "                          )\n",
    "\n",
    "model.fit(train_pool, eval_set=test_pool)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# print('training pool')\n",
    "# get_precision_recall(model, train_pool)\n",
    "\n",
    "# print('testing pool')\n",
    "# get_precision_recall(model, test_pool)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>y</th>\n",
       "      <th>pred</th>\n",
       "      <th>occurence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     y  pred  occurence\n",
       "0 -1.0  -1.0         27\n",
       "1 -1.0   1.0          5\n",
       "2  0.0  -1.0        240\n",
       "3  0.0   1.0        173\n",
       "4  1.0  -1.0          7\n",
       "5  1.0   1.0         21"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7f72f8291d60>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7f72f8b2a370>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_pred = (\n",
    "    pd.DataFrame(model.predict(test_pool), columns=['pred']).\n",
    "    assign(y=test_pool.get_label())\n",
    ")\n",
    "\n",
    "df_pred_summary = df_pred.groupby(['y', 'pred']).size().to_frame('occurence').reset_index()\n",
    "df_pred_summary\n",
    "\n",
    "\n",
    "df_pred[['pred', 'y']].hist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# plot entry and exit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def plot_turnpt(price, turnpts, dates_index):\n",
    "#     index_upward = np.where(turnpts==1)[0]\n",
    "#     index_downward = np.where(turnpts==-1)[0]\n",
    "    \n",
    "#     plt.figure(figsize=(20, 5))\n",
    "#     plt.plot(dates_index, price, color='black', markevery=index_upward.tolist(), marker='^', markerfacecolor='red', markeredgewidth=0.0)\n",
    "#     plt.plot(dates_index, price, color='black', markevery=index_downward.tolist(), marker='v', markerfacecolor='green', markeredgewidth=0.0)\n",
    "\n",
    "\n",
    "# test_df_y = df_y[df_y.index.isin(y_test.index)]\n",
    "# test_price = test_df_y.close.values\n",
    "# test_turnpts_true = test_df_y.is_turnpt.values\n",
    "# test_dates_index = test_df_y.index\n",
    "# plot_turnpt(test_price, test_turnpts_true, test_dates_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def get_predict_class(model, pool, threshold = 0):\n",
    "#     max_prob = model.predict_proba(pool).max(axis=1)\n",
    "#     turnpts_pred = model.predict(pool).flatten()\n",
    "#     turnpts_pred_filtered = np.where(max_prob>=threshold, turnpts_pred, 0)\n",
    "#     return turnpts_pred_filtered\n",
    "\n",
    "# test_turnpts_pred = get_predict_class(model, test_pool)\n",
    "# plot_turnpt(test_price, test_turnpts_pred, test_dates_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# # only plot 1 or -1 when prob > thereshold\n",
    "# # noises during 2018-M7 disappear\n",
    "# test_turnpts_pred = get_predict_class(model, test_pool, threshold=0.6)\n",
    "# plot_turnpt(test_price, test_turnpts_pred, test_dates_index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# plot NAV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-231-74b3a4708c79>:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  test_df_y['is_turnpt_pred'] =get_predict_class(model, test_pool)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tx_datetime</th>\n",
       "      <th>N</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>vol</th>\n",
       "      <th>maturity</th>\n",
       "      <th>day_to_maturity</th>\n",
       "      <th>is_turnpt_upward</th>\n",
       "      <th>is_turnpt_downward</th>\n",
       "      <th>is_turnpt</th>\n",
       "      <th>next_dt_turnpt</th>\n",
       "      <th>prc_diff</th>\n",
       "      <th>duration</th>\n",
       "      <th>slope</th>\n",
       "      <th>is_turnpt_pred</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tx_datetime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-03-06</th>\n",
       "      <td>2017-03-06</td>\n",
       "      <td>1</td>\n",
       "      <td>9639.0</td>\n",
       "      <td>9684.0</td>\n",
       "      <td>9619.0</td>\n",
       "      <td>9683.0</td>\n",
       "      <td>196244.0</td>\n",
       "      <td>2017-03-15</td>\n",
       "      <td>-9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2017-03-07</td>\n",
       "      <td>44.0</td>\n",
       "      <td>1</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03-07</th>\n",
       "      <td>2017-03-07</td>\n",
       "      <td>1</td>\n",
       "      <td>9693.0</td>\n",
       "      <td>9734.0</td>\n",
       "      <td>9682.0</td>\n",
       "      <td>9727.0</td>\n",
       "      <td>188700.0</td>\n",
       "      <td>2017-03-15</td>\n",
       "      <td>-8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2017-03-08</td>\n",
       "      <td>21.0</td>\n",
       "      <td>1</td>\n",
       "      <td>21.000000</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03-08</th>\n",
       "      <td>2017-03-08</td>\n",
       "      <td>1</td>\n",
       "      <td>9736.0</td>\n",
       "      <td>9765.0</td>\n",
       "      <td>9701.0</td>\n",
       "      <td>9748.0</td>\n",
       "      <td>226610.0</td>\n",
       "      <td>2017-03-15</td>\n",
       "      <td>-7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>2017-03-09</td>\n",
       "      <td>-96.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-96.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03-09</th>\n",
       "      <td>2017-03-09</td>\n",
       "      <td>1</td>\n",
       "      <td>9739.0</td>\n",
       "      <td>9742.0</td>\n",
       "      <td>9642.0</td>\n",
       "      <td>9652.0</td>\n",
       "      <td>277266.0</td>\n",
       "      <td>2017-03-15</td>\n",
       "      <td>-6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2017-03-10</td>\n",
       "      <td>-27.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-27.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-03-10</th>\n",
       "      <td>2017-03-10</td>\n",
       "      <td>1</td>\n",
       "      <td>9647.0</td>\n",
       "      <td>9652.0</td>\n",
       "      <td>9555.0</td>\n",
       "      <td>9625.0</td>\n",
       "      <td>307656.0</td>\n",
       "      <td>2017-03-15</td>\n",
       "      <td>-5</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2017-03-13</td>\n",
       "      <td>67.0</td>\n",
       "      <td>3</td>\n",
       "      <td>22.333333</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            tx_datetime  N    open    high     low   close       vol  \\\n",
       "tx_datetime                                                            \n",
       "2017-03-06   2017-03-06  1  9639.0  9684.0  9619.0  9683.0  196244.0   \n",
       "2017-03-07   2017-03-07  1  9693.0  9734.0  9682.0  9727.0  188700.0   \n",
       "2017-03-08   2017-03-08  1  9736.0  9765.0  9701.0  9748.0  226610.0   \n",
       "2017-03-09   2017-03-09  1  9739.0  9742.0  9642.0  9652.0  277266.0   \n",
       "2017-03-10   2017-03-10  1  9647.0  9652.0  9555.0  9625.0  307656.0   \n",
       "\n",
       "               maturity  day_to_maturity  is_turnpt_upward  \\\n",
       "tx_datetime                                                  \n",
       "2017-03-06   2017-03-15               -9               0.0   \n",
       "2017-03-07   2017-03-15               -8               0.0   \n",
       "2017-03-08   2017-03-15               -7               0.0   \n",
       "2017-03-09   2017-03-15               -6               0.0   \n",
       "2017-03-10   2017-03-15               -5               1.0   \n",
       "\n",
       "             is_turnpt_downward  is_turnpt next_dt_turnpt  prc_diff  duration  \\\n",
       "tx_datetime                                                                     \n",
       "2017-03-06                  0.0        0.0     2017-03-07      44.0         1   \n",
       "2017-03-07                  0.0        0.0     2017-03-08      21.0         1   \n",
       "2017-03-08                  1.0       -1.0     2017-03-09     -96.0         1   \n",
       "2017-03-09                  0.0        0.0     2017-03-10     -27.0         1   \n",
       "2017-03-10                  0.0        1.0     2017-03-13      67.0         3   \n",
       "\n",
       "                 slope  is_turnpt_pred  \n",
       "tx_datetime                             \n",
       "2017-03-06   44.000000            -1.0  \n",
       "2017-03-07   21.000000            -1.0  \n",
       "2017-03-08  -96.000000             0.0  \n",
       "2017-03-09  -27.000000             1.0  \n",
       "2017-03-10   22.333333             1.0  "
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# test_df_y['is_turnpt_pred'] = get_predict_class(model, test_pool)\n",
    "# test_df_y.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # test_df_y\n",
    "# (\n",
    "#     test_df_y.query('is_turnpt_pred != 0').loc[:, ['tx_datetime', 'close']].\n",
    "#     assign(entry = lambda x: x.tx_datetime, \n",
    "#            exit = lambda x: x.tx_datetime.shift(-1)).dropna().\n",
    "#     assign(p_entry = lambda x: x.close.loc[x.entry],# \n",
    "# #            p_exit = lambda x: x.close[x.exit],# \n",
    "# #            ret = lambda x: x.p_exit.div(x.p_entry)-1\n",
    "#     )\n",
    "# #     loc[:, ['entry', 'exit']].\n",
    "# #     join(test_df_y, how='right')\n",
    "# )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# check X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# show metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "jupytext": {
   "formats": "ipynb,py:percent"
  },
  "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": "3.8.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}