op0_60_XGboost.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.90\n"
     ]
    }
   ],
   "source": [
    "import xgboost as xgb\n",
    "print(xgb.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import plot_importance\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_csv('./EGR_Cooler_data/material_train_data.csv')\n",
    "test = pd.read_csv('./EGR_Cooler_data/material_test_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>body_length</th>\n",
       "      <th>body_width</th>\n",
       "      <th>body_height</th>\n",
       "      <th>wavyfin_length</th>\n",
       "      <th>wavyfin_width</th>\n",
       "      <th>wavyfin_height</th>\n",
       "      <th>pipe1_length</th>\n",
       "      <th>pipe1_width</th>\n",
       "      <th>pipe1_height</th>\n",
       "      <th>pipe2_length</th>\n",
       "      <th>pipe2_width</th>\n",
       "      <th>pipe2_height</th>\n",
       "      <th>flange1_length</th>\n",
       "      <th>flange1_width</th>\n",
       "      <th>flange1_height</th>\n",
       "      <th>flange2_length</th>\n",
       "      <th>flange2_width</th>\n",
       "      <th>flange2_height</th>\n",
       "      <th>op60_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>199.996</td>\n",
       "      <td>100.002</td>\n",
       "      <td>50.001</td>\n",
       "      <td>100.000</td>\n",
       "      <td>49.997</td>\n",
       "      <td>59.997</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.003</td>\n",
       "      <td>29.996</td>\n",
       "      <td>30.000</td>\n",
       "      <td>49.996</td>\n",
       "      <td>29.999</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.000</td>\n",
       "      <td>40.003</td>\n",
       "      <td>30.001</td>\n",
       "      <td>79.998</td>\n",
       "      <td>39.999</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>200.002</td>\n",
       "      <td>100.000</td>\n",
       "      <td>49.998</td>\n",
       "      <td>99.999</td>\n",
       "      <td>50.002</td>\n",
       "      <td>60.000</td>\n",
       "      <td>29.999</td>\n",
       "      <td>49.994</td>\n",
       "      <td>29.999</td>\n",
       "      <td>30.007</td>\n",
       "      <td>50.000</td>\n",
       "      <td>29.998</td>\n",
       "      <td>30.000</td>\n",
       "      <td>79.997</td>\n",
       "      <td>40.002</td>\n",
       "      <td>30.000</td>\n",
       "      <td>79.995</td>\n",
       "      <td>40.000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>199.998</td>\n",
       "      <td>100.000</td>\n",
       "      <td>49.998</td>\n",
       "      <td>100.003</td>\n",
       "      <td>50.003</td>\n",
       "      <td>59.998</td>\n",
       "      <td>29.995</td>\n",
       "      <td>50.002</td>\n",
       "      <td>30.000</td>\n",
       "      <td>30.003</td>\n",
       "      <td>50.005</td>\n",
       "      <td>30.001</td>\n",
       "      <td>30.000</td>\n",
       "      <td>80.001</td>\n",
       "      <td>39.993</td>\n",
       "      <td>29.997</td>\n",
       "      <td>79.999</td>\n",
       "      <td>39.995</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>200.002</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.000</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.003</td>\n",
       "      <td>60.000</td>\n",
       "      <td>30.001</td>\n",
       "      <td>49.995</td>\n",
       "      <td>29.998</td>\n",
       "      <td>30.002</td>\n",
       "      <td>50.000</td>\n",
       "      <td>29.997</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.002</td>\n",
       "      <td>40.001</td>\n",
       "      <td>29.998</td>\n",
       "      <td>79.997</td>\n",
       "      <td>39.998</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>200.001</td>\n",
       "      <td>100.001</td>\n",
       "      <td>49.999</td>\n",
       "      <td>100.003</td>\n",
       "      <td>49.997</td>\n",
       "      <td>59.996</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.001</td>\n",
       "      <td>29.996</td>\n",
       "      <td>30.002</td>\n",
       "      <td>50.004</td>\n",
       "      <td>29.997</td>\n",
       "      <td>29.999</td>\n",
       "      <td>80.004</td>\n",
       "      <td>40.000</td>\n",
       "      <td>29.997</td>\n",
       "      <td>79.998</td>\n",
       "      <td>40.000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   body_length  body_width  body_height  wavyfin_length  wavyfin_width  \\\n",
       "0      199.996     100.002       50.001         100.000         49.997   \n",
       "1      200.002     100.000       49.998          99.999         50.002   \n",
       "2      199.998     100.000       49.998         100.003         50.003   \n",
       "3      200.002     100.000       50.000         100.000         50.003   \n",
       "4      200.001     100.001       49.999         100.003         49.997   \n",
       "\n",
       "   wavyfin_height  pipe1_length  pipe1_width  pipe1_height  pipe2_length  \\\n",
       "0          59.997        30.000       50.003        29.996        30.000   \n",
       "1          60.000        29.999       49.994        29.999        30.007   \n",
       "2          59.998        29.995       50.002        30.000        30.003   \n",
       "3          60.000        30.001       49.995        29.998        30.002   \n",
       "4          59.996        30.000       50.001        29.996        30.002   \n",
       "\n",
       "   pipe2_width  pipe2_height  flange1_length  flange1_width  flange1_height  \\\n",
       "0       49.996        29.999          30.002         80.000          40.003   \n",
       "1       50.000        29.998          30.000         79.997          40.002   \n",
       "2       50.005        30.001          30.000         80.001          39.993   \n",
       "3       50.000        29.997          30.002         80.002          40.001   \n",
       "4       50.004        29.997          29.999         80.004          40.000   \n",
       "\n",
       "   flange2_length  flange2_width  flange2_height  op60_test  \n",
       "0          30.001         79.998          39.999          1  \n",
       "1          30.000         79.995          40.000          0  \n",
       "2          29.997         79.999          39.995          0  \n",
       "3          29.998         79.997          39.998          0  \n",
       "4          29.997         79.998          40.000          0  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>body_length</th>\n",
       "      <th>body_width</th>\n",
       "      <th>body_height</th>\n",
       "      <th>wavyfin_length</th>\n",
       "      <th>wavyfin_width</th>\n",
       "      <th>wavyfin_height</th>\n",
       "      <th>pipe1_length</th>\n",
       "      <th>pipe1_width</th>\n",
       "      <th>pipe1_height</th>\n",
       "      <th>pipe2_length</th>\n",
       "      <th>pipe2_width</th>\n",
       "      <th>pipe2_height</th>\n",
       "      <th>flange1_length</th>\n",
       "      <th>flange1_width</th>\n",
       "      <th>flange1_height</th>\n",
       "      <th>flange2_length</th>\n",
       "      <th>flange2_width</th>\n",
       "      <th>flange2_height</th>\n",
       "      <th>op60_test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>199.997</td>\n",
       "      <td>99.999</td>\n",
       "      <td>50.003</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.005</td>\n",
       "      <td>60.003</td>\n",
       "      <td>29.996</td>\n",
       "      <td>50.004</td>\n",
       "      <td>30.002</td>\n",
       "      <td>29.998</td>\n",
       "      <td>50.002</td>\n",
       "      <td>30.003</td>\n",
       "      <td>30.001</td>\n",
       "      <td>80.001</td>\n",
       "      <td>40.007</td>\n",
       "      <td>30.002</td>\n",
       "      <td>79.999</td>\n",
       "      <td>40.004</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>199.999</td>\n",
       "      <td>100.001</td>\n",
       "      <td>50.002</td>\n",
       "      <td>99.996</td>\n",
       "      <td>50.000</td>\n",
       "      <td>60.004</td>\n",
       "      <td>29.998</td>\n",
       "      <td>50.004</td>\n",
       "      <td>29.995</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.000</td>\n",
       "      <td>30.002</td>\n",
       "      <td>30.003</td>\n",
       "      <td>80.001</td>\n",
       "      <td>39.999</td>\n",
       "      <td>29.998</td>\n",
       "      <td>79.999</td>\n",
       "      <td>39.999</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>199.999</td>\n",
       "      <td>100.001</td>\n",
       "      <td>50.002</td>\n",
       "      <td>100.001</td>\n",
       "      <td>50.001</td>\n",
       "      <td>60.003</td>\n",
       "      <td>29.997</td>\n",
       "      <td>50.001</td>\n",
       "      <td>30.003</td>\n",
       "      <td>30.001</td>\n",
       "      <td>50.002</td>\n",
       "      <td>30.001</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.003</td>\n",
       "      <td>39.998</td>\n",
       "      <td>29.999</td>\n",
       "      <td>80.001</td>\n",
       "      <td>39.997</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>200.002</td>\n",
       "      <td>99.996</td>\n",
       "      <td>49.999</td>\n",
       "      <td>100.001</td>\n",
       "      <td>49.998</td>\n",
       "      <td>60.000</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.006</td>\n",
       "      <td>30.001</td>\n",
       "      <td>30.001</td>\n",
       "      <td>49.999</td>\n",
       "      <td>29.999</td>\n",
       "      <td>29.996</td>\n",
       "      <td>80.002</td>\n",
       "      <td>39.998</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.002</td>\n",
       "      <td>39.999</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>199.999</td>\n",
       "      <td>100.003</td>\n",
       "      <td>50.003</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.001</td>\n",
       "      <td>60.000</td>\n",
       "      <td>30.000</td>\n",
       "      <td>49.998</td>\n",
       "      <td>30.002</td>\n",
       "      <td>30.001</td>\n",
       "      <td>50.000</td>\n",
       "      <td>29.999</td>\n",
       "      <td>30.004</td>\n",
       "      <td>79.998</td>\n",
       "      <td>39.999</td>\n",
       "      <td>30.003</td>\n",
       "      <td>80.002</td>\n",
       "      <td>40.000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   body_length  body_width  body_height  wavyfin_length  wavyfin_width  \\\n",
       "0      199.997      99.999       50.003         100.000         50.005   \n",
       "1      199.999     100.001       50.002          99.996         50.000   \n",
       "2      199.999     100.001       50.002         100.001         50.001   \n",
       "3      200.002      99.996       49.999         100.001         49.998   \n",
       "4      199.999     100.003       50.003         100.000         50.001   \n",
       "\n",
       "   wavyfin_height  pipe1_length  pipe1_width  pipe1_height  pipe2_length  \\\n",
       "0          60.003        29.996       50.004        30.002        29.998   \n",
       "1          60.004        29.998       50.004        29.995        30.000   \n",
       "2          60.003        29.997       50.001        30.003        30.001   \n",
       "3          60.000        30.000       50.006        30.001        30.001   \n",
       "4          60.000        30.000       49.998        30.002        30.001   \n",
       "\n",
       "   pipe2_width  pipe2_height  flange1_length  flange1_width  flange1_height  \\\n",
       "0       50.002        30.003          30.001         80.001          40.007   \n",
       "1       50.000        30.002          30.003         80.001          39.999   \n",
       "2       50.002        30.001          30.002         80.003          39.998   \n",
       "3       49.999        29.999          29.996         80.002          39.998   \n",
       "4       50.000        29.999          30.004         79.998          39.999   \n",
       "\n",
       "   flange2_length  flange2_width  flange2_height  op60_test  \n",
       "0          30.002         79.999          40.004          0  \n",
       "1          29.998         79.999          39.999          0  \n",
       "2          29.999         80.001          39.997          0  \n",
       "3          30.002         80.002          39.999          0  \n",
       "4          30.003         80.002          40.000          0  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = train.drop(['op60_test'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>body_length</th>\n",
       "      <th>body_width</th>\n",
       "      <th>body_height</th>\n",
       "      <th>wavyfin_length</th>\n",
       "      <th>wavyfin_width</th>\n",
       "      <th>wavyfin_height</th>\n",
       "      <th>pipe1_length</th>\n",
       "      <th>pipe1_width</th>\n",
       "      <th>pipe1_height</th>\n",
       "      <th>pipe2_length</th>\n",
       "      <th>pipe2_width</th>\n",
       "      <th>pipe2_height</th>\n",
       "      <th>flange1_length</th>\n",
       "      <th>flange1_width</th>\n",
       "      <th>flange1_height</th>\n",
       "      <th>flange2_length</th>\n",
       "      <th>flange2_width</th>\n",
       "      <th>flange2_height</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>199.996</td>\n",
       "      <td>100.002</td>\n",
       "      <td>50.001</td>\n",
       "      <td>100.000</td>\n",
       "      <td>49.997</td>\n",
       "      <td>59.997</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.003</td>\n",
       "      <td>29.996</td>\n",
       "      <td>30.000</td>\n",
       "      <td>49.996</td>\n",
       "      <td>29.999</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.000</td>\n",
       "      <td>40.003</td>\n",
       "      <td>30.001</td>\n",
       "      <td>79.998</td>\n",
       "      <td>39.999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>200.002</td>\n",
       "      <td>100.000</td>\n",
       "      <td>49.998</td>\n",
       "      <td>99.999</td>\n",
       "      <td>50.002</td>\n",
       "      <td>60.000</td>\n",
       "      <td>29.999</td>\n",
       "      <td>49.994</td>\n",
       "      <td>29.999</td>\n",
       "      <td>30.007</td>\n",
       "      <td>50.000</td>\n",
       "      <td>29.998</td>\n",
       "      <td>30.000</td>\n",
       "      <td>79.997</td>\n",
       "      <td>40.002</td>\n",
       "      <td>30.000</td>\n",
       "      <td>79.995</td>\n",
       "      <td>40.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>199.998</td>\n",
       "      <td>100.000</td>\n",
       "      <td>49.998</td>\n",
       "      <td>100.003</td>\n",
       "      <td>50.003</td>\n",
       "      <td>59.998</td>\n",
       "      <td>29.995</td>\n",
       "      <td>50.002</td>\n",
       "      <td>30.000</td>\n",
       "      <td>30.003</td>\n",
       "      <td>50.005</td>\n",
       "      <td>30.001</td>\n",
       "      <td>30.000</td>\n",
       "      <td>80.001</td>\n",
       "      <td>39.993</td>\n",
       "      <td>29.997</td>\n",
       "      <td>79.999</td>\n",
       "      <td>39.995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>200.002</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.000</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.003</td>\n",
       "      <td>60.000</td>\n",
       "      <td>30.001</td>\n",
       "      <td>49.995</td>\n",
       "      <td>29.998</td>\n",
       "      <td>30.002</td>\n",
       "      <td>50.000</td>\n",
       "      <td>29.997</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.002</td>\n",
       "      <td>40.001</td>\n",
       "      <td>29.998</td>\n",
       "      <td>79.997</td>\n",
       "      <td>39.998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>200.001</td>\n",
       "      <td>100.001</td>\n",
       "      <td>49.999</td>\n",
       "      <td>100.003</td>\n",
       "      <td>49.997</td>\n",
       "      <td>59.996</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.001</td>\n",
       "      <td>29.996</td>\n",
       "      <td>30.002</td>\n",
       "      <td>50.004</td>\n",
       "      <td>29.997</td>\n",
       "      <td>29.999</td>\n",
       "      <td>80.004</td>\n",
       "      <td>40.000</td>\n",
       "      <td>29.997</td>\n",
       "      <td>79.998</td>\n",
       "      <td>40.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   body_length  body_width  body_height  wavyfin_length  wavyfin_width  \\\n",
       "0      199.996     100.002       50.001         100.000         49.997   \n",
       "1      200.002     100.000       49.998          99.999         50.002   \n",
       "2      199.998     100.000       49.998         100.003         50.003   \n",
       "3      200.002     100.000       50.000         100.000         50.003   \n",
       "4      200.001     100.001       49.999         100.003         49.997   \n",
       "\n",
       "   wavyfin_height  pipe1_length  pipe1_width  pipe1_height  pipe2_length  \\\n",
       "0          59.997        30.000       50.003        29.996        30.000   \n",
       "1          60.000        29.999       49.994        29.999        30.007   \n",
       "2          59.998        29.995       50.002        30.000        30.003   \n",
       "3          60.000        30.001       49.995        29.998        30.002   \n",
       "4          59.996        30.000       50.001        29.996        30.002   \n",
       "\n",
       "   pipe2_width  pipe2_height  flange1_length  flange1_width  flange1_height  \\\n",
       "0       49.996        29.999          30.002         80.000          40.003   \n",
       "1       50.000        29.998          30.000         79.997          40.002   \n",
       "2       50.005        30.001          30.000         80.001          39.993   \n",
       "3       50.000        29.997          30.002         80.002          40.001   \n",
       "4       50.004        29.997          29.999         80.004          40.000   \n",
       "\n",
       "   flange2_length  flange2_width  flange2_height  \n",
       "0          30.001         79.998          39.999  \n",
       "1          30.000         79.995          40.000  \n",
       "2          29.997         79.999          39.995  \n",
       "3          29.998         79.997          39.998  \n",
       "4          29.997         79.998          40.000  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1\n",
       "1    0\n",
       "2    0\n",
       "3    0\n",
       "4    0\n",
       "Name: op60_test, dtype: int64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_train = train['op60_test']\n",
    "Y_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>body_length</th>\n",
       "      <th>body_width</th>\n",
       "      <th>body_height</th>\n",
       "      <th>wavyfin_length</th>\n",
       "      <th>wavyfin_width</th>\n",
       "      <th>wavyfin_height</th>\n",
       "      <th>pipe1_length</th>\n",
       "      <th>pipe1_width</th>\n",
       "      <th>pipe1_height</th>\n",
       "      <th>pipe2_length</th>\n",
       "      <th>pipe2_width</th>\n",
       "      <th>pipe2_height</th>\n",
       "      <th>flange1_length</th>\n",
       "      <th>flange1_width</th>\n",
       "      <th>flange1_height</th>\n",
       "      <th>flange2_length</th>\n",
       "      <th>flange2_width</th>\n",
       "      <th>flange2_height</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>199.997</td>\n",
       "      <td>99.999</td>\n",
       "      <td>50.003</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.005</td>\n",
       "      <td>60.003</td>\n",
       "      <td>29.996</td>\n",
       "      <td>50.004</td>\n",
       "      <td>30.002</td>\n",
       "      <td>29.998</td>\n",
       "      <td>50.002</td>\n",
       "      <td>30.003</td>\n",
       "      <td>30.001</td>\n",
       "      <td>80.001</td>\n",
       "      <td>40.007</td>\n",
       "      <td>30.002</td>\n",
       "      <td>79.999</td>\n",
       "      <td>40.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>199.999</td>\n",
       "      <td>100.001</td>\n",
       "      <td>50.002</td>\n",
       "      <td>99.996</td>\n",
       "      <td>50.000</td>\n",
       "      <td>60.004</td>\n",
       "      <td>29.998</td>\n",
       "      <td>50.004</td>\n",
       "      <td>29.995</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.000</td>\n",
       "      <td>30.002</td>\n",
       "      <td>30.003</td>\n",
       "      <td>80.001</td>\n",
       "      <td>39.999</td>\n",
       "      <td>29.998</td>\n",
       "      <td>79.999</td>\n",
       "      <td>39.999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>199.999</td>\n",
       "      <td>100.001</td>\n",
       "      <td>50.002</td>\n",
       "      <td>100.001</td>\n",
       "      <td>50.001</td>\n",
       "      <td>60.003</td>\n",
       "      <td>29.997</td>\n",
       "      <td>50.001</td>\n",
       "      <td>30.003</td>\n",
       "      <td>30.001</td>\n",
       "      <td>50.002</td>\n",
       "      <td>30.001</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.003</td>\n",
       "      <td>39.998</td>\n",
       "      <td>29.999</td>\n",
       "      <td>80.001</td>\n",
       "      <td>39.997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>200.002</td>\n",
       "      <td>99.996</td>\n",
       "      <td>49.999</td>\n",
       "      <td>100.001</td>\n",
       "      <td>49.998</td>\n",
       "      <td>60.000</td>\n",
       "      <td>30.000</td>\n",
       "      <td>50.006</td>\n",
       "      <td>30.001</td>\n",
       "      <td>30.001</td>\n",
       "      <td>49.999</td>\n",
       "      <td>29.999</td>\n",
       "      <td>29.996</td>\n",
       "      <td>80.002</td>\n",
       "      <td>39.998</td>\n",
       "      <td>30.002</td>\n",
       "      <td>80.002</td>\n",
       "      <td>39.999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>199.999</td>\n",
       "      <td>100.003</td>\n",
       "      <td>50.003</td>\n",
       "      <td>100.000</td>\n",
       "      <td>50.001</td>\n",
       "      <td>60.000</td>\n",
       "      <td>30.000</td>\n",
       "      <td>49.998</td>\n",
       "      <td>30.002</td>\n",
       "      <td>30.001</td>\n",
       "      <td>50.000</td>\n",
       "      <td>29.999</td>\n",
       "      <td>30.004</td>\n",
       "      <td>79.998</td>\n",
       "      <td>39.999</td>\n",
       "      <td>30.003</td>\n",
       "      <td>80.002</td>\n",
       "      <td>40.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   body_length  body_width  body_height  wavyfin_length  wavyfin_width  \\\n",
       "0      199.997      99.999       50.003         100.000         50.005   \n",
       "1      199.999     100.001       50.002          99.996         50.000   \n",
       "2      199.999     100.001       50.002         100.001         50.001   \n",
       "3      200.002      99.996       49.999         100.001         49.998   \n",
       "4      199.999     100.003       50.003         100.000         50.001   \n",
       "\n",
       "   wavyfin_height  pipe1_length  pipe1_width  pipe1_height  pipe2_length  \\\n",
       "0          60.003        29.996       50.004        30.002        29.998   \n",
       "1          60.004        29.998       50.004        29.995        30.000   \n",
       "2          60.003        29.997       50.001        30.003        30.001   \n",
       "3          60.000        30.000       50.006        30.001        30.001   \n",
       "4          60.000        30.000       49.998        30.002        30.001   \n",
       "\n",
       "   pipe2_width  pipe2_height  flange1_length  flange1_width  flange1_height  \\\n",
       "0       50.002        30.003          30.001         80.001          40.007   \n",
       "1       50.000        30.002          30.003         80.001          39.999   \n",
       "2       50.002        30.001          30.002         80.003          39.998   \n",
       "3       49.999        29.999          29.996         80.002          39.998   \n",
       "4       50.000        29.999          30.004         79.998          39.999   \n",
       "\n",
       "   flange2_length  flange2_width  flange2_height  \n",
       "0          30.002         79.999          40.004  \n",
       "1          29.998         79.999          39.999  \n",
       "2          29.999         80.001          39.997  \n",
       "3          30.002         80.002          39.999  \n",
       "4          30.003         80.002          40.000  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test = test.drop(['op60_test'], axis = 1)\n",
    "X_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0\n",
       "1    0\n",
       "2    0\n",
       "3    0\n",
       "4    0\n",
       "Name: op60_test, dtype: int64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_test = test['op60_test']\n",
    "Y_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "dtrain = xgb.DMatrix(data = X_train, label = Y_train)\n",
    "dtest = xgb.DMatrix(data = X_test, label = Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "wlist = [(dtrain, 'train'), (dtest, 'eval')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {'max_depth' : 20,\n",
    "         'eta': 0.001,\n",
    "         'objective' : 'binary:logistic',\n",
    "          'eval_metric' : 'error',\n",
    "         'early_stopping':200\n",
    "         }\n",
    "num_rounds = 1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\ttrain-error:0.0425\teval-error:0.059\n",
      "Multiple eval metrics have been passed: 'eval-error' will be used for early stopping.\n",
      "\n",
      "Will train until eval-error hasn't improved in 1000 rounds.\n",
      "[1]\ttrain-error:0.042\teval-error:0.059\n",
      "[2]\ttrain-error:0.042625\teval-error:0.0565\n",
      "[3]\ttrain-error:0.04275\teval-error:0.057\n",
      "[4]\ttrain-error:0.042625\teval-error:0.0565\n",
      "[5]\ttrain-error:0.04275\teval-error:0.056\n",
      "[6]\ttrain-error:0.042625\teval-error:0.057\n",
      "[7]\ttrain-error:0.043\teval-error:0.058\n",
      "[8]\ttrain-error:0.042875\teval-error:0.058\n",
      "[9]\ttrain-error:0.043\teval-error:0.058\n",
      "[10]\ttrain-error:0.04275\teval-error:0.058\n",
      "[11]\ttrain-error:0.043125\teval-error:0.058\n",
      "[12]\ttrain-error:0.043\teval-error:0.058\n",
      "[13]\ttrain-error:0.043125\teval-error:0.058\n",
      "[14]\ttrain-error:0.043\teval-error:0.058\n",
      "[15]\ttrain-error:0.043125\teval-error:0.058\n",
      "[16]\ttrain-error:0.043125\teval-error:0.058\n",
      "[17]\ttrain-error:0.043125\teval-error:0.058\n",
      "[18]\ttrain-error:0.043125\teval-error:0.058\n",
      "[19]\ttrain-error:0.043125\teval-error:0.058\n",
      "[20]\ttrain-error:0.043125\teval-error:0.058\n",
      "[21]\ttrain-error:0.04325\teval-error:0.058\n",
      "[22]\ttrain-error:0.043125\teval-error:0.058\n",
      "[23]\ttrain-error:0.043125\teval-error:0.058\n",
      "[24]\ttrain-error:0.043125\teval-error:0.058\n",
      "[25]\ttrain-error:0.043125\teval-error:0.058\n",
      "[26]\ttrain-error:0.043125\teval-error:0.058\n",
      "[27]\ttrain-error:0.043125\teval-error:0.058\n",
      "[28]\ttrain-error:0.043125\teval-error:0.058\n",
      "[29]\ttrain-error:0.043125\teval-error:0.058\n",
      "[30]\ttrain-error:0.043125\teval-error:0.058\n",
      "[31]\ttrain-error:0.043125\teval-error:0.058\n",
      "[32]\ttrain-error:0.043125\teval-error:0.058\n",
      "[33]\ttrain-error:0.043125\teval-error:0.058\n",
      "[34]\ttrain-error:0.043125\teval-error:0.058\n",
      "[35]\ttrain-error:0.043125\teval-error:0.058\n",
      "[36]\ttrain-error:0.043125\teval-error:0.058\n",
      "[37]\ttrain-error:0.043125\teval-error:0.058\n",
      "[38]\ttrain-error:0.04325\teval-error:0.058\n",
      "[39]\ttrain-error:0.04325\teval-error:0.058\n",
      "[40]\ttrain-error:0.043125\teval-error:0.058\n",
      "[41]\ttrain-error:0.043125\teval-error:0.058\n",
      "[42]\ttrain-error:0.04325\teval-error:0.058\n",
      "[43]\ttrain-error:0.04325\teval-error:0.058\n",
      "[44]\ttrain-error:0.04325\teval-error:0.058\n",
      "[45]\ttrain-error:0.043375\teval-error:0.058\n",
      "[46]\ttrain-error:0.04325\teval-error:0.058\n",
      "[47]\ttrain-error:0.04325\teval-error:0.0575\n",
      "[48]\ttrain-error:0.043375\teval-error:0.0575\n",
      "[49]\ttrain-error:0.04325\teval-error:0.057\n",
      "[50]\ttrain-error:0.0435\teval-error:0.057\n",
      "[51]\ttrain-error:0.0435\teval-error:0.057\n",
      "[52]\ttrain-error:0.043375\teval-error:0.057\n",
      "[53]\ttrain-error:0.0435\teval-error:0.057\n",
      "[54]\ttrain-error:0.043375\teval-error:0.057\n",
      "[55]\ttrain-error:0.043375\teval-error:0.057\n",
      "[56]\ttrain-error:0.043375\teval-error:0.057\n",
      "[57]\ttrain-error:0.043375\teval-error:0.057\n",
      "[58]\ttrain-error:0.043375\teval-error:0.057\n",
      "[59]\ttrain-error:0.043375\teval-error:0.057\n",
      "[60]\ttrain-error:0.043375\teval-error:0.057\n",
      "[61]\ttrain-error:0.043375\teval-error:0.057\n",
      "[62]\ttrain-error:0.043375\teval-error:0.057\n",
      "[63]\ttrain-error:0.043375\teval-error:0.057\n",
      "[64]\ttrain-error:0.043375\teval-error:0.057\n",
      "[65]\ttrain-error:0.043375\teval-error:0.057\n",
      "[66]\ttrain-error:0.043375\teval-error:0.057\n",
      "[67]\ttrain-error:0.043375\teval-error:0.057\n",
      "[68]\ttrain-error:0.043375\teval-error:0.057\n",
      "[69]\ttrain-error:0.043375\teval-error:0.057\n",
      "[70]\ttrain-error:0.043375\teval-error:0.057\n",
      "[71]\ttrain-error:0.043375\teval-error:0.057\n",
      "[72]\ttrain-error:0.043375\teval-error:0.057\n",
      "[73]\ttrain-error:0.043375\teval-error:0.057\n",
      "[74]\ttrain-error:0.043375\teval-error:0.057\n",
      "[75]\ttrain-error:0.043375\teval-error:0.057\n",
      "[76]\ttrain-error:0.043375\teval-error:0.057\n",
      "[77]\ttrain-error:0.04325\teval-error:0.057\n",
      "[78]\ttrain-error:0.04325\teval-error:0.057\n",
      "[79]\ttrain-error:0.04325\teval-error:0.057\n",
      "[80]\ttrain-error:0.04325\teval-error:0.057\n",
      "[81]\ttrain-error:0.04325\teval-error:0.057\n",
      "[82]\ttrain-error:0.04325\teval-error:0.057\n",
      "[83]\ttrain-error:0.04325\teval-error:0.057\n",
      "[84]\ttrain-error:0.04325\teval-error:0.057\n",
      "[85]\ttrain-error:0.04325\teval-error:0.057\n",
      "[86]\ttrain-error:0.04325\teval-error:0.057\n",
      "[87]\ttrain-error:0.043125\teval-error:0.057\n",
      "[88]\ttrain-error:0.043125\teval-error:0.057\n",
      "[89]\ttrain-error:0.043125\teval-error:0.057\n",
      "[90]\ttrain-error:0.043125\teval-error:0.057\n",
      "[91]\ttrain-error:0.043125\teval-error:0.057\n",
      "[92]\ttrain-error:0.043125\teval-error:0.057\n",
      "[93]\ttrain-error:0.043125\teval-error:0.057\n",
      "[94]\ttrain-error:0.043125\teval-error:0.057\n",
      "[95]\ttrain-error:0.043125\teval-error:0.057\n",
      "[96]\ttrain-error:0.043125\teval-error:0.057\n",
      "[97]\ttrain-error:0.043125\teval-error:0.057\n",
      "[98]\ttrain-error:0.043125\teval-error:0.057\n",
      "[99]\ttrain-error:0.043125\teval-error:0.057\n",
      "[100]\ttrain-error:0.043125\teval-error:0.057\n",
      "[101]\ttrain-error:0.043125\teval-error:0.057\n",
      "[102]\ttrain-error:0.043125\teval-error:0.057\n",
      "[103]\ttrain-error:0.043125\teval-error:0.057\n",
      "[104]\ttrain-error:0.043125\teval-error:0.057\n",
      "[105]\ttrain-error:0.043125\teval-error:0.057\n",
      "[106]\ttrain-error:0.043125\teval-error:0.057\n",
      "[107]\ttrain-error:0.043125\teval-error:0.057\n",
      "[108]\ttrain-error:0.043125\teval-error:0.057\n",
      "[109]\ttrain-error:0.04325\teval-error:0.057\n",
      "[110]\ttrain-error:0.04325\teval-error:0.057\n",
      "[111]\ttrain-error:0.04325\teval-error:0.0575\n",
      "[112]\ttrain-error:0.04325\teval-error:0.0575\n",
      "[113]\ttrain-error:0.04325\teval-error:0.0575\n",
      "[114]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[115]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[116]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[117]\ttrain-error:0.043\teval-error:0.0575\n",
      "[118]\ttrain-error:0.043125\teval-error:0.058\n",
      "[119]\ttrain-error:0.043\teval-error:0.0575\n",
      "[120]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[121]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[122]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[123]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[124]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[125]\ttrain-error:0.043\teval-error:0.0575\n",
      "[126]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[127]\ttrain-error:0.043\teval-error:0.0575\n",
      "[128]\ttrain-error:0.043\teval-error:0.0575\n",
      "[129]\ttrain-error:0.043\teval-error:0.0565\n",
      "[130]\ttrain-error:0.043\teval-error:0.0575\n",
      "[131]\ttrain-error:0.043\teval-error:0.0575\n",
      "[132]\ttrain-error:0.043\teval-error:0.0575\n",
      "[133]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[134]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[135]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[136]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[137]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[138]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[139]\ttrain-error:0.043\teval-error:0.0575\n",
      "[140]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[141]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[142]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[143]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[144]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[145]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[146]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[147]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[148]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[149]\ttrain-error:0.043\teval-error:0.0575\n",
      "[150]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[151]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[152]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[153]\ttrain-error:0.04275\teval-error:0.0575\n",
      "[154]\ttrain-error:0.04275\teval-error:0.0575\n",
      "[155]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[156]\ttrain-error:0.043\teval-error:0.0575\n",
      "[157]\ttrain-error:0.043\teval-error:0.0575\n",
      "[158]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[159]\ttrain-error:0.043\teval-error:0.0575\n",
      "[160]\ttrain-error:0.043\teval-error:0.0575\n",
      "[161]\ttrain-error:0.043\teval-error:0.0575\n",
      "[162]\ttrain-error:0.043\teval-error:0.0575\n",
      "[163]\ttrain-error:0.043\teval-error:0.0575\n",
      "[164]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[165]\ttrain-error:0.043125\teval-error:0.0575\n",
      "[166]\ttrain-error:0.043\teval-error:0.0575\n",
      "[167]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[168]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[169]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[170]\ttrain-error:0.04275\teval-error:0.0575\n",
      "[171]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[172]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[173]\ttrain-error:0.042875\teval-error:0.0575\n",
      "[174]\ttrain-error:0.042875\teval-error:0.057\n",
      "[175]\ttrain-error:0.042875\teval-error:0.057\n",
      "[176]\ttrain-error:0.043\teval-error:0.0565\n",
      "[177]\ttrain-error:0.042875\teval-error:0.0565\n",
      "[178]\ttrain-error:0.042875\teval-error:0.0565\n",
      "[179]\ttrain-error:0.042875\teval-error:0.0565\n",
      "[180]\ttrain-error:0.043\teval-error:0.056\n",
      "[181]\ttrain-error:0.043\teval-error:0.056\n",
      "[182]\ttrain-error:0.043\teval-error:0.056\n",
      "[183]\ttrain-error:0.043\teval-error:0.056\n",
      "[184]\ttrain-error:0.043\teval-error:0.056\n",
      "[185]\ttrain-error:0.043\teval-error:0.056\n",
      "[186]\ttrain-error:0.043\teval-error:0.056\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[187]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[188]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[189]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[190]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[191]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[192]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[193]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[194]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[195]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[196]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[197]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[198]\ttrain-error:0.042875\teval-error:0.0555\n",
      "[199]\ttrain-error:0.042625\teval-error:0.0555\n",
      "[200]\ttrain-error:0.042625\teval-error:0.0555\n",
      "[201]\ttrain-error:0.042625\teval-error:0.0555\n",
      "[202]\ttrain-error:0.0425\teval-error:0.0555\n",
      "[203]\ttrain-error:0.0425\teval-error:0.0555\n",
      "[204]\ttrain-error:0.0425\teval-error:0.0555\n",
      "[205]\ttrain-error:0.0425\teval-error:0.0555\n",
      "[206]\ttrain-error:0.042375\teval-error:0.0555\n",
      "[207]\ttrain-error:0.04225\teval-error:0.056\n",
      "[208]\ttrain-error:0.04225\teval-error:0.056\n",
      "[209]\ttrain-error:0.04225\teval-error:0.056\n",
      "[210]\ttrain-error:0.04225\teval-error:0.056\n",
      "[211]\ttrain-error:0.042125\teval-error:0.056\n",
      "[212]\ttrain-error:0.042125\teval-error:0.056\n",
      "[213]\ttrain-error:0.042125\teval-error:0.056\n",
      "[214]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[215]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[216]\ttrain-error:0.042125\teval-error:0.056\n",
      "[217]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[218]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[219]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[220]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[221]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[222]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[223]\ttrain-error:0.04225\teval-error:0.0565\n",
      "[224]\ttrain-error:0.042125\teval-error:0.0565\n",
      "[225]\ttrain-error:0.042125\teval-error:0.0565\n",
      "[226]\ttrain-error:0.042125\teval-error:0.0565\n",
      "[227]\ttrain-error:0.042125\teval-error:0.0565\n",
      "[228]\ttrain-error:0.042\teval-error:0.056\n",
      "[229]\ttrain-error:0.042\teval-error:0.056\n",
      "[230]\ttrain-error:0.041875\teval-error:0.0555\n",
      "[231]\ttrain-error:0.041875\teval-error:0.0555\n",
      "[232]\ttrain-error:0.041875\teval-error:0.0555\n",
      "[233]\ttrain-error:0.041875\teval-error:0.0555\n",
      "[234]\ttrain-error:0.041875\teval-error:0.0555\n",
      "[235]\ttrain-error:0.042\teval-error:0.056\n",
      "[236]\ttrain-error:0.042\teval-error:0.056\n",
      "[237]\ttrain-error:0.042\teval-error:0.056\n",
      "[238]\ttrain-error:0.041875\teval-error:0.0555\n",
      "[239]\ttrain-error:0.041875\teval-error:0.0555\n",
      "[240]\ttrain-error:0.042\teval-error:0.0555\n",
      "[241]\ttrain-error:0.042\teval-error:0.0555\n",
      "[242]\ttrain-error:0.042125\teval-error:0.056\n",
      "[243]\ttrain-error:0.042125\teval-error:0.056\n",
      "[244]\ttrain-error:0.042\teval-error:0.056\n",
      "[245]\ttrain-error:0.042\teval-error:0.0565\n",
      "[246]\ttrain-error:0.042\teval-error:0.0565\n",
      "[247]\ttrain-error:0.042\teval-error:0.0565\n",
      "[248]\ttrain-error:0.042\teval-error:0.0565\n",
      "[249]\ttrain-error:0.04175\teval-error:0.056\n",
      "[250]\ttrain-error:0.04175\teval-error:0.056\n",
      "[251]\ttrain-error:0.042\teval-error:0.056\n",
      "[252]\ttrain-error:0.041625\teval-error:0.056\n",
      "[253]\ttrain-error:0.04175\teval-error:0.056\n",
      "[254]\ttrain-error:0.0415\teval-error:0.056\n",
      "[255]\ttrain-error:0.041625\teval-error:0.056\n",
      "[256]\ttrain-error:0.0415\teval-error:0.056\n",
      "[257]\ttrain-error:0.04175\teval-error:0.056\n",
      "[258]\ttrain-error:0.0415\teval-error:0.056\n",
      "[259]\ttrain-error:0.041375\teval-error:0.056\n",
      "[260]\ttrain-error:0.0415\teval-error:0.056\n",
      "[261]\ttrain-error:0.041625\teval-error:0.056\n",
      "[262]\ttrain-error:0.0415\teval-error:0.056\n",
      "[263]\ttrain-error:0.0415\teval-error:0.056\n",
      "[264]\ttrain-error:0.041625\teval-error:0.056\n",
      "[265]\ttrain-error:0.041625\teval-error:0.056\n",
      "[266]\ttrain-error:0.041625\teval-error:0.056\n",
      "[267]\ttrain-error:0.041625\teval-error:0.056\n",
      "[268]\ttrain-error:0.041625\teval-error:0.056\n",
      "[269]\ttrain-error:0.04175\teval-error:0.056\n",
      "[270]\ttrain-error:0.041875\teval-error:0.056\n",
      "[271]\ttrain-error:0.041875\teval-error:0.056\n",
      "[272]\ttrain-error:0.041875\teval-error:0.056\n",
      "[273]\ttrain-error:0.041875\teval-error:0.056\n",
      "[274]\ttrain-error:0.041875\teval-error:0.056\n",
      "[275]\ttrain-error:0.041875\teval-error:0.056\n",
      "[276]\ttrain-error:0.041875\teval-error:0.056\n",
      "[277]\ttrain-error:0.04175\teval-error:0.056\n",
      "[278]\ttrain-error:0.042\teval-error:0.056\n",
      "[279]\ttrain-error:0.041625\teval-error:0.0555\n",
      "[280]\ttrain-error:0.04175\teval-error:0.0555\n",
      "[281]\ttrain-error:0.041625\teval-error:0.0545\n",
      "[282]\ttrain-error:0.0415\teval-error:0.054\n",
      "[283]\ttrain-error:0.0415\teval-error:0.054\n",
      "[284]\ttrain-error:0.041625\teval-error:0.054\n",
      "[285]\ttrain-error:0.041625\teval-error:0.054\n",
      "[286]\ttrain-error:0.041625\teval-error:0.054\n",
      "[287]\ttrain-error:0.041625\teval-error:0.054\n",
      "[288]\ttrain-error:0.041375\teval-error:0.054\n",
      "[289]\ttrain-error:0.041375\teval-error:0.054\n",
      "[290]\ttrain-error:0.041375\teval-error:0.054\n",
      "[291]\ttrain-error:0.041375\teval-error:0.054\n",
      "[292]\ttrain-error:0.04125\teval-error:0.0535\n",
      "[293]\ttrain-error:0.041375\teval-error:0.0535\n",
      "[294]\ttrain-error:0.041375\teval-error:0.0535\n",
      "[295]\ttrain-error:0.041375\teval-error:0.053\n",
      "[296]\ttrain-error:0.041375\teval-error:0.053\n",
      "[297]\ttrain-error:0.041375\teval-error:0.053\n",
      "[298]\ttrain-error:0.041375\teval-error:0.053\n",
      "[299]\ttrain-error:0.041375\teval-error:0.053\n",
      "[300]\ttrain-error:0.041375\teval-error:0.053\n",
      "[301]\ttrain-error:0.041375\teval-error:0.053\n",
      "[302]\ttrain-error:0.041375\teval-error:0.053\n",
      "[303]\ttrain-error:0.041375\teval-error:0.053\n",
      "[304]\ttrain-error:0.041375\teval-error:0.053\n",
      "[305]\ttrain-error:0.041375\teval-error:0.053\n",
      "[306]\ttrain-error:0.041375\teval-error:0.053\n",
      "[307]\ttrain-error:0.041\teval-error:0.053\n",
      "[308]\ttrain-error:0.041\teval-error:0.053\n",
      "[309]\ttrain-error:0.041\teval-error:0.053\n",
      "[310]\ttrain-error:0.041\teval-error:0.053\n",
      "[311]\ttrain-error:0.041\teval-error:0.053\n",
      "[312]\ttrain-error:0.041\teval-error:0.053\n",
      "[313]\ttrain-error:0.041\teval-error:0.053\n",
      "[314]\ttrain-error:0.041\teval-error:0.053\n",
      "[315]\ttrain-error:0.041\teval-error:0.053\n",
      "[316]\ttrain-error:0.041\teval-error:0.053\n",
      "[317]\ttrain-error:0.041\teval-error:0.053\n",
      "[318]\ttrain-error:0.041\teval-error:0.053\n",
      "[319]\ttrain-error:0.04075\teval-error:0.053\n",
      "[320]\ttrain-error:0.04075\teval-error:0.053\n",
      "[321]\ttrain-error:0.04075\teval-error:0.053\n",
      "[322]\ttrain-error:0.04075\teval-error:0.053\n",
      "[323]\ttrain-error:0.04075\teval-error:0.053\n",
      "[324]\ttrain-error:0.04075\teval-error:0.053\n",
      "[325]\ttrain-error:0.04075\teval-error:0.053\n",
      "[326]\ttrain-error:0.0405\teval-error:0.053\n",
      "[327]\ttrain-error:0.040375\teval-error:0.053\n",
      "[328]\ttrain-error:0.040375\teval-error:0.053\n",
      "[329]\ttrain-error:0.04025\teval-error:0.053\n",
      "[330]\ttrain-error:0.04025\teval-error:0.053\n",
      "[331]\ttrain-error:0.04025\teval-error:0.053\n",
      "[332]\ttrain-error:0.040125\teval-error:0.053\n",
      "[333]\ttrain-error:0.040125\teval-error:0.053\n",
      "[334]\ttrain-error:0.040125\teval-error:0.053\n",
      "[335]\ttrain-error:0.039875\teval-error:0.053\n",
      "[336]\ttrain-error:0.039875\teval-error:0.053\n",
      "[337]\ttrain-error:0.039875\teval-error:0.053\n",
      "[338]\ttrain-error:0.03975\teval-error:0.053\n",
      "[339]\ttrain-error:0.039625\teval-error:0.053\n",
      "[340]\ttrain-error:0.039625\teval-error:0.053\n",
      "[341]\ttrain-error:0.0395\teval-error:0.053\n",
      "[342]\ttrain-error:0.039625\teval-error:0.053\n",
      "[343]\ttrain-error:0.039625\teval-error:0.053\n",
      "[344]\ttrain-error:0.039625\teval-error:0.053\n",
      "[345]\ttrain-error:0.039625\teval-error:0.053\n",
      "[346]\ttrain-error:0.039625\teval-error:0.053\n",
      "[347]\ttrain-error:0.039625\teval-error:0.053\n",
      "[348]\ttrain-error:0.0395\teval-error:0.053\n",
      "[349]\ttrain-error:0.0395\teval-error:0.053\n",
      "[350]\ttrain-error:0.0395\teval-error:0.053\n",
      "[351]\ttrain-error:0.0395\teval-error:0.053\n",
      "[352]\ttrain-error:0.0395\teval-error:0.053\n",
      "[353]\ttrain-error:0.039375\teval-error:0.053\n",
      "[354]\ttrain-error:0.039375\teval-error:0.053\n",
      "[355]\ttrain-error:0.039375\teval-error:0.053\n",
      "[356]\ttrain-error:0.039375\teval-error:0.053\n",
      "[357]\ttrain-error:0.039375\teval-error:0.053\n",
      "[358]\ttrain-error:0.039375\teval-error:0.053\n",
      "[359]\ttrain-error:0.03925\teval-error:0.052\n",
      "[360]\ttrain-error:0.03925\teval-error:0.052\n",
      "[361]\ttrain-error:0.03925\teval-error:0.052\n",
      "[362]\ttrain-error:0.03925\teval-error:0.052\n",
      "[363]\ttrain-error:0.03925\teval-error:0.052\n",
      "[364]\ttrain-error:0.03925\teval-error:0.052\n",
      "[365]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[366]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[367]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[368]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[369]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[370]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[371]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[372]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[373]\ttrain-error:0.039375\teval-error:0.0515\n",
      "[374]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[375]\ttrain-error:0.03925\teval-error:0.0515\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[376]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[377]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[378]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[379]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[380]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[381]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[382]\ttrain-error:0.039125\teval-error:0.0515\n",
      "[383]\ttrain-error:0.03925\teval-error:0.0515\n",
      "[384]\ttrain-error:0.039125\teval-error:0.0515\n",
      "[385]\ttrain-error:0.039\teval-error:0.0515\n",
      "[386]\ttrain-error:0.039125\teval-error:0.0515\n",
      "[387]\ttrain-error:0.039\teval-error:0.0515\n",
      "[388]\ttrain-error:0.039\teval-error:0.0515\n",
      "[389]\ttrain-error:0.039\teval-error:0.0515\n",
      "[390]\ttrain-error:0.039\teval-error:0.0515\n",
      "[391]\ttrain-error:0.039\teval-error:0.0515\n",
      "[392]\ttrain-error:0.039\teval-error:0.0515\n",
      "[393]\ttrain-error:0.039\teval-error:0.0515\n",
      "[394]\ttrain-error:0.039\teval-error:0.0515\n",
      "[395]\ttrain-error:0.039\teval-error:0.0515\n",
      "[396]\ttrain-error:0.038875\teval-error:0.051\n",
      "[397]\ttrain-error:0.039\teval-error:0.051\n",
      "[398]\ttrain-error:0.039\teval-error:0.051\n",
      "[399]\ttrain-error:0.039\teval-error:0.051\n",
      "[400]\ttrain-error:0.039\teval-error:0.051\n",
      "[401]\ttrain-error:0.038875\teval-error:0.051\n",
      "[402]\ttrain-error:0.038875\teval-error:0.051\n",
      "[403]\ttrain-error:0.03875\teval-error:0.051\n",
      "[404]\ttrain-error:0.03875\teval-error:0.051\n",
      "[405]\ttrain-error:0.03875\teval-error:0.051\n",
      "[406]\ttrain-error:0.03875\teval-error:0.051\n",
      "[407]\ttrain-error:0.03875\teval-error:0.051\n",
      "[408]\ttrain-error:0.03875\teval-error:0.051\n",
      "[409]\ttrain-error:0.038875\teval-error:0.051\n",
      "[410]\ttrain-error:0.03875\teval-error:0.051\n",
      "[411]\ttrain-error:0.03875\teval-error:0.051\n",
      "[412]\ttrain-error:0.038625\teval-error:0.051\n",
      "[413]\ttrain-error:0.038625\teval-error:0.051\n",
      "[414]\ttrain-error:0.038625\teval-error:0.051\n",
      "[415]\ttrain-error:0.0385\teval-error:0.051\n",
      "[416]\ttrain-error:0.0385\teval-error:0.051\n",
      "[417]\ttrain-error:0.0385\teval-error:0.051\n",
      "[418]\ttrain-error:0.0385\teval-error:0.051\n",
      "[419]\ttrain-error:0.0385\teval-error:0.051\n",
      "[420]\ttrain-error:0.0385\teval-error:0.051\n",
      "[421]\ttrain-error:0.0385\teval-error:0.051\n",
      "[422]\ttrain-error:0.0385\teval-error:0.051\n",
      "[423]\ttrain-error:0.0385\teval-error:0.051\n",
      "[424]\ttrain-error:0.0385\teval-error:0.051\n",
      "[425]\ttrain-error:0.038375\teval-error:0.051\n",
      "[426]\ttrain-error:0.038375\teval-error:0.051\n",
      "[427]\ttrain-error:0.03825\teval-error:0.051\n",
      "[428]\ttrain-error:0.03825\teval-error:0.051\n",
      "[429]\ttrain-error:0.03825\teval-error:0.051\n",
      "[430]\ttrain-error:0.03825\teval-error:0.051\n",
      "[431]\ttrain-error:0.03825\teval-error:0.051\n",
      "[432]\ttrain-error:0.03825\teval-error:0.051\n",
      "[433]\ttrain-error:0.03825\teval-error:0.051\n",
      "[434]\ttrain-error:0.038125\teval-error:0.051\n",
      "[435]\ttrain-error:0.038125\teval-error:0.051\n",
      "[436]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[437]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[438]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[439]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[440]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[441]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[442]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[443]\ttrain-error:0.038125\teval-error:0.0505\n",
      "[444]\ttrain-error:0.038\teval-error:0.0505\n",
      "[445]\ttrain-error:0.038\teval-error:0.0505\n",
      "[446]\ttrain-error:0.038\teval-error:0.0505\n",
      "[447]\ttrain-error:0.038\teval-error:0.05\n",
      "[448]\ttrain-error:0.038\teval-error:0.05\n",
      "[449]\ttrain-error:0.038\teval-error:0.05\n",
      "[450]\ttrain-error:0.037875\teval-error:0.05\n",
      "[451]\ttrain-error:0.037875\teval-error:0.05\n",
      "[452]\ttrain-error:0.037875\teval-error:0.05\n",
      "[453]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[454]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[455]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[456]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[457]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[458]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[459]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[460]\ttrain-error:0.037875\teval-error:0.0495\n",
      "[461]\ttrain-error:0.037625\teval-error:0.0495\n",
      "[462]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[463]\ttrain-error:0.037625\teval-error:0.0495\n",
      "[464]\ttrain-error:0.037625\teval-error:0.0495\n",
      "[465]\ttrain-error:0.037625\teval-error:0.0495\n",
      "[466]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[467]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[468]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[469]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[470]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[471]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[472]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[473]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[474]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[475]\ttrain-error:0.03775\teval-error:0.0495\n",
      "[476]\ttrain-error:0.037625\teval-error:0.0495\n",
      "[477]\ttrain-error:0.037625\teval-error:0.0495\n",
      "[478]\ttrain-error:0.0375\teval-error:0.0495\n",
      "[479]\ttrain-error:0.0375\teval-error:0.0495\n",
      "[480]\ttrain-error:0.0375\teval-error:0.0495\n",
      "[481]\ttrain-error:0.0375\teval-error:0.0495\n",
      "[482]\ttrain-error:0.03725\teval-error:0.05\n",
      "[483]\ttrain-error:0.03725\teval-error:0.05\n",
      "[484]\ttrain-error:0.03725\teval-error:0.05\n",
      "[485]\ttrain-error:0.03725\teval-error:0.05\n",
      "[486]\ttrain-error:0.03725\teval-error:0.05\n",
      "[487]\ttrain-error:0.03725\teval-error:0.05\n",
      "[488]\ttrain-error:0.03725\teval-error:0.05\n",
      "[489]\ttrain-error:0.03725\teval-error:0.05\n",
      "[490]\ttrain-error:0.03725\teval-error:0.05\n",
      "[491]\ttrain-error:0.037125\teval-error:0.05\n",
      "[492]\ttrain-error:0.037125\teval-error:0.05\n",
      "[493]\ttrain-error:0.037125\teval-error:0.05\n",
      "[494]\ttrain-error:0.037125\teval-error:0.05\n",
      "[495]\ttrain-error:0.037125\teval-error:0.05\n",
      "[496]\ttrain-error:0.03725\teval-error:0.05\n",
      "[497]\ttrain-error:0.03725\teval-error:0.05\n",
      "[498]\ttrain-error:0.03725\teval-error:0.05\n",
      "[499]\ttrain-error:0.03725\teval-error:0.05\n",
      "[500]\ttrain-error:0.03725\teval-error:0.05\n",
      "[501]\ttrain-error:0.03725\teval-error:0.05\n",
      "[502]\ttrain-error:0.03725\teval-error:0.05\n",
      "[503]\ttrain-error:0.03725\teval-error:0.05\n",
      "[504]\ttrain-error:0.03725\teval-error:0.05\n",
      "[505]\ttrain-error:0.03725\teval-error:0.05\n",
      "[506]\ttrain-error:0.03725\teval-error:0.05\n",
      "[507]\ttrain-error:0.03725\teval-error:0.05\n",
      "[508]\ttrain-error:0.03725\teval-error:0.05\n",
      "[509]\ttrain-error:0.03725\teval-error:0.05\n",
      "[510]\ttrain-error:0.037375\teval-error:0.05\n",
      "[511]\ttrain-error:0.037375\teval-error:0.05\n",
      "[512]\ttrain-error:0.037375\teval-error:0.05\n",
      "[513]\ttrain-error:0.037375\teval-error:0.05\n",
      "[514]\ttrain-error:0.037375\teval-error:0.05\n",
      "[515]\ttrain-error:0.037375\teval-error:0.05\n",
      "[516]\ttrain-error:0.037375\teval-error:0.05\n",
      "[517]\ttrain-error:0.037375\teval-error:0.05\n",
      "[518]\ttrain-error:0.03725\teval-error:0.0505\n",
      "[519]\ttrain-error:0.037375\teval-error:0.0505\n",
      "[520]\ttrain-error:0.03725\teval-error:0.0505\n",
      "[521]\ttrain-error:0.03725\teval-error:0.0505\n",
      "[522]\ttrain-error:0.037375\teval-error:0.0505\n",
      "[523]\ttrain-error:0.037375\teval-error:0.0505\n",
      "[524]\ttrain-error:0.03725\teval-error:0.0505\n",
      "[525]\ttrain-error:0.037125\teval-error:0.0505\n",
      "[526]\ttrain-error:0.03725\teval-error:0.0505\n",
      "[527]\ttrain-error:0.037125\teval-error:0.0505\n",
      "[528]\ttrain-error:0.037125\teval-error:0.0505\n",
      "[529]\ttrain-error:0.037125\teval-error:0.05\n",
      "[530]\ttrain-error:0.037125\teval-error:0.05\n",
      "[531]\ttrain-error:0.037125\teval-error:0.05\n",
      "[532]\ttrain-error:0.037125\teval-error:0.05\n",
      "[533]\ttrain-error:0.037125\teval-error:0.05\n",
      "[534]\ttrain-error:0.037125\teval-error:0.05\n",
      "[535]\ttrain-error:0.037125\teval-error:0.05\n",
      "[536]\ttrain-error:0.037\teval-error:0.05\n",
      "[537]\ttrain-error:0.037125\teval-error:0.05\n",
      "[538]\ttrain-error:0.037\teval-error:0.0505\n",
      "[539]\ttrain-error:0.037\teval-error:0.0505\n",
      "[540]\ttrain-error:0.037\teval-error:0.0505\n",
      "[541]\ttrain-error:0.037\teval-error:0.0505\n",
      "[542]\ttrain-error:0.037\teval-error:0.0505\n",
      "[543]\ttrain-error:0.037\teval-error:0.0505\n",
      "[544]\ttrain-error:0.037\teval-error:0.0505\n",
      "[545]\ttrain-error:0.037\teval-error:0.0505\n",
      "[546]\ttrain-error:0.037\teval-error:0.0505\n",
      "[547]\ttrain-error:0.037125\teval-error:0.0505\n",
      "[548]\ttrain-error:0.037\teval-error:0.0505\n",
      "[549]\ttrain-error:0.037\teval-error:0.0505\n",
      "[550]\ttrain-error:0.037\teval-error:0.0505\n",
      "[551]\ttrain-error:0.037\teval-error:0.0505\n",
      "[552]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[553]\ttrain-error:0.037\teval-error:0.0505\n",
      "[554]\ttrain-error:0.037\teval-error:0.0505\n",
      "[555]\ttrain-error:0.037\teval-error:0.0505\n",
      "[556]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[557]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[558]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[559]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[560]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[561]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[562]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[563]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[564]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[565]\ttrain-error:0.036875\teval-error:0.0505\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[566]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[567]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[568]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[569]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[570]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[571]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[572]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[573]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[574]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[575]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[576]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[577]\ttrain-error:0.036875\teval-error:0.0505\n",
      "[578]\ttrain-error:0.036625\teval-error:0.0505\n",
      "[579]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[580]\ttrain-error:0.03675\teval-error:0.0505\n",
      "[581]\ttrain-error:0.036625\teval-error:0.0505\n",
      "[582]\ttrain-error:0.036625\teval-error:0.0505\n",
      "[583]\ttrain-error:0.0365\teval-error:0.0505\n",
      "[584]\ttrain-error:0.0365\teval-error:0.0505\n",
      "[585]\ttrain-error:0.0365\teval-error:0.0505\n",
      "[586]\ttrain-error:0.0365\teval-error:0.0505\n",
      "[587]\ttrain-error:0.0365\teval-error:0.0505\n",
      "[588]\ttrain-error:0.0365\teval-error:0.0505\n",
      "[589]\ttrain-error:0.036375\teval-error:0.0505\n",
      "[590]\ttrain-error:0.036375\teval-error:0.0505\n",
      "[591]\ttrain-error:0.03625\teval-error:0.0505\n",
      "[592]\ttrain-error:0.03625\teval-error:0.0505\n",
      "[593]\ttrain-error:0.03625\teval-error:0.0505\n",
      "[594]\ttrain-error:0.036125\teval-error:0.0505\n",
      "[595]\ttrain-error:0.036125\teval-error:0.0505\n",
      "[596]\ttrain-error:0.036125\teval-error:0.0505\n",
      "[597]\ttrain-error:0.036\teval-error:0.0505\n",
      "[598]\ttrain-error:0.036\teval-error:0.0505\n",
      "[599]\ttrain-error:0.036125\teval-error:0.0505\n",
      "[600]\ttrain-error:0.036125\teval-error:0.0505\n",
      "[601]\ttrain-error:0.036\teval-error:0.0505\n",
      "[602]\ttrain-error:0.036\teval-error:0.0505\n",
      "[603]\ttrain-error:0.036\teval-error:0.0495\n",
      "[604]\ttrain-error:0.036\teval-error:0.0495\n",
      "[605]\ttrain-error:0.036\teval-error:0.0495\n",
      "[606]\ttrain-error:0.036125\teval-error:0.0495\n",
      "[607]\ttrain-error:0.036125\teval-error:0.0495\n",
      "[608]\ttrain-error:0.036\teval-error:0.0485\n",
      "[609]\ttrain-error:0.036\teval-error:0.0485\n",
      "[610]\ttrain-error:0.036\teval-error:0.049\n",
      "[611]\ttrain-error:0.036\teval-error:0.049\n",
      "[612]\ttrain-error:0.036125\teval-error:0.049\n",
      "[613]\ttrain-error:0.036\teval-error:0.049\n",
      "[614]\ttrain-error:0.036125\teval-error:0.049\n",
      "[615]\ttrain-error:0.036125\teval-error:0.0485\n",
      "[616]\ttrain-error:0.03625\teval-error:0.0485\n",
      "[617]\ttrain-error:0.036125\teval-error:0.0485\n",
      "[618]\ttrain-error:0.036125\teval-error:0.0485\n",
      "[619]\ttrain-error:0.03625\teval-error:0.0485\n",
      "[620]\ttrain-error:0.03625\teval-error:0.0485\n",
      "[621]\ttrain-error:0.03625\teval-error:0.0485\n",
      "[622]\ttrain-error:0.03625\teval-error:0.048\n",
      "[623]\ttrain-error:0.036125\teval-error:0.048\n",
      "[624]\ttrain-error:0.036125\teval-error:0.048\n",
      "[625]\ttrain-error:0.036125\teval-error:0.048\n",
      "[626]\ttrain-error:0.036125\teval-error:0.048\n",
      "[627]\ttrain-error:0.036125\teval-error:0.048\n",
      "[628]\ttrain-error:0.036125\teval-error:0.048\n",
      "[629]\ttrain-error:0.036125\teval-error:0.048\n",
      "[630]\ttrain-error:0.036125\teval-error:0.048\n",
      "[631]\ttrain-error:0.036125\teval-error:0.048\n",
      "[632]\ttrain-error:0.03625\teval-error:0.048\n",
      "[633]\ttrain-error:0.03625\teval-error:0.048\n",
      "[634]\ttrain-error:0.036\teval-error:0.048\n",
      "[635]\ttrain-error:0.036125\teval-error:0.048\n",
      "[636]\ttrain-error:0.036125\teval-error:0.048\n",
      "[637]\ttrain-error:0.036125\teval-error:0.048\n",
      "[638]\ttrain-error:0.036125\teval-error:0.048\n",
      "[639]\ttrain-error:0.036125\teval-error:0.048\n",
      "[640]\ttrain-error:0.036125\teval-error:0.048\n",
      "[641]\ttrain-error:0.036125\teval-error:0.048\n",
      "[642]\ttrain-error:0.036125\teval-error:0.048\n",
      "[643]\ttrain-error:0.036125\teval-error:0.048\n",
      "[644]\ttrain-error:0.036125\teval-error:0.048\n",
      "[645]\ttrain-error:0.036125\teval-error:0.048\n",
      "[646]\ttrain-error:0.036125\teval-error:0.048\n",
      "[647]\ttrain-error:0.036125\teval-error:0.048\n",
      "[648]\ttrain-error:0.036125\teval-error:0.048\n",
      "[649]\ttrain-error:0.036125\teval-error:0.048\n",
      "[650]\ttrain-error:0.036\teval-error:0.048\n",
      "[651]\ttrain-error:0.036125\teval-error:0.048\n",
      "[652]\ttrain-error:0.036125\teval-error:0.048\n",
      "[653]\ttrain-error:0.035875\teval-error:0.048\n",
      "[654]\ttrain-error:0.036\teval-error:0.048\n",
      "[655]\ttrain-error:0.035875\teval-error:0.048\n",
      "[656]\ttrain-error:0.035875\teval-error:0.048\n",
      "[657]\ttrain-error:0.035875\teval-error:0.048\n",
      "[658]\ttrain-error:0.035875\teval-error:0.0475\n",
      "[659]\ttrain-error:0.035875\teval-error:0.0475\n",
      "[660]\ttrain-error:0.035875\teval-error:0.0475\n",
      "[661]\ttrain-error:0.03575\teval-error:0.0475\n",
      "[662]\ttrain-error:0.03575\teval-error:0.0475\n",
      "[663]\ttrain-error:0.035875\teval-error:0.0475\n",
      "[664]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[665]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[666]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[667]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[668]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[669]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[670]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[671]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[672]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[673]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[674]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[675]\ttrain-error:0.035625\teval-error:0.0475\n",
      "[676]\ttrain-error:0.0355\teval-error:0.0475\n",
      "[677]\ttrain-error:0.0355\teval-error:0.0475\n",
      "[678]\ttrain-error:0.0355\teval-error:0.0475\n",
      "[679]\ttrain-error:0.0355\teval-error:0.0475\n",
      "[680]\ttrain-error:0.035375\teval-error:0.0475\n",
      "[681]\ttrain-error:0.035375\teval-error:0.0475\n",
      "[682]\ttrain-error:0.035375\teval-error:0.047\n",
      "[683]\ttrain-error:0.035375\teval-error:0.047\n",
      "[684]\ttrain-error:0.035375\teval-error:0.047\n",
      "[685]\ttrain-error:0.035375\teval-error:0.047\n",
      "[686]\ttrain-error:0.03525\teval-error:0.047\n",
      "[687]\ttrain-error:0.03525\teval-error:0.047\n",
      "[688]\ttrain-error:0.035\teval-error:0.047\n",
      "[689]\ttrain-error:0.035\teval-error:0.047\n",
      "[690]\ttrain-error:0.035\teval-error:0.047\n",
      "[691]\ttrain-error:0.035\teval-error:0.047\n",
      "[692]\ttrain-error:0.035\teval-error:0.047\n",
      "[693]\ttrain-error:0.035\teval-error:0.047\n",
      "[694]\ttrain-error:0.035\teval-error:0.047\n",
      "[695]\ttrain-error:0.035\teval-error:0.047\n",
      "[696]\ttrain-error:0.034875\teval-error:0.047\n",
      "[697]\ttrain-error:0.035\teval-error:0.047\n",
      "[698]\ttrain-error:0.035\teval-error:0.047\n",
      "[699]\ttrain-error:0.035\teval-error:0.0465\n",
      "[700]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[701]\ttrain-error:0.035\teval-error:0.0465\n",
      "[702]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[703]\ttrain-error:0.035\teval-error:0.0465\n",
      "[704]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[705]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[706]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[707]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[708]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[709]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[710]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[711]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[712]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[713]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[714]\ttrain-error:0.034625\teval-error:0.0465\n",
      "[715]\ttrain-error:0.034875\teval-error:0.0465\n",
      "[716]\ttrain-error:0.034625\teval-error:0.047\n",
      "[717]\ttrain-error:0.03475\teval-error:0.047\n",
      "[718]\ttrain-error:0.03475\teval-error:0.047\n",
      "[719]\ttrain-error:0.034375\teval-error:0.047\n",
      "[720]\ttrain-error:0.0345\teval-error:0.047\n",
      "[721]\ttrain-error:0.034375\teval-error:0.047\n",
      "[722]\ttrain-error:0.034375\teval-error:0.047\n",
      "[723]\ttrain-error:0.03425\teval-error:0.047\n",
      "[724]\ttrain-error:0.034375\teval-error:0.047\n",
      "[725]\ttrain-error:0.03425\teval-error:0.047\n",
      "[726]\ttrain-error:0.03425\teval-error:0.047\n",
      "[727]\ttrain-error:0.03425\teval-error:0.047\n",
      "[728]\ttrain-error:0.03425\teval-error:0.047\n",
      "[729]\ttrain-error:0.03425\teval-error:0.047\n",
      "[730]\ttrain-error:0.03425\teval-error:0.047\n",
      "[731]\ttrain-error:0.03425\teval-error:0.047\n",
      "[732]\ttrain-error:0.034125\teval-error:0.047\n",
      "[733]\ttrain-error:0.034125\teval-error:0.047\n",
      "[734]\ttrain-error:0.034125\teval-error:0.047\n",
      "[735]\ttrain-error:0.034125\teval-error:0.047\n",
      "[736]\ttrain-error:0.034125\teval-error:0.047\n",
      "[737]\ttrain-error:0.034125\teval-error:0.047\n",
      "[738]\ttrain-error:0.034125\teval-error:0.047\n",
      "[739]\ttrain-error:0.034125\teval-error:0.047\n",
      "[740]\ttrain-error:0.034125\teval-error:0.047\n",
      "[741]\ttrain-error:0.034125\teval-error:0.047\n",
      "[742]\ttrain-error:0.034125\teval-error:0.047\n",
      "[743]\ttrain-error:0.034125\teval-error:0.047\n",
      "[744]\ttrain-error:0.034\teval-error:0.047\n",
      "[745]\ttrain-error:0.034\teval-error:0.047\n",
      "[746]\ttrain-error:0.034\teval-error:0.047\n",
      "[747]\ttrain-error:0.034\teval-error:0.047\n",
      "[748]\ttrain-error:0.033875\teval-error:0.047\n",
      "[749]\ttrain-error:0.033875\teval-error:0.047\n",
      "[750]\ttrain-error:0.033875\teval-error:0.047\n",
      "[751]\ttrain-error:0.033875\teval-error:0.047\n",
      "[752]\ttrain-error:0.033875\teval-error:0.047\n",
      "[753]\ttrain-error:0.03375\teval-error:0.047\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[754]\ttrain-error:0.03375\teval-error:0.047\n",
      "[755]\ttrain-error:0.03375\teval-error:0.047\n",
      "[756]\ttrain-error:0.03375\teval-error:0.047\n",
      "[757]\ttrain-error:0.03375\teval-error:0.047\n",
      "[758]\ttrain-error:0.03375\teval-error:0.047\n",
      "[759]\ttrain-error:0.033625\teval-error:0.047\n",
      "[760]\ttrain-error:0.0335\teval-error:0.047\n",
      "[761]\ttrain-error:0.033375\teval-error:0.047\n",
      "[762]\ttrain-error:0.033625\teval-error:0.047\n",
      "[763]\ttrain-error:0.0335\teval-error:0.047\n",
      "[764]\ttrain-error:0.033375\teval-error:0.047\n",
      "[765]\ttrain-error:0.0335\teval-error:0.047\n",
      "[766]\ttrain-error:0.03325\teval-error:0.047\n",
      "[767]\ttrain-error:0.03325\teval-error:0.047\n",
      "[768]\ttrain-error:0.03325\teval-error:0.047\n",
      "[769]\ttrain-error:0.03325\teval-error:0.047\n",
      "[770]\ttrain-error:0.03325\teval-error:0.047\n",
      "[771]\ttrain-error:0.03325\teval-error:0.047\n",
      "[772]\ttrain-error:0.033375\teval-error:0.047\n",
      "[773]\ttrain-error:0.033125\teval-error:0.047\n",
      "[774]\ttrain-error:0.032875\teval-error:0.047\n",
      "[775]\ttrain-error:0.033\teval-error:0.047\n",
      "[776]\ttrain-error:0.032875\teval-error:0.047\n",
      "[777]\ttrain-error:0.033\teval-error:0.047\n",
      "[778]\ttrain-error:0.03275\teval-error:0.047\n",
      "[779]\ttrain-error:0.033\teval-error:0.047\n",
      "[780]\ttrain-error:0.03275\teval-error:0.047\n",
      "[781]\ttrain-error:0.03275\teval-error:0.047\n",
      "[782]\ttrain-error:0.03275\teval-error:0.047\n",
      "[783]\ttrain-error:0.03275\teval-error:0.047\n",
      "[784]\ttrain-error:0.03275\teval-error:0.047\n",
      "[785]\ttrain-error:0.032875\teval-error:0.0465\n",
      "[786]\ttrain-error:0.03275\teval-error:0.0465\n",
      "[787]\ttrain-error:0.03275\teval-error:0.0465\n",
      "[788]\ttrain-error:0.032625\teval-error:0.0465\n",
      "[789]\ttrain-error:0.032625\teval-error:0.0465\n",
      "[790]\ttrain-error:0.032625\teval-error:0.0465\n",
      "[791]\ttrain-error:0.032625\teval-error:0.0465\n",
      "[792]\ttrain-error:0.032625\teval-error:0.0465\n",
      "[793]\ttrain-error:0.032625\teval-error:0.0465\n",
      "[794]\ttrain-error:0.032625\teval-error:0.047\n",
      "[795]\ttrain-error:0.032625\teval-error:0.047\n",
      "[796]\ttrain-error:0.032625\teval-error:0.0465\n",
      "[797]\ttrain-error:0.0325\teval-error:0.047\n",
      "[798]\ttrain-error:0.0325\teval-error:0.0465\n",
      "[799]\ttrain-error:0.0325\teval-error:0.0465\n",
      "[800]\ttrain-error:0.0325\teval-error:0.0465\n",
      "[801]\ttrain-error:0.0325\teval-error:0.0465\n",
      "[802]\ttrain-error:0.032375\teval-error:0.0465\n",
      "[803]\ttrain-error:0.032375\teval-error:0.0465\n",
      "[804]\ttrain-error:0.03225\teval-error:0.0465\n",
      "[805]\ttrain-error:0.03225\teval-error:0.0465\n",
      "[806]\ttrain-error:0.032125\teval-error:0.0465\n",
      "[807]\ttrain-error:0.032\teval-error:0.0465\n",
      "[808]\ttrain-error:0.032\teval-error:0.0465\n",
      "[809]\ttrain-error:0.032\teval-error:0.0465\n",
      "[810]\ttrain-error:0.032\teval-error:0.0465\n",
      "[811]\ttrain-error:0.03225\teval-error:0.0465\n",
      "[812]\ttrain-error:0.032\teval-error:0.0465\n",
      "[813]\ttrain-error:0.031875\teval-error:0.0465\n",
      "[814]\ttrain-error:0.031875\teval-error:0.0465\n",
      "[815]\ttrain-error:0.031875\teval-error:0.0465\n",
      "[816]\ttrain-error:0.03175\teval-error:0.0465\n",
      "[817]\ttrain-error:0.03175\teval-error:0.0465\n",
      "[818]\ttrain-error:0.03175\teval-error:0.0465\n",
      "[819]\ttrain-error:0.03175\teval-error:0.0465\n",
      "[820]\ttrain-error:0.03175\teval-error:0.0465\n",
      "[821]\ttrain-error:0.03175\teval-error:0.0465\n",
      "[822]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[823]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[824]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[825]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[826]\ttrain-error:0.031625\teval-error:0.046\n",
      "[827]\ttrain-error:0.031625\teval-error:0.046\n",
      "[828]\ttrain-error:0.031625\teval-error:0.046\n",
      "[829]\ttrain-error:0.03175\teval-error:0.046\n",
      "[830]\ttrain-error:0.03175\teval-error:0.046\n",
      "[831]\ttrain-error:0.03175\teval-error:0.046\n",
      "[832]\ttrain-error:0.03175\teval-error:0.046\n",
      "[833]\ttrain-error:0.03175\teval-error:0.046\n",
      "[834]\ttrain-error:0.03175\teval-error:0.046\n",
      "[835]\ttrain-error:0.03175\teval-error:0.046\n",
      "[836]\ttrain-error:0.03175\teval-error:0.046\n",
      "[837]\ttrain-error:0.031625\teval-error:0.046\n",
      "[838]\ttrain-error:0.031625\teval-error:0.046\n",
      "[839]\ttrain-error:0.031625\teval-error:0.046\n",
      "[840]\ttrain-error:0.031625\teval-error:0.046\n",
      "[841]\ttrain-error:0.031625\teval-error:0.046\n",
      "[842]\ttrain-error:0.031625\teval-error:0.046\n",
      "[843]\ttrain-error:0.0315\teval-error:0.046\n",
      "[844]\ttrain-error:0.0315\teval-error:0.046\n",
      "[845]\ttrain-error:0.0315\teval-error:0.046\n",
      "[846]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[847]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[848]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[849]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[850]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[851]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[852]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[853]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[854]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[855]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[856]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[857]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[858]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[859]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[860]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[861]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[862]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[863]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[864]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[865]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[866]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[867]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[868]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[869]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[870]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[871]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[872]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[873]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[874]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[875]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[876]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[877]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[878]\ttrain-error:0.031625\teval-error:0.0465\n",
      "[879]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[880]\ttrain-error:0.0315\teval-error:0.0465\n",
      "[881]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[882]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[883]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[884]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[885]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[886]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[887]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[888]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[889]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[890]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[891]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[892]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[893]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[894]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[895]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[896]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[897]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[898]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[899]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[900]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[901]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[902]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[903]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[904]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[905]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[906]\ttrain-error:0.031375\teval-error:0.0465\n",
      "[907]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[908]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[909]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[910]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[911]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[912]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[913]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[914]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[915]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[916]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[917]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[918]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[919]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[920]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[921]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[922]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[923]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[924]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[925]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[926]\ttrain-error:0.03125\teval-error:0.046\n",
      "[927]\ttrain-error:0.03125\teval-error:0.046\n",
      "[928]\ttrain-error:0.03125\teval-error:0.046\n",
      "[929]\ttrain-error:0.03125\teval-error:0.046\n",
      "[930]\ttrain-error:0.03125\teval-error:0.046\n",
      "[931]\ttrain-error:0.03125\teval-error:0.046\n",
      "[932]\ttrain-error:0.03125\teval-error:0.046\n",
      "[933]\ttrain-error:0.03125\teval-error:0.046\n",
      "[934]\ttrain-error:0.03125\teval-error:0.046\n",
      "[935]\ttrain-error:0.03125\teval-error:0.046\n",
      "[936]\ttrain-error:0.03125\teval-error:0.046\n",
      "[937]\ttrain-error:0.03125\teval-error:0.046\n",
      "[938]\ttrain-error:0.03125\teval-error:0.046\n",
      "[939]\ttrain-error:0.03125\teval-error:0.046\n",
      "[940]\ttrain-error:0.03125\teval-error:0.0465\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[941]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[942]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[943]\ttrain-error:0.03125\teval-error:0.0465\n",
      "[944]\ttrain-error:0.031125\teval-error:0.0465\n",
      "[945]\ttrain-error:0.031\teval-error:0.046\n",
      "[946]\ttrain-error:0.031\teval-error:0.046\n",
      "[947]\ttrain-error:0.031\teval-error:0.046\n",
      "[948]\ttrain-error:0.031\teval-error:0.046\n",
      "[949]\ttrain-error:0.031\teval-error:0.046\n",
      "[950]\ttrain-error:0.031\teval-error:0.046\n",
      "[951]\ttrain-error:0.031\teval-error:0.046\n",
      "[952]\ttrain-error:0.031\teval-error:0.046\n",
      "[953]\ttrain-error:0.031\teval-error:0.046\n",
      "[954]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[955]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[956]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[957]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[958]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[959]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[960]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[961]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[962]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[963]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[964]\ttrain-error:0.03125\teval-error:0.0455\n",
      "[965]\ttrain-error:0.03125\teval-error:0.045\n",
      "[966]\ttrain-error:0.03125\teval-error:0.045\n",
      "[967]\ttrain-error:0.031125\teval-error:0.045\n",
      "[968]\ttrain-error:0.031125\teval-error:0.045\n",
      "[969]\ttrain-error:0.031125\teval-error:0.045\n",
      "[970]\ttrain-error:0.031125\teval-error:0.045\n",
      "[971]\ttrain-error:0.031125\teval-error:0.045\n",
      "[972]\ttrain-error:0.031125\teval-error:0.045\n",
      "[973]\ttrain-error:0.031125\teval-error:0.045\n",
      "[974]\ttrain-error:0.031125\teval-error:0.045\n",
      "[975]\ttrain-error:0.031125\teval-error:0.045\n",
      "[976]\ttrain-error:0.031125\teval-error:0.045\n",
      "[977]\ttrain-error:0.031125\teval-error:0.045\n",
      "[978]\ttrain-error:0.031125\teval-error:0.045\n",
      "[979]\ttrain-error:0.031125\teval-error:0.045\n",
      "[980]\ttrain-error:0.031125\teval-error:0.045\n",
      "[981]\ttrain-error:0.031125\teval-error:0.045\n",
      "[982]\ttrain-error:0.031125\teval-error:0.045\n",
      "[983]\ttrain-error:0.031125\teval-error:0.045\n",
      "[984]\ttrain-error:0.031125\teval-error:0.045\n",
      "[985]\ttrain-error:0.031125\teval-error:0.045\n",
      "[986]\ttrain-error:0.031125\teval-error:0.045\n",
      "[987]\ttrain-error:0.031125\teval-error:0.045\n",
      "[988]\ttrain-error:0.031125\teval-error:0.045\n",
      "[989]\ttrain-error:0.031125\teval-error:0.045\n",
      "[990]\ttrain-error:0.031125\teval-error:0.045\n",
      "[991]\ttrain-error:0.031125\teval-error:0.045\n",
      "[992]\ttrain-error:0.031125\teval-error:0.045\n",
      "[993]\ttrain-error:0.031125\teval-error:0.045\n",
      "[994]\ttrain-error:0.031125\teval-error:0.045\n",
      "[995]\ttrain-error:0.031125\teval-error:0.045\n",
      "[996]\ttrain-error:0.031125\teval-error:0.045\n",
      "[997]\ttrain-error:0.031125\teval-error:0.045\n",
      "[998]\ttrain-error:0.031\teval-error:0.045\n",
      "[999]\ttrain-error:0.031\teval-error:0.045\n"
     ]
    }
   ],
   "source": [
    "xgb_model = xgb.train(params = params, dtrain = dtrain, num_boost_round = num_rounds,\n",
    "                     early_stopping_rounds = 1000, evals = wlist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.198 0.216 0.2   0.238 0.2   0.2   0.199 0.197 0.247 0.203]\n",
      "predict() 수행 결과값을 10개만 표시, 예측 확률 값으로 표시됨 \n",
      "예측값 100개만 표시:  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "pred_probs = xgb_model.predict(dtest)\n",
    "print(np.round(pred_probs[:10], 3))\n",
    "\n",
    "print('predict() 수행 결과값을 10개만 표시, 예측 확률 값으로 표시됨 ')\n",
    "preds = [1 if x > 0.5 else 0 for x in pred_probs]\n",
    "print('예측값 100개만 표시: ', preds[:100])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score, precision_score , recall_score , confusion_matrix\n",
    "\n",
    "def get_clf_eval(y_test , pred):\n",
    "    confusion = confusion_matrix( y_test, pred)\n",
    "    accuracy = accuracy_score(y_test , pred)\n",
    "    precision = precision_score(y_test , pred)\n",
    "    recall = recall_score(y_test , pred)\n",
    "    print('오차 행렬')\n",
    "    print(confusion)\n",
    "    print('정확도: {0:.4f}, 정밀도: {1:.4f}, 재현율: {2:.4f}'.format(accuracy , precision ,recall))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "오차 행렬\n",
      "[[1876    5]\n",
      " [  85   34]]\n",
      "정확도: 0.9550, 정밀도: 0.8718, 재현율: 0.2857\n"
     ]
    }
   ],
   "source": [
    "get_clf_eval(Y_test, preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "# depth : eta : stop = 20, 0.001, 1000  = 정확도: 0.9550, 정밀도: 0.8718, 재현율: 0.2857\n",
    "# depth : eta : stop = 10, 0.001, 200  = 09505 0.8846 0.1933\n",
    "# depth : eta : stop = 15, 0.001, 1000  = 09535 0.8421 0.2689\n",
    "# depth : eta : stop = 8, 0.001, 1000  = 정확도: 0.9480, 정밀도: 0.8947, 재현율: 0.1429\n",
    "# depth : eta : stop = 6, 0.001, 1000  = 정확도: 0.9455, 정밀도: 0.9167, 재현율: 0.0924\n",
    "# depth : eta : stop = 5, 0.001, 1000  = 정확도: 0.9450, 정밀도: 1.0000, 재현율: 0.0756\n",
    "# depth : eta : stop = 4, 0.001, 1000  = 정확도: 0.9420, 정밀도: 1.0000, 재현율: 0.0252\n",
    "# depth : eta : stop = 3, 0.001, 1000  = 정확도: 0.9410, 정밀도: 1.0000, 재현율: 0.0084"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}