BAF_xgboost_multicass_12122020_0834pm.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicción de nivel de penetracion de Banda Ancha fija (fibra óptica o cable coaxial) en municipios\n",
    "\n",
    "El problema a evaluar es el nivel de penetración de BAF basada en fibra óptica o cable coaxial en los municipios (ver nota explicativa de la definición del nivel de penetración, que tomó como referecia los datos de países de la OCDE a finales de 2018).\n",
    "\n",
    "**Nota:** Se empleó un modelo de regresión logística, XGBoost y Random Forest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preámbulo de paquetes a utilizarse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns \n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "\n",
    "import xgboost as xgb\n",
    "import shap\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Carga de datos de penetracion BAF y municipios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cargamos los datos en crudo y creamos una copia para trabajar\n",
    "raw_data = pd.read_csv ('BAF_06209_P2.csv')\n",
    "data = raw_data\n",
    "\n",
    "# Elimina columnas de municipio y clases de penetracion (solo estudiaremos si hay \n",
    "# o no penetracion de F.O y Cable Coax.)\n",
    "del data['K_ENTIDAD_MUNICIPIO']\n",
    "del data['IS_PEN_BAF_HABS_COAXFO'] #\n",
    "del data['ANALF'] #\n",
    "\n",
    "# Renombramos una columna para que no genere problemas con Scikit Learn\n",
    "data.rename(columns={\"PL<5000\": \"PL5000\"}, inplace = True)\n",
    "\n",
    "# Renombramos una columnas para facilitar manipulacion\n",
    "data.rename(columns={\"ANOS_PROMEDIO_DE_ESCOLARIDAD\": \"AVG_ESCOLARIDAD\"}, inplace = True )\n",
    "data.rename(columns={\"CLASS_PEN_BAF_HABS_COAXFO\": \"LEVEL_PEN\"}, inplace = True)\n",
    "\n",
    "# Sustitumos valores de la columnas NUM_OPS (1= Hay mas de dos operadores, 0 = en otro caso)\n",
    "data['NUM_OPS'] = np.where(data['NUM_OPS']>1,1,0)\n",
    "\n",
    "# Descartamos los municipios que no poseen indicadores de ingreso anual bruto per capita\n",
    "data.dropna(inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Resumen de las variables en estudio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>HOGARES</th>\n",
       "      <th>POBLACION</th>\n",
       "      <th>SUPERFICIE</th>\n",
       "      <th>DENS_HOGS</th>\n",
       "      <th>DENS_HABS</th>\n",
       "      <th>SPRIM</th>\n",
       "      <th>AVG_ESCOLARIDAD</th>\n",
       "      <th>OVSAE</th>\n",
       "      <th>OVSEE</th>\n",
       "      <th>PL5000</th>\n",
       "      <th>PO2SM</th>\n",
       "      <th>INGRESOPC_ANUAL</th>\n",
       "      <th>DISP_INTERNET</th>\n",
       "      <th>DISP_TV_PAGA</th>\n",
       "      <th>DISP_TEL_CELULAR</th>\n",
       "      <th>DISP_TEL_FIJO</th>\n",
       "      <th>NUM_OPS</th>\n",
       "      <th>LEVEL_PEN</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2.446000e+03</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "      <td>2446.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>12991.372036</td>\n",
       "      <td>4.860599e+04</td>\n",
       "      <td>791.321885</td>\n",
       "      <td>7916.018110</td>\n",
       "      <td>7916.018110</td>\n",
       "      <td>29.213050</td>\n",
       "      <td>6.423426</td>\n",
       "      <td>8.651132</td>\n",
       "      <td>2.169366</td>\n",
       "      <td>71.898684</td>\n",
       "      <td>55.381063</td>\n",
       "      <td>1935.145724</td>\n",
       "      <td>12.054015</td>\n",
       "      <td>31.158202</td>\n",
       "      <td>57.439859</td>\n",
       "      <td>20.204734</td>\n",
       "      <td>0.226492</td>\n",
       "      <td>0.401881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>38143.144359</td>\n",
       "      <td>1.389142e+05</td>\n",
       "      <td>2104.590464</td>\n",
       "      <td>35001.243252</td>\n",
       "      <td>35001.243252</td>\n",
       "      <td>11.862555</td>\n",
       "      <td>1.767869</td>\n",
       "      <td>11.487796</td>\n",
       "      <td>3.410109</td>\n",
       "      <td>34.685724</td>\n",
       "      <td>16.985484</td>\n",
       "      <td>1020.473455</td>\n",
       "      <td>12.770534</td>\n",
       "      <td>18.829657</td>\n",
       "      <td>25.072398</td>\n",
       "      <td>14.044259</td>\n",
       "      <td>0.418647</td>\n",
       "      <td>0.661519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>31.000000</td>\n",
       "      <td>8.700000e+01</td>\n",
       "      <td>2.210000</td>\n",
       "      <td>4.995331</td>\n",
       "      <td>4.995331</td>\n",
       "      <td>2.490000</td>\n",
       "      <td>1.460000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.250000</td>\n",
       "      <td>185.290000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1111.750000</td>\n",
       "      <td>4.253000e+03</td>\n",
       "      <td>85.765000</td>\n",
       "      <td>521.867640</td>\n",
       "      <td>521.867640</td>\n",
       "      <td>20.500000</td>\n",
       "      <td>5.190000</td>\n",
       "      <td>1.492500</td>\n",
       "      <td>0.490000</td>\n",
       "      <td>42.687500</td>\n",
       "      <td>42.930000</td>\n",
       "      <td>1198.730000</td>\n",
       "      <td>2.100553</td>\n",
       "      <td>15.313693</td>\n",
       "      <td>42.198730</td>\n",
       "      <td>9.047367</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3471.500000</td>\n",
       "      <td>1.340400e+04</td>\n",
       "      <td>233.485000</td>\n",
       "      <td>1387.721595</td>\n",
       "      <td>1387.721595</td>\n",
       "      <td>29.405000</td>\n",
       "      <td>6.280000</td>\n",
       "      <td>4.060000</td>\n",
       "      <td>1.175000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>57.020000</td>\n",
       "      <td>1789.905000</td>\n",
       "      <td>7.954121</td>\n",
       "      <td>29.018092</td>\n",
       "      <td>64.877177</td>\n",
       "      <td>17.645177</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>8877.500000</td>\n",
       "      <td>3.439950e+04</td>\n",
       "      <td>654.895000</td>\n",
       "      <td>3462.189955</td>\n",
       "      <td>3462.189955</td>\n",
       "      <td>37.345000</td>\n",
       "      <td>7.520000</td>\n",
       "      <td>10.865000</td>\n",
       "      <td>2.560000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>68.470000</td>\n",
       "      <td>2447.497500</td>\n",
       "      <td>17.788759</td>\n",
       "      <td>45.173174</td>\n",
       "      <td>76.445478</td>\n",
       "      <td>29.454433</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>495665.000000</td>\n",
       "      <td>1.827868e+06</td>\n",
       "      <td>53138.790000</td>\n",
       "      <td>598127.340824</td>\n",
       "      <td>598127.340824</td>\n",
       "      <td>71.240000</td>\n",
       "      <td>13.830000</td>\n",
       "      <td>98.880000</td>\n",
       "      <td>57.960000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>94.120000</td>\n",
       "      <td>9748.530000</td>\n",
       "      <td>81.882586</td>\n",
       "      <td>85.097192</td>\n",
       "      <td>95.002027</td>\n",
       "      <td>84.047612</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             HOGARES     POBLACION    SUPERFICIE      DENS_HOGS  \\\n",
       "count    2446.000000  2.446000e+03   2446.000000    2446.000000   \n",
       "mean    12991.372036  4.860599e+04    791.321885    7916.018110   \n",
       "std     38143.144359  1.389142e+05   2104.590464   35001.243252   \n",
       "min        31.000000  8.700000e+01      2.210000       4.995331   \n",
       "25%      1111.750000  4.253000e+03     85.765000     521.867640   \n",
       "50%      3471.500000  1.340400e+04    233.485000    1387.721595   \n",
       "75%      8877.500000  3.439950e+04    654.895000    3462.189955   \n",
       "max    495665.000000  1.827868e+06  53138.790000  598127.340824   \n",
       "\n",
       "           DENS_HABS        SPRIM  AVG_ESCOLARIDAD        OVSAE        OVSEE  \\\n",
       "count    2446.000000  2446.000000      2446.000000  2446.000000  2446.000000   \n",
       "mean     7916.018110    29.213050         6.423426     8.651132     2.169366   \n",
       "std     35001.243252    11.862555         1.767869    11.487796     3.410109   \n",
       "min         4.995331     2.490000         1.460000     0.000000     0.000000   \n",
       "25%       521.867640    20.500000         5.190000     1.492500     0.490000   \n",
       "50%      1387.721595    29.405000         6.280000     4.060000     1.175000   \n",
       "75%      3462.189955    37.345000         7.520000    10.865000     2.560000   \n",
       "max    598127.340824    71.240000        13.830000    98.880000    57.960000   \n",
       "\n",
       "            PL5000        PO2SM  INGRESOPC_ANUAL  DISP_INTERNET  DISP_TV_PAGA  \\\n",
       "count  2446.000000  2446.000000      2446.000000    2446.000000   2446.000000   \n",
       "mean     71.898684    55.381063      1935.145724      12.054015     31.158202   \n",
       "std      34.685724    16.985484      1020.473455      12.770534     18.829657   \n",
       "min       0.000000     8.250000       185.290000       0.000000      0.000000   \n",
       "25%      42.687500    42.930000      1198.730000       2.100553     15.313693   \n",
       "50%     100.000000    57.020000      1789.905000       7.954121     29.018092   \n",
       "75%     100.000000    68.470000      2447.497500      17.788759     45.173174   \n",
       "max     100.000000    94.120000      9748.530000      81.882586     85.097192   \n",
       "\n",
       "       DISP_TEL_CELULAR  DISP_TEL_FIJO      NUM_OPS    LEVEL_PEN  \n",
       "count       2446.000000    2446.000000  2446.000000  2446.000000  \n",
       "mean          57.439859      20.204734     0.226492     0.401881  \n",
       "std           25.072398      14.044259     0.418647     0.661519  \n",
       "min            0.000000       0.000000     0.000000     0.000000  \n",
       "25%           42.198730       9.047367     0.000000     0.000000  \n",
       "50%           64.877177      17.645177     0.000000     0.000000  \n",
       "75%           76.445478      29.454433     0.000000     1.000000  \n",
       "max           95.002027      84.047612     1.000000     4.000000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Variables de la base"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "['HOGARES', 'POBLACION', 'SUPERFICIE', 'DENS_HOGS', 'DENS_HABS', 'SPRIM', 'AVG_ESCOLARIDAD', 'OVSAE', 'OVSEE', 'PL5000', 'PO2SM', 'INGRESOPC_ANUAL', 'DISP_INTERNET', 'DISP_TV_PAGA', 'DISP_TEL_CELULAR', 'DISP_TEL_FIJO', 'NUM_OPS', 'LEVEL_PEN']"
      ],
      "text/plain": [
       "['HOGARES',\n",
       " 'POBLACION',\n",
       " 'SUPERFICIE',\n",
       " 'DENS_HOGS',\n",
       " 'DENS_HABS',\n",
       " 'SPRIM',\n",
       " 'AVG_ESCOLARIDAD',\n",
       " 'OVSAE',\n",
       " 'OVSEE',\n",
       " 'PL5000',\n",
       " 'PO2SM',\n",
       " 'INGRESOPC_ANUAL',\n",
       " 'DISP_INTERNET',\n",
       " 'DISP_TV_PAGA',\n",
       " 'DISP_TEL_CELULAR',\n",
       " 'DISP_TEL_FIJO',\n",
       " 'NUM_OPS',\n",
       " 'LEVEL_PEN']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(data.columns.values)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dimensiones de la base"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "m, n = data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Matrices de correlación"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f14294e1668>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "corrmat = data.corr() \n",
    "  \n",
    "f, ax = plt.subplots(figsize =(9, 8)) \n",
    "sns.heatmap(corrmat, ax = ax, cmap =\"YlGn\", linewidths = 0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.matrix.ClusterGrid at 0x7f1429566860>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#corrmat = data.corr()   \n",
    "cg = sns.clustermap(corrmat, cmap =\"RdYlBu\", linewidths = 0.1); \n",
    "plt.setp(cg.ax_heatmap.yaxis.get_majorticklabels(), rotation = 0) \n",
    "  \n",
    "cg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LEVEL_PEN\n",
       "0    1660\n",
       "1     625\n",
       "2     133\n",
       "3      20\n",
       "4       8\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(x='LEVEL_PEN',data=data,palette='hls')\n",
    "data.groupby(['LEVEL_PEN']).size()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preparamos los datos para entrenamiento y prueba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/cesar/.local/lib/python3.7/site-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "seed = 712 # Seed para poder replicar\n",
    "X_train, X_test, y_train, y_test = train_test_split(data.iloc[:, :n-1], data.iloc[:, n-1], random_state=seed)\n",
    "\n",
    "xg_train = xgb.DMatrix(X_train, label=y_train)\n",
    "xg_test = xgb.DMatrix(X_test, label=y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo con regresion logística 1 (All vs All)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_lr1 = LogisticRegression(random_state=seed,solver='lbfgs', max_iter=1000000, multi_class = \"multinomial\").fit(preprocessing.scale(X_train), y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_model_lr1_predicted = model_lr1.predict(preprocessing.scale(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8839869281045751\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.91      0.95       447\n",
      "           1       0.70      0.86      0.77       133\n",
      "           2       0.59      0.67      0.62        30\n",
      "           3       0.25      1.00      0.40         1\n",
      "           4       0.00      0.00      0.00         1\n",
      "\n",
      "    accuracy                           0.88       612\n",
      "   macro avg       0.51      0.69      0.55       612\n",
      "weighted avg       0.90      0.88      0.89       612\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(91.68, 0.5, 'Valor real')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Accuracy\n",
    "from sklearn.metrics import accuracy_score\n",
    "print(\"Accuracy: {}\".format(accuracy_score(y_test, y_model_lr1_predicted)))\n",
    "\n",
    "## Métricas clasicas de Clasificadores\n",
    "from sklearn import metrics\n",
    "print(metrics.classification_report(y_model_lr1_predicted, y_test))\n",
    "\n",
    "## Matriz de confusión\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "mat_lr = confusion_matrix(y_test, y_model_lr1_predicted)\n",
    "\n",
    "plt.close()\n",
    "\n",
    "sns.heatmap(mat_lr, square = True, annot = True, cbar = False)\n",
    "plt.xlabel('Valor predicho')\n",
    "plt.ylabel('Valor real')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo con regresion logística 2 (One vs All)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_lr2 = LogisticRegression(random_state=seed,solver='sag', max_iter=1000000, multi_class = \"ovr\").fit(preprocessing.scale(X_train), y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_model_lr2_predicted = model_lr2.predict(preprocessing.scale(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8839869281045751\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.91      0.95       447\n",
      "           1       0.70      0.86      0.77       133\n",
      "           2       0.59      0.67      0.62        30\n",
      "           3       0.25      1.00      0.40         1\n",
      "           4       0.00      0.00      0.00         1\n",
      "\n",
      "    accuracy                           0.88       612\n",
      "   macro avg       0.51      0.69      0.55       612\n",
      "weighted avg       0.90      0.88      0.89       612\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(91.68, 0.5, 'Valor real')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Accuracy\n",
    "from sklearn.metrics import accuracy_score\n",
    "print(\"Accuracy: {}\".format(accuracy_score(y_test, y_model_lr1_predicted)))\n",
    "\n",
    "## Métricas clasicas de Clasificadores\n",
    "from sklearn import metrics\n",
    "print(metrics.classification_report(y_model_lr1_predicted, y_test))\n",
    "\n",
    "## Matriz de confusión\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "mat_lr = confusion_matrix(y_test, y_model_lr1_predicted)\n",
    "\n",
    "plt.close()\n",
    "\n",
    "sns.heatmap(mat_lr, square = True, annot = True, cbar = False)\n",
    "plt.xlabel('Valor predicho')\n",
    "plt.ylabel('Valor real')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo con XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 6 folds for each of 27 candidates, totalling 162 fits\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.4s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.1s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.0s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.5s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.1s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   3.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.5s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.01, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.5s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.0s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.0s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.05, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   2.0s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=6, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.6s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=7, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.7s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.4s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.2s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=10, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   0.3s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.8s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n",
      "[CV] colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5 \n",
      "[CV]  colsample_bytree=1.0, eta=0.1, gamma=0.5, max_depth=8, n_estimators=100, nthread=10, num_class=5, objective=multi:softmax, seed=27, silent=1, subsample=0.5, total=   1.9s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 162 out of 162 | elapsed:  3.6min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=6, error_score='raise-deprecating',\n",
       "             estimator=XGBClassifier(base_score=0.5, booster='gbtree',\n",
       "                                     colsample_bylevel=1, colsample_bynode=1,\n",
       "                                     colsample_bytree=1, gamma=0,\n",
       "                                     learning_rate=0.1, max_delta_step=0,\n",
       "                                     max_depth=3, min_child_weight=1,\n",
       "                                     missing=None, n_estimators=100, n_jobs=1,\n",
       "                                     nthread=None, objective='binary:logistic',\n",
       "                                     random_state=0, reg_alpha=0, reg_l...\n",
       "                                     subsample=1, verbosity=1),\n",
       "             iid='warn', n_jobs=None,\n",
       "             param_grid={'colsample_bytree': [1.0], 'eta': [0.01, 0.05, 0.1],\n",
       "                         'gamma': [0.5], 'max_depth': [6, 7, 8],\n",
       "                         'n_estimators': [10, 100, 100], 'nthread': [10],\n",
       "                         'num_class': [5], 'objective': ['multi:softmax'],\n",
       "                         'seed': [27], 'silent': [1], 'subsample': [0.5]},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=2)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "# Instanciamos el clasificador\n",
    "classifier_xgb = xgb.XGBClassifier() \n",
    "\n",
    "# Hiper-parametros para hacer el grid search\n",
    "hyper_param_grid = {\n",
    "    'objective':['multi:softmax'],\n",
    "    'eta':[0.01, 0.05,0.1], # Tasa de aprendizaje\n",
    "    'gamma': [0.5],\n",
    "    'max_depth':[6,7,8],\n",
    "    'subsample':[0.8],\n",
    "    'silent':[1],\n",
    "    'nthread':[10],\n",
    "    'num_class':[5],\n",
    "    'subsample':[0.5],\n",
    "    'colsample_bytree':[1.0],\n",
    "    'n_estimators': [10, 100, 100], #number of trees, change it to 1000 for better results\n",
    "    'seed':[27]\n",
    "    }\n",
    "\n",
    "model_gsearch_xgb = GridSearchCV(classifier_xgb, hyper_param_grid, cv = 6, verbose = 2)\n",
    "\n",
    "model_gsearch_xgb.fit(X_train, y_train, eval_metric='auc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'colsample_bytree': 1.0,\n",
       " 'eta': 0.01,\n",
       " 'gamma': 0.5,\n",
       " 'max_depth': 7,\n",
       " 'n_estimators': 10,\n",
       " 'nthread': 10,\n",
       " 'num_class': 5,\n",
       " 'objective': 'multi:softmax',\n",
       " 'seed': 27,\n",
       " 'silent': 1,\n",
       " 'subsample': 0.5}"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#model_gsearch_xgb.grid_scores_ \n",
    "model_gsearch_xgb.best_params_\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8833151581243184"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_gsearch_xgb.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8790849673202614\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.96      0.93      0.95       427\n",
      "           1       0.72      0.82      0.77       141\n",
      "           2       0.71      0.57      0.63        42\n",
      "           3       0.50      1.00      0.67         2\n",
      "           4       0.00      0.00      0.00         0\n",
      "\n",
      "    accuracy                           0.88       612\n",
      "   macro avg       0.58      0.66      0.60       612\n",
      "weighted avg       0.89      0.88      0.88       612\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/cesar/anaconda3/lib/python3.7/site-packages/sklearn/metrics/classification.py:1439: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples.\n",
      "  'recall', 'true', average, warn_for)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(91.68, 0.5, 'Valor real')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_model_xgb = model_gsearch_xgb.predict(X_test)\n",
    "## Accuracy\n",
    "from sklearn.metrics import accuracy_score\n",
    "print(\"Accuracy: {}\".format(accuracy_score(y_test, y_model_xgb)))\n",
    "\n",
    "## Métricas clasicas de Clasificadores\n",
    "from sklearn import metrics\n",
    "print(metrics.classification_report(y_model_xgb, y_test))\n",
    "\n",
    "## Matriz de confusión\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "mat = confusion_matrix(y_test, y_model_xgb)\n",
    "\n",
    "plt.close()\n",
    "\n",
    "sns.heatmap(mat, square = True, annot = True, cbar = False)\n",
    "plt.xlabel('Valor predicho')\n",
    "plt.ylabel('Valor real')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# How the importance is calculated: either \"weight\", \"gain\", or \"cover\" \n",
    "#    \"weight\" is the number of times a feature appears in a tree \n",
    "#    \"gain\" is the average\"gain\"of splits which use the feature \n",
    "#    \"cover\" is the average coverage of splits which use the feature\n",
    "#      where coverage is defined as the number of samples affected by the split\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10,5))\n",
    "xgb.plot_importance(model_gsearch_xgb.best_estimator_, ax=ax, importance_type='weight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# How the importance is calculated: either \"weight\", \"gain\", or \"cover\" \n",
    "#    \"weight\" is the number of times a feature appears in a tree \n",
    "#    \"gain\" is the average\"gain\"of splits which use the feature \n",
    "#    \"cover\" is the average coverage of splits which use the feature\n",
    "#      where coverage is defined as the number of samples affected by the split\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10,5))\n",
    "xgb.plot_importance(model_gsearch_xgb.best_estimator_, ax=ax, importance_type='gain')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# How the importance is calculated: either \"weight\", \"gain\", or \"cover\" \n",
    "#    \"weight\" is the number of times a feature appears in a tree \n",
    "#    \"gain\" is the average\"gain\"of splits which use the feature \n",
    "#    \"cover\" is the average coverage of splits which use the feature\n",
    "#      where coverage is defined as the number of samples affected by the split\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10,5))\n",
    "xgb.plot_importance(model_gsearch_xgb.best_estimator_, ax=ax, importance_type='cover')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create object that can calculate shap values\n",
    "explainer = shap.TreeExplainer(model_gsearch_xgb.best_estimator_)\n",
    "\n",
    "# Calculate shap_values for all of val_X rather than a single row, to have more data for plot.\n",
    "shap_values = explainer.shap_values(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x597.6 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.summary_plot(shap_values, X_test) # Plot de los que no tienen conectividad de BAF (f.o. + coaxial)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x597.6 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Make plot. Index of [0] is explained in text below.\n",
    "shap.summary_plot(shap_values[0], X_test) # Plot de los que no tienen conectividad de BAF (f.o. + coaxial)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x597.6 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.summary_plot(shap_values[1], X_test) # Plot de los que si tienen conectividad de BAF (f.o. + coaxial)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x597.6 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.summary_plot(shap_values[2], X_test) # Plot de los que si tienen conectividad de BAF (f.o. + coaxial)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x597.6 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.summary_plot(shap_values[3], X_test) # Plot de los que si tienen conectividad de BAF (f.o. + coaxial)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x597.6 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.summary_plot(shap_values[4], X_test) # Plot de los que si tienen conectividad de BAF (f.o. + coaxial)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 540x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# https://www.kaggle.com/dansbecker/advanced-uses-of-shap-values\n",
    "\n",
    "shap.dependence_plot('DENS_HOGS', shap_values[1], X_test, interaction_index=\"INGRESOPC_ANUAL\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random forests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 6 folds for each of 90 candidates, totalling 540 fits\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.856, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.807, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.829, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.896, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.0s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.915, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.837, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.868, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.845, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.906, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.925, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.840, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.892, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.868, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.837, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.807, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.829, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.900, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.918, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.833, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.889, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.868, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.903, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.918, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.872, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.837, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.801, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.826, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.819, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.903, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.840, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.882, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.906, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.840, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.885, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.868, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.856, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.807, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.829, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.896, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.915, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.837, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.868, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.845, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.906, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.925, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.840, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.892, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.868, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.837, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.807, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.852, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.829, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.900, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.918, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.833, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.889, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.868, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.903, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.918, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.872, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.837, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.801, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.826, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.832, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.819, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.903, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.840, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.882, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.906, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.840, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.885, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.868, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=5, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.6s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.843, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.810, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.822, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.887, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.840, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.879, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.845, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.906, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.928, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.862, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.872, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.836, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.877, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.879, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.850, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.849, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.915, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.846, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.862, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.868, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.848, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.863, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.810, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.807, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.803, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.799, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.893, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.899, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.853, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.855, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.843, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.810, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.822, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.887, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.840, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.879, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.845, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.906, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.928, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.862, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.872, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.836, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.877, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.879, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.850, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.849, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.915, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.846, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.862, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.868, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.848, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.863, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.810, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.807, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.803, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.799, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.893, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.899, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.853, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.855, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=10, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.809, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.788, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.823, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.806, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.871, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.817, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.790, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.793, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.880, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.922, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.898, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.872, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.853, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.776, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.812, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.874, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.843, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.885, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.853, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.855, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.809, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.788, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.823, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.806, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.871, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.817, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.790, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.793, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.880, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.922, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.898, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.872, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.853, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.776, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.812, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.874, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.843, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.885, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.853, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.855, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=20, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.809, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.788, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.823, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.806, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.871, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.893, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.817, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.790, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.793, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.880, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.922, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.898, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.872, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.853, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.776, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.812, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.874, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.843, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.885, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.853, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.855, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.809, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.788, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.823, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.806, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.871, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.893, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.817, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.790, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.793, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.880, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.922, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.898, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.872, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.853, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.776, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.812, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.874, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.843, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.885, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.853, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.855, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=50, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.809, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.788, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.823, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.806, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.871, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.893, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.817, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.790, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.793, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.880, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.922, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.898, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.872, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.853, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.776, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.812, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.874, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.843, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.885, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.853, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.855, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=sqrt, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.809, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.788, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.823, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.806, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.871, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.905, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.866, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.859, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.893, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.856, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.882, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.8s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=2, n_estimators=100, random_state=712, warm_start=TRUE, score=0.865, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.796, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.820, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.817, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.790, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.816, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=1, random_state=712, warm_start=TRUE, score=0.793, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.880, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.922, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.846, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.898, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.872, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=10, random_state=712, warm_start=TRUE, score=0.862, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.900, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.843, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.859, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=5, n_estimators=100, random_state=712, warm_start=TRUE, score=0.878, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.825, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.853, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.797, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.833, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.776, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=1, random_state=712, warm_start=TRUE, score=0.812, total=   0.0s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.874, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.908, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.843, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.885, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.865, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=10, random_state=712, warm_start=TRUE, score=0.875, total=   0.1s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.896, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.922, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.853, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.889, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.855, total=   0.7s\n",
      "[CV] criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE \n",
      "[CV]  criterion=entropy, max_depth=100, max_features=log2, min_samples_split=10, n_estimators=100, random_state=712, warm_start=TRUE, score=0.875, total=   0.7s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 540 out of 540 | elapsed:  2.4min finished\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "rf_classifier = RandomForestClassifier() \n",
    "hyper_param_grid = {'n_estimators': [1,10,100],#[1,10,100,1000,10000], \n",
    "                    'max_depth': [5,10,20,50,100],#5,10,20,50,100], \n",
    "                    'max_features': ['sqrt','log2'],\n",
    "                    'min_samples_split': [2,5,10],\n",
    "                    'criterion':['entropy'],\n",
    "                    #'class_weight':['balanced'],\n",
    "                    'warm_start' :['TRUE'],\n",
    "                    'random_state':[712]\n",
    "                   },\n",
    "\n",
    "\n",
    "model_RF_gsearch = GridSearchCV(rf_classifier, hyper_param_grid, cv = 6, verbose = 3)\n",
    "\n",
    "model_RF_gsearch.fit(X_train, y_train)\n",
    "\n",
    "y_model_forest = model_RF_gsearch.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'criterion': 'entropy',\n",
       " 'max_depth': 10,\n",
       " 'max_features': 'sqrt',\n",
       " 'min_samples_split': 2,\n",
       " 'n_estimators': 100,\n",
       " 'random_state': 712,\n",
       " 'warm_start': 'TRUE'}"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_RF_gsearch.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8854961832061069"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_RF_gsearch.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8905228758169934\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.93      0.96       430\n",
      "           1       0.75      0.85      0.79       143\n",
      "           2       0.59      0.57      0.58        35\n",
      "           3       0.50      0.67      0.57         3\n",
      "           4       0.00      0.00      0.00         1\n",
      "\n",
      "    accuracy                           0.89       612\n",
      "   macro avg       0.56      0.60      0.58       612\n",
      "weighted avg       0.90      0.89      0.89       612\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(91.68, 0.5, 'Valor real')"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Accuracy\n",
    "from sklearn.metrics import accuracy_score\n",
    "print(\"Accuracy: {}\".format(accuracy_score(y_test, y_model_forest)))\n",
    "\n",
    "## Métricas clasicas de Clasificadores\n",
    "from sklearn import metrics\n",
    "print(metrics.classification_report(y_model_forest, y_test))\n",
    "\n",
    "## Matriz de confusión\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "mat = confusion_matrix(y_test, y_model_forest)\n",
    "\n",
    "plt.close()\n",
    "\n",
    "sns.heatmap(mat, square = True, annot = True, cbar = False)\n",
    "plt.xlabel('Valor predicho')\n",
    "plt.ylabel('Valor real')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}