learn.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ac = pd.read_csv('data/trainset/action_train.csv')\n",
    "order = pd.read_csv('data/trainset/orderFuture_train.csv')\n",
    "order_his = pd.read_csv('data/trainset/orderHistory_train.csv')\n",
    "comment = pd.read_csv('data/trainset/userComment_train.csv')\n",
    "profile = pd.read_csv('data/trainset/userProfile_train.csv')\n",
    "\n",
    "ac_test = pd.read_csv('data/test/action_test.csv')\n",
    "order_test = pd.read_csv('data/test/orderFuture_test.csv')\n",
    "order_his_test = pd.read_csv('data/test/orderHistory_test.csv')\n",
    "comment_test = pd.read_csv('data/test/userComment_test.csv')\n",
    "profile_test = pd.read_csv('data/test/userProfile_test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>actionType</th>\n",
       "      <th>actionTime</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>1</td>\n",
       "      <td>1474300753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>5</td>\n",
       "      <td>1474300763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>6</td>\n",
       "      <td>1474300874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>5</td>\n",
       "      <td>1474300911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>6</td>\n",
       "      <td>1474300936</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  actionType  actionTime\n",
       "0  100000000013           1  1474300753\n",
       "1  100000000013           5  1474300763\n",
       "2  100000000013           6  1474300874\n",
       "3  100000000013           5  1474300911\n",
       "4  100000000013           6  1474300936"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 用户id，行为类型，发生时间\n",
    "# 行为类型一共有9个，其中1是唤醒app；2~4是浏览产品，无先后关系；5~9则是有先后关系的，从填写表单到提交订单再到最后支付\n",
    "ac.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderType</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000111</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000127</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000231</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000379</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderType\n",
       "0  100000000013          0\n",
       "1  100000000111          0\n",
       "2  100000000127          0\n",
       "3  100000000231          0\n",
       "4  100000000379          0"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "order.head()\n",
    "#两列，分别是用户id和订单类型。供参赛者训练模型使用。其中1表示购买了精品旅游服务，0表示未购买精品旅游服务（包括普通旅游服务和未下订单）\n",
    "\n",
    "# 预测订单类型"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderid</th>\n",
       "      <th>rating</th>\n",
       "      <th>tags</th>\n",
       "      <th>commentsKeyWords</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>1000015</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['很','简陋','太','随便']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000231</td>\n",
       "      <td>1000024</td>\n",
       "      <td>5.0</td>\n",
       "      <td>提前联系|耐心等候</td>\n",
       "      <td>['很','细心']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000471</td>\n",
       "      <td>1000038</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000637</td>\n",
       "      <td>1000040</td>\n",
       "      <td>5.0</td>\n",
       "      <td>主动热情|提前联系|举牌迎接|主动搬运行李</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000755</td>\n",
       "      <td>1000045</td>\n",
       "      <td>1.0</td>\n",
       "      <td>未举牌服务</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderid  rating                   tags     commentsKeyWords\n",
       "0  100000000013  1000015     4.0                    NaN  ['很','简陋','太','随便']\n",
       "1  100000000231  1000024     5.0              提前联系|耐心等候           ['很','细心']\n",
       "2  100000000471  1000038     5.0                    NaN                  NaN\n",
       "3  100000000637  1000040     5.0  主动热情|提前联系|举牌迎接|主动搬运行李                  NaN\n",
       "4  100000000755  1000045     1.0                  未举牌服务                  NaN"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "comment.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderid</th>\n",
       "      <th>orderTime</th>\n",
       "      <th>orderType</th>\n",
       "      <th>city</th>\n",
       "      <th>country</th>\n",
       "      <th>continent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>1000015</td>\n",
       "      <td>1481714516</td>\n",
       "      <td>0</td>\n",
       "      <td>柏林</td>\n",
       "      <td>德国</td>\n",
       "      <td>欧洲</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>1000014</td>\n",
       "      <td>1501959643</td>\n",
       "      <td>0</td>\n",
       "      <td>旧金山</td>\n",
       "      <td>美国</td>\n",
       "      <td>北美洲</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000393</td>\n",
       "      <td>1000033</td>\n",
       "      <td>1499440296</td>\n",
       "      <td>0</td>\n",
       "      <td>巴黎</td>\n",
       "      <td>法国</td>\n",
       "      <td>欧洲</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000459</td>\n",
       "      <td>1000036</td>\n",
       "      <td>1480601668</td>\n",
       "      <td>0</td>\n",
       "      <td>纽约</td>\n",
       "      <td>美国</td>\n",
       "      <td>北美洲</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000459</td>\n",
       "      <td>1000034</td>\n",
       "      <td>1479146723</td>\n",
       "      <td>0</td>\n",
       "      <td>巴厘岛</td>\n",
       "      <td>印度尼西亚</td>\n",
       "      <td>亚洲</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderid   orderTime  orderType city country continent\n",
       "0  100000000013  1000015  1481714516          0   柏林      德国        欧洲\n",
       "1  100000000013  1000014  1501959643          0  旧金山      美国       北美洲\n",
       "2  100000000393  1000033  1499440296          0   巴黎      法国        欧洲\n",
       "3  100000000459  1000036  1480601668          0   纽约      美国       北美洲\n",
       "4  100000000459  1000034  1479146723          0  巴厘岛   印度尼西亚        亚洲"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "order_his.head()\n",
    "# 数据共有7列，分别是用户id，订单id，订单时间，订单类型，旅游城市，国家，大陆。其中1表示购买了精品旅游服务，0表示普通旅游服务。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>gender</th>\n",
       "      <th>province</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>男</td>\n",
       "      <td>NaN</td>\n",
       "      <td>60后</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000111</td>\n",
       "      <td>NaN</td>\n",
       "      <td>上海</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000127</td>\n",
       "      <td>NaN</td>\n",
       "      <td>上海</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000231</td>\n",
       "      <td>男</td>\n",
       "      <td>北京</td>\n",
       "      <td>70后</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000379</td>\n",
       "      <td>男</td>\n",
       "      <td>北京</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid gender province  age\n",
       "0  100000000013      男      NaN  60后\n",
       "1  100000000111    NaN       上海  NaN\n",
       "2  100000000127    NaN       上海  NaN\n",
       "3  100000000231      男       北京  70后\n",
       "4  100000000379      男       北京  NaN"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profile.head()\n",
    "# 数据共有四列，分别是用户id、性别、省份、年龄段"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2016-09-19 23:59:13'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "timeArray = time.localtime(1474300753)\n",
    "time.strftime('%Y-%m-%d %H:%M:%S', timeArray)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderType</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000231</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderType\n",
       "3  100000000231          0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "order[order.userid == 100000000231]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0                               ['很','简陋','太','随便']\n",
       "1                                        ['很','细心']\n",
       "2                                               NaN\n",
       "3                                               NaN\n",
       "4                                               NaN\n",
       "5                                               NaN\n",
       "6            ['真心','好','惊喜','惊喜','合理','还','真心','赞']\n",
       "7                                               NaN\n",
       "8     ['非常','满意','认真','负责','诚恳','合理','很','精致','希望']\n",
       "9                                               NaN\n",
       "10                                              NaN\n",
       "11                                              NaN\n",
       "12                                              NaN\n",
       "13                                  ['very','good']\n",
       "14                                              NaN\n",
       "15                              ['很','负责','辛苦','还']\n",
       "16        ['非常','好','还','好','还','很','热情','美丽','肯定']\n",
       "17                                      ['非常','满意']\n",
       "18                                              NaN\n",
       "19                                              NaN\n",
       "Name: commentsKeyWords, dtype: object"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "comment.commentsKeyWords.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def time_help(x):\n",
    "    timeArray = time.localtime(x)\n",
    "    return time.strftime('%Y-%m-%d %H:%M:%S', timeArray)\n",
    "    \n",
    "ac.actionTime = pd.to_datetime(ac.actionTime.apply(lambda x:time_help(x)), format='%Y-%m-%d %H:%M:%S')\n",
    "order_his.orderTime = pd.to_datetime(order_his.orderTime.apply(lambda x:time_help(x)), format='%Y-%m-%d %H:%M:%S')\n",
    "ac_test.actionTime = pd.to_datetime(ac_test.actionTime.apply(lambda x:time_help(x)), format='%Y-%m-%d %H:%M:%S')\n",
    "order_his_test.orderTime = pd.to_datetime(order_his_test.orderTime.apply(lambda x:time_help(x)), format='%Y-%m-%d %H:%M:%S')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>actionType</th>\n",
       "      <th>actionTime</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-09-19 23:59:13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>5</td>\n",
       "      <td>2016-09-19 23:59:23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>6</td>\n",
       "      <td>2016-09-20 00:01:14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>5</td>\n",
       "      <td>2016-09-20 00:01:51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>6</td>\n",
       "      <td>2016-09-20 00:02:16</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  actionType          actionTime\n",
       "0  100000000013           1 2016-09-19 23:59:13\n",
       "1  100000000013           5 2016-09-19 23:59:23\n",
       "2  100000000013           6 2016-09-20 00:01:14\n",
       "3  100000000013           5 2016-09-20 00:01:51\n",
       "4  100000000013           6 2016-09-20 00:02:16"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ac.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# t = ac.actionTime.iloc[1]\n",
    "ac['day'] = ac.actionTime.apply(lambda x:x.day)\n",
    "ac_test['day'] = ac_test.actionTime.apply(lambda x:x.day)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# print(t.year, t.month, t.day, t.second, t.minute,t.hour)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# ac['count'] = 1\n",
    "# ac[['userid', 'actionType']].groupby('userid', as_index=False).agg('last')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# ac[(ac.userid == 100000000459) && (ac.actionT)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# order[order.userid == 100000000459]\n",
    "# order['count'] = 1\n",
    "# order[['userid', 'count']].groupby('userid', as_index=False).agg('count')\n",
    "# del order['count']\n",
    "\n",
    "# ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg('last')\n",
    "# ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg('first')\n",
    "\n",
    "# ac.actionTime.\n",
    "# ac = pd.read_csv('data/trainset/action_train.csv')\n",
    "# ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg('last')\n",
    "# ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg('mean')\n",
    "\n",
    "# order_his[['userid', 'orderType']].groupby('userid', as_index=False).agg('max')\n",
    "\n",
    "# order  对应 userid,  所在  order_his的 最后一个 的 城市\n",
    "\n",
    "# order_his.head()\n",
    "# order_his.drop_duplicates(subset='userid', keep='last')\n",
    "# order_his[[order_his[['userid']].drop_duplicates(keep='last').index]]\n",
    "# pd.merge(order_his[['userid']].drop_duplicates(keep='last'), order_his, on='userid', how='left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "feature \n",
    "1. order  的 actionType 的最后一个\n",
    "  ac\n",
    "2. 时间间隔的第1个值和最后一个值\n",
    "  ac  actiontime\n",
    "3. 之前 是否定过精品，flag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# len(order_his_test.city.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "##user过去是否订购过orderType    以及比率\n",
    "def orderHistory_feat(df):\n",
    "    grouped=df[['userid','orderType']].groupby('userid',as_index=False)\n",
    "    df_count=grouped.count()\n",
    "    df_count.rename(columns={'orderType':'df_count'},inplace=True)\n",
    "    df_sum=grouped.sum()\n",
    "    \n",
    "    df_sum.rename(columns={'orderType':'df_sum'},inplace=True)\n",
    "    df_merge=pd.merge(df_count,df_sum,on='userid',how='left')\n",
    "    df_merge['rate']=df_merge['df_sum']/df_merge['df_count']\n",
    "    del df_merge['df_count']\n",
    "    df_merge.rename(columns={'df_sum':'orderHistory_feat_sum','rate':'orderHistory_feat_rate'},inplace=True)\n",
    "    \n",
    "    # rate  的 score 很低\n",
    "    del df_merge['orderHistory_feat_rate']\n",
    "    return df_merge\n",
    "\n",
    "# 点击比率\n",
    "def actions_orderType(df):\n",
    "    df['count']=1\n",
    "    df_count=df[['userid','count']].groupby('userid',as_index=False).count()\n",
    "    actionType=pd.get_dummies(df['actionType'],prefix='actionType')\n",
    "    df=pd.concat([df['userid'],actionType],axis=1)\n",
    "    df=df.groupby('userid',as_index=False).sum()\n",
    "    for column in range(1,df.shape[1]):\n",
    "        df['actionType_{}'.format(column)]=df['actionType_{}'.format(column)]/df_count['count']\n",
    "    return df\n",
    "\n",
    "\n",
    "def get_two(arr):\n",
    "    try:\n",
    "        tem = int(arr.iloc[-2])\n",
    "        return tem\n",
    "    except:\n",
    "        return np.nan\n",
    "\n",
    "def get_three(arr):\n",
    "    try:\n",
    "        tem = int(arr.iloc[-3])\n",
    "        return tem\n",
    "    except:\n",
    "        return np.nan\n",
    "\n",
    "# order 的 actionType 的最后一个 行为， 倒数第二个， 倒数第三个\n",
    "def last_action (df):\n",
    "    last_type = df[['userid', 'actionType']].groupby('userid', as_index=False).agg('last')\n",
    "    last_type.rename(columns={'actionType':'last_type'}, inplace=True)\n",
    "    \n",
    "    two_type = df[['userid', 'actionType']].groupby('userid', as_index=False).agg(get_two)\n",
    "    two_type.rename(columns={'actionType':'two_type'}, inplace=True)\n",
    "    \n",
    "    three_type = df[['userid', 'actionType']].groupby('userid', as_index=False).agg(get_three)\n",
    "    three_type.rename(columns={'actionType':'three_type'}, inplace=True)\n",
    "    two_type\n",
    "    \n",
    "    last = pd.merge(last_type, two_type, on='userid', how='left')\n",
    "    last = pd.merge(last, three_type, on='userid', how='left')\n",
    "    \n",
    "    return last\n",
    "    \n",
    "\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "le = LabelEncoder()\n",
    "\n",
    "# order所在城市\n",
    "def last_city(train, test):\n",
    "    city = list(order_his['city'].unique())\n",
    "    city_test = list(order_his_test['city'].unique())\n",
    "    for i in city_test:\n",
    "        if i not in city:\n",
    "            city.append(i)\n",
    "    le.fit(city)\n",
    "    \n",
    "    new = order_his[['userid', 'city', 'continent']].drop_duplicates(subset='userid', keep='last')\n",
    "    new['city'] = le.transform(new['city'])\n",
    "   \n",
    "    new_test = order_his_test[['userid', 'city', 'continent']].drop_duplicates(subset='userid', keep='last')\n",
    "    new_test['city'] = le.transform(new_test['city']) \n",
    "    \n",
    "    new['continent'] = le.fit_transform(new['continent'])\n",
    "    new_test['continent'] = le.transform(new_test['continent'])\n",
    "    \n",
    "    train = pd.merge(train, new, on='userid', how='left')\n",
    "    test = pd.merge(test, new_test, on='userid', how='left')\n",
    "    return train, test\n",
    "\n",
    "\n",
    "def profile_me(train, test):\n",
    "    \n",
    "    for i in ['province', 'gender', 'age']:\n",
    "        train[i] = train[i].factorize()[0]\n",
    "        test[i] = test[i].factorize()[0]\n",
    "        \n",
    "        train[i] = le.fit_transform(train[i])\n",
    "        test[i] = le.transform(test[i])\n",
    "    \n",
    "    return train, test\n",
    "\n",
    "\n",
    "def fun1(arr):\n",
    "#     print(arr.iloc[-1], arr.iloc[-2])\n",
    "    try:\n",
    "        return ((arr.iloc[-1]) - (arr.iloc[-2])).seconds\n",
    "    except:\n",
    "        return np.nan\n",
    "\n",
    "def fun2(arr):\n",
    "    try:\n",
    "        return ((arr.iloc[-1]) - (arr.iloc[-2])).days\n",
    "    except:\n",
    "        return np.nan\n",
    "\n",
    "def fun3(arr):\n",
    "    try:\n",
    "        return ((arr.iloc[-2]) - (arr.iloc[-3])).days\n",
    "    except:\n",
    "        return np.nan\n",
    "\n",
    "def fun4(arr):\n",
    "    try:\n",
    "        return ((arr.iloc[-2]) - (arr.iloc[-3])).seconds\n",
    "    except:\n",
    "        return np.nan\n",
    "        \n",
    "# ac 倒数第一次和 倒数第二次， 倒数第三次的间隔时间\n",
    "def ac_time(train, test):\n",
    "\n",
    "    las_two_day = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun2)\n",
    "    las_two_day.rename(columns={'actionTime':'last_two_day'}, inplace=True)\n",
    "    las_two_second = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun1)\n",
    "    las_two_second.rename(columns={'actionTime':'last_two_second'}, inplace=True)\n",
    "    \n",
    "    two_three_day = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun3)\n",
    "    two_three_day.rename(columns={'actionTime':'two_three_day'}, inplace=True)\n",
    "    two_three_second = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun4)\n",
    "    two_three_second.rename(columns={'actionTime':'two_three_second'}, inplace=True)\n",
    "    \n",
    "#         time_mean = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun3)\n",
    "#         time_mean.rename(columns={'actionTime':'time_mean'}, inplace=True)\n",
    "#         time_std = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun4)\n",
    "#         time_std.rename(columns={'actionTime':'time_std'}, inplace=True)\n",
    "\n",
    "    train = pd.merge(train, las_two_day, on='userid', how='left')\n",
    "    train = pd.merge(train, las_two_second, on='userid', how='left')\n",
    "    train = pd.merge(train, two_three_day, on='userid', how='left')\n",
    "    train = pd.merge(train, two_three_second, on='userid', how='left')\n",
    "\n",
    "    las_two_day_ = ac_test[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun2)\n",
    "    las_two_day_.rename(columns={'actionTime':'last_two_day'}, inplace=True)\n",
    "    las_two_second_ = ac_test[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun1)\n",
    "    las_two_second_.rename(columns={'actionTime':'last_two_second'}, inplace=True)\n",
    "    \n",
    "    two_three_day = ac_test[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun3)\n",
    "    two_three_day.rename(columns={'actionTime':'two_three_day'}, inplace=True)\n",
    "    two_three_second = ac_test[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun4)\n",
    "    two_three_second.rename(columns={'actionTime':'two_three_second'}, inplace=True)\n",
    "    \n",
    "#         time_std = ac_test[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun4)\n",
    "#         time_std.rename(columns={'actionTime':'time_std'}, inplace=True)\n",
    "\n",
    "\n",
    "    test = pd.merge(test, las_two_day_, on='userid', how='left')\n",
    "    test = pd.merge(test, las_two_second_, on='userid', how='left')\n",
    "    test = pd.merge(test, two_three_day, on='userid', how='left')\n",
    "    test = pd.merge(test, two_three_second, on='userid', how='left')\n",
    "\n",
    "    return train, test\n",
    "\n",
    "# comment rating\n",
    "# def ratting_score()\n",
    "\n",
    "def fun5(timeseries):\n",
    "# 时间间隔的均值\n",
    "    sum = []\n",
    "    for i in range(len(timeseries) - 1):\n",
    "        sum.append((timeseries.iloc[i+ 1] - timeseries.iloc[i]).seconds)\n",
    "    return np.mean(sum)\n",
    "\n",
    "def fun6(timeseries):\n",
    "# 时间间隔的均值\n",
    "    sum = []\n",
    "    for i in range(len(timeseries) - 1):\n",
    "        sum.append((timeseries.iloc[i+1] - timeseries.iloc[i]).seconds)\n",
    "    return np.std(sum)\n",
    "\n",
    "def time_delta(train, test):\n",
    "    time_mean = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun5)\n",
    "    time_mean.rename(columns={'actionTime':'timedelta_mean'}, inplace=True)\n",
    "    time_std = ac[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun5)\n",
    "    time_std.rename(columns={'actionTime':'timedelta_std'}, inplace=True)\n",
    "    \n",
    "    train = pd.merge(train, time_mean, on='userid', how='left')\n",
    "    train = pd.merge(train, time_std, on='userid', how='left')\n",
    "    \n",
    "    time_mean = ac_test[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun5)\n",
    "    time_mean.rename(columns={'actionTime':'timedelta_mean'}, inplace=True)\n",
    "    time_std = ac_test[['userid', 'actionTime']].groupby('userid', as_index=False).agg(fun6)\n",
    "    time_std.rename(columns={'actionTime':'timedelta_std'}, inplace=True)\n",
    "    \n",
    "    test = pd.merge(test, time_mean, on='userid', how='left')\n",
    "    test = pd.merge(test, time_std, on='userid', how='left')\n",
    "    return train, test\n",
    "\n",
    "# 历史最后一次时间对应action中点击的什么, 组合\n",
    "def his_s(train, test):\n",
    "    df = order_his.drop_duplicates(subset='userid', keep='last').reset_index(drop=True)\n",
    "    last_type = []\n",
    "    for i in range(len(df)):\n",
    "        temp_time = df.loc[i, 'orderTime']\n",
    "        temp_id = df.loc[i, 'userid']\n",
    "        new = ac[ac.userid == temp_id]\n",
    "        new = new[new.actionTime < temp_time]\n",
    "        if(len(new) != 0):\n",
    "            new = new.iloc[-1]\n",
    "            last_type.append(new.actionType)\n",
    "        else:\n",
    "            last_type.append(np.nan)\n",
    "    last_type = pd.Series(np.array(last_type))\n",
    "    last_type.index = df.index\n",
    "    last_type = pd.DataFrame(last_type, columns=['hist_type'])\n",
    "    last_type['userid'] = df['userid']\n",
    "    train = pd.merge(train, last_type, on='userid', how='left')\n",
    "    \n",
    "    df = order_his_test.drop_duplicates(subset='userid', keep='last').reset_index(drop=True)\n",
    "    last_type = []\n",
    "    for i in range(len(df)):\n",
    "        temp_time = df.loc[i, 'orderTime']\n",
    "        temp_id = df.loc[i, 'userid']\n",
    "        new = ac_test[ac_test.userid == temp_id]\n",
    "        new = new[new.actionTime < temp_time]\n",
    "        if(len(new) != 0):\n",
    "            new = new.iloc[-1]\n",
    "            last_type.append(new.actionType)\n",
    "        else:\n",
    "            last_type.append(np.nan)\n",
    "    last_type = pd.Series(np.array(last_type))\n",
    "    last_type.index = df.index\n",
    "    last_type = pd.DataFrame(last_type, columns=['hist_type'])\n",
    "    last_type['userid'] = df['userid']\n",
    "    \n",
    "    test = pd.merge(test, last_type, on='userid', how='left')\n",
    "    return train, test\n",
    "\n",
    "\n",
    "# def fun7(type)\n",
    "\n",
    "# 距离 最近的 5, 6 , 7 , 8的 seconds\n",
    "def distance_last(odf, a=5):\n",
    "    name = 'flag_'+ str(a)\n",
    "    odf[name] = (odf.actionType == a) * 1\n",
    "    \n",
    "    id = []\n",
    "    dis = []\n",
    "    for i in list(odf.userid.unique()):\n",
    "        id.append(i)\n",
    "        \n",
    "        df = odf[odf.userid == i]\n",
    "        time1 = df.actionTime.iloc[-1]\n",
    "        \n",
    "        df = df[df[name] == 1]\n",
    "        if (len(df) != 0):\n",
    "            if df.iloc[-1].actionTime == time1:\n",
    "                df = df[:-1]\n",
    "            if len(df) >= 1:\n",
    "                dis.append((time1 - df.iloc[-1].actionTime).seconds)\n",
    "            else:\n",
    "                dis.append(np.nan)\n",
    "        else:\n",
    "            dis.append(np.nan)\n",
    "    column_name = 'last_' + str(a) + 'dis'\n",
    "    new_df = pd.DataFrame({'userid':id,\n",
    "                           column_name:dis})\n",
    "    return new_df\n",
    "    \n",
    "        \n",
    "    \n",
    "\n",
    "# 最后操作那一天 action 的count\n",
    "def action_count(train, test):\n",
    "    count = []\n",
    "    id = []\n",
    "    for i in (ac.userid.unique()):\n",
    "        id.append(i)\n",
    "        df = ac[ac.userid == i]\n",
    "        df = df.loc[df.day == df.iloc[-1].day]\n",
    "        count.append(len(df))\n",
    "    new = pd.DataFrame({'userid':id, \n",
    "                       'count':count})\n",
    "    train = pd.merge(train, new, on='userid', how='left')\n",
    "    \n",
    "    count = []\n",
    "    id = []\n",
    "    for i in (ac_test.userid.unique()):\n",
    "        id.append(i)\n",
    "        df = ac_test[ac_test.userid == i]\n",
    "        df = df.loc[df.day == df.iloc[-1].day]\n",
    "        count.append(len(df))\n",
    "    new = pd.DataFrame({'userid':id, \n",
    "                       'count':count})\n",
    "    test = pd.merge(test, new, on='userid', how='left')\n",
    "    return train, test\n",
    "\n",
    "# orderType\n",
    "def flags_buy(df):\n",
    "    flags = []\n",
    "    id = []\n",
    "    for i in list(df.userid.unique()):\n",
    "        id.append(i)\n",
    "        df_ = df[df.userid == i]\n",
    "        if df_.orderType.sum() != 0:\n",
    "            flags.append(1)\n",
    "        else:\n",
    "            flags.append(0)\n",
    "    new = pd.DataFrame({\n",
    "        'userid':id,\n",
    "        'buy_flags':flags\n",
    "    })\n",
    "    return new"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def gen_train_feat():\n",
    "    actions = order\n",
    "    actions_test = order_test\n",
    "    \n",
    "   \n",
    "    actions = pd.merge(actions, distance_last(ac, 5), on='userid', how='left')\n",
    "    actions_test = pd.merge(actions_test, distance_last(ac_test, 5), on='userid', how='left')\n",
    "    \n",
    "    actions = pd.merge(actions, distance_last(ac, 6), on='userid', how='left')\n",
    "    actions_test = pd.merge(actions_test, distance_last(ac_test, 6), on='userid', how='left')\n",
    "    \n",
    "    actions = pd.merge(actions, distance_last(ac, 7), on='userid', how='left')\n",
    "    actions_test = pd.merge(actions_test, distance_last(ac_test, 7), on='userid', how='left')\n",
    "    \n",
    "    actions = pd.merge(actions, distance_last(ac, 8), on='userid', how='left')\n",
    "    actions_test = pd.merge(actions_test, distance_last(ac_test, 8), on='userid', how='left')\n",
    "    \n",
    "    actions = pd.merge(actions, distance_last(ac, 9), on='userid', how='left')\n",
    "    actions_test = pd.merge(actions_test, distance_last(ac_test, 9), on='userid', how='left')\n",
    "    actions, actions_test = action_count(actions, actions_test)\n",
    "    actions = pd.merge(actions,orderHistory_feat(order_his),on='userid',how='left')\n",
    "    actions = pd.merge(actions,actions_orderType(ac),on='userid',how='left')\n",
    "    actions = pd.merge(actions,last_action(ac),on='userid',how='left')  \n",
    "    \n",
    "    actions_test = pd.merge(actions_test,orderHistory_feat(order_his_test),on='userid',how='left')\n",
    "    actions_test = pd.merge(actions_test,actions_orderType(ac_test),on='userid',how='left')\n",
    "    actions_test = pd.merge(actions_test,last_action(ac_test),on='userid',how='left')\n",
    "    \n",
    "    actions, actions_test = last_city(actions, actions_test)\n",
    "    \n",
    "    profile_, profile_test_ = profile_me(profile, profile_test)\n",
    "    actions = pd.merge(actions, profile_, on='userid', how='left')\n",
    "    actions_test = pd.merge(actions_test, profile_test_, on='userid', how='left')\n",
    "    actions, actions_test = ac_time(actions, actions_test)\n",
    "    \n",
    "    actions = pd.merge(actions, comment[['userid', 'rating']], on='userid', how='left')\n",
    "    actions_test = pd.merge(actions_test, comment_test[['userid', 'rating']], on='userid', how='left')\n",
    "    \n",
    "    actions['flag_five'] = (actions['last_type'] == 5) * 1\n",
    "    actions['flag_six'] = (actions['last_type'] == 6) * 1\n",
    "    actions['flag_seven'] = (actions['last_type'] == 1) * 1\n",
    "    \n",
    "    actions_test['flag_five'] = (actions_test['last_type'] == 5) * 1\n",
    "    actions_test['flag_six'] = (actions_test['last_type'] == 6) * 1\n",
    "    actions_test['flag_seven'] = (actions_test['last_type'] == 7) * 1\n",
    "    \n",
    "    \n",
    "#     actions = pd.merge(actions, flags_buy(order_his), on='userid', how='left')\n",
    "#     actions_test = pd.merge(actions_test, flags_buy(order_his_test), on='userid', how='left')\n",
    "    \n",
    "#     actions, actions_test = time_delta(actions, actions_test)\n",
    "\n",
    "#     actions, actions_test = his_s(actions, actions_test)\n",
    "    return actions, actions_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train_data.to_csv('data/checkpoint_train.csv', index=False)\n",
    "# test_data.to_csv('data/checkpoint_test.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>label</th>\n",
       "      <th>buy_flags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000111</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000127</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000231</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000379</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  label  buy_flags\n",
       "0  100000000013      0        0.0\n",
       "1  100000000111      0        NaN\n",
       "2  100000000127      0        NaN\n",
       "3  100000000231      0        NaN\n",
       "4  100000000379      0        NaN"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "order.rename(columns={'orderType':'label'},inplace=True)\n",
    "train_data, test_data=gen_train_feat()\n",
    "train_data.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "#c_train = pd.read_csv('data/checkpoint_train.csv')\n",
    "#c_test = pd.read_csv('data/checkpoint_test.csv')\n",
    "#del c_train['orderType']\n",
    "#train_data = pd.merge(c_train, train_data, how='left', on='userid')\n",
    "#test_data = pd.merge(c_test, test_data, how='left', on='userid')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "train_label=train_data['label']\n",
    "del train_data['label']\n",
    "x_train,x_val,y_train,y_val=train_test_split(train_data,train_label,test_size=0.2,random_state=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 32245 entries, 5544 to 38408\n",
      "Data columns (total 34 columns):\n",
      "userid                   32245 non-null int64\n",
      "last_5dis                30476 non-null float64\n",
      "last_6dis                26282 non-null float64\n",
      "last_7dis                7203 non-null float64\n",
      "last_8dis                9526 non-null float64\n",
      "last_9dis                7135 non-null float64\n",
      "count                    32245 non-null int64\n",
      "orderHistory_feat_sum    8518 non-null float64\n",
      "actionType_1             32245 non-null float64\n",
      "actionType_2             32245 non-null float64\n",
      "actionType_3             32245 non-null float64\n",
      "actionType_4             32245 non-null float64\n",
      "actionType_5             32245 non-null float64\n",
      "actionType_6             32245 non-null float64\n",
      "actionType_7             32245 non-null float64\n",
      "actionType_8             32245 non-null float64\n",
      "actionType_9             32245 non-null float64\n",
      "last_type                32245 non-null int64\n",
      "two_type                 31240 non-null float64\n",
      "three_type               30649 non-null float64\n",
      "city                     8518 non-null float64\n",
      "continent                8518 non-null float64\n",
      "gender                   32245 non-null int64\n",
      "province                 32245 non-null int64\n",
      "age                      32245 non-null int64\n",
      "last_two_day             31240 non-null float64\n",
      "last_two_second          31240 non-null float64\n",
      "two_three_day            30649 non-null float64\n",
      "two_three_second         30649 non-null float64\n",
      "rating                   7857 non-null float64\n",
      "flag_five                32245 non-null int64\n",
      "flag_six                 32245 non-null int64\n",
      "flag_seven               32245 non-null int64\n",
      "buy_flags                8518 non-null float64\n",
      "dtypes: float64(25), int64(9)\n",
      "memory usage: 8.6 MB\n"
     ]
    }
   ],
   "source": [
    "x_train.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 10076 entries, 0 to 10075\n",
      "Data columns (total 34 columns):\n",
      "userid                   10076 non-null int64\n",
      "last_5dis                9501 non-null float64\n",
      "last_6dis                8200 non-null float64\n",
      "last_7dis                2252 non-null float64\n",
      "last_8dis                2995 non-null float64\n",
      "last_9dis                2219 non-null float64\n",
      "count                    10076 non-null int64\n",
      "orderHistory_feat_sum    2686 non-null float64\n",
      "actionType_1             10076 non-null float64\n",
      "actionType_2             10076 non-null float64\n",
      "actionType_3             10076 non-null float64\n",
      "actionType_4             10076 non-null float64\n",
      "actionType_5             10076 non-null float64\n",
      "actionType_6             10076 non-null float64\n",
      "actionType_7             10076 non-null float64\n",
      "actionType_8             10076 non-null float64\n",
      "actionType_9             10076 non-null float64\n",
      "last_type                10076 non-null int64\n",
      "two_type                 9722 non-null float64\n",
      "three_type               9544 non-null float64\n",
      "city                     2686 non-null float64\n",
      "continent                2686 non-null float64\n",
      "gender                   10076 non-null int64\n",
      "province                 10076 non-null int64\n",
      "age                      10076 non-null int64\n",
      "last_two_day             9722 non-null float64\n",
      "last_two_second          9722 non-null float64\n",
      "two_three_day            9544 non-null float64\n",
      "two_three_second         9544 non-null float64\n",
      "rating                   2474 non-null float64\n",
      "flag_five                10076 non-null int64\n",
      "flag_six                 10076 non-null int64\n",
      "flag_seven               10076 non-null int64\n",
      "buy_flags                2686 non-null float64\n",
      "dtypes: float64(25), int64(9)\n",
      "memory usage: 3.0 MB\n"
     ]
    }
   ],
   "source": [
    "# del test_data['orderType']\n",
    "test_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 调参失败， 电脑炸了"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# del test_data['orderType']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "start running ....\n",
      "[0]\teval-auc:0.819828\ttrain-auc:0.816973\n",
      "Multiple eval metrics have been passed: 'train-auc' will be used for early stopping.\n",
      "\n",
      "Will train until train-auc hasn't improved in 100 rounds.\n",
      "[1]\teval-auc:0.839604\ttrain-auc:0.834355\n",
      "[2]\teval-auc:0.841011\ttrain-auc:0.836347\n",
      "[3]\teval-auc:0.860001\ttrain-auc:0.853497\n",
      "[4]\teval-auc:0.864739\ttrain-auc:0.855608\n",
      "[5]\teval-auc:0.86561\ttrain-auc:0.856885\n",
      "[6]\teval-auc:0.867917\ttrain-auc:0.854782\n",
      "[7]\teval-auc:0.864301\ttrain-auc:0.851976\n",
      "[8]\teval-auc:0.866368\ttrain-auc:0.853649\n",
      "[9]\teval-auc:0.869583\ttrain-auc:0.858388\n",
      "[10]\teval-auc:0.87903\ttrain-auc:0.867452\n",
      "[11]\teval-auc:0.87965\ttrain-auc:0.866644\n",
      "[12]\teval-auc:0.878375\ttrain-auc:0.866654\n",
      "[13]\teval-auc:0.881434\ttrain-auc:0.871081\n",
      "[14]\teval-auc:0.883573\ttrain-auc:0.87275\n",
      "[15]\teval-auc:0.885187\ttrain-auc:0.874749\n",
      "[16]\teval-auc:0.887769\ttrain-auc:0.876837\n",
      "[17]\teval-auc:0.887781\ttrain-auc:0.877328\n",
      "[18]\teval-auc:0.889127\ttrain-auc:0.879404\n",
      "[19]\teval-auc:0.893457\ttrain-auc:0.88439\n",
      "[20]\teval-auc:0.893514\ttrain-auc:0.885517\n",
      "[21]\teval-auc:0.894028\ttrain-auc:0.885562\n",
      "[22]\teval-auc:0.894319\ttrain-auc:0.885727\n",
      "[23]\teval-auc:0.89546\ttrain-auc:0.886197\n",
      "[24]\teval-auc:0.899876\ttrain-auc:0.890706\n",
      "[25]\teval-auc:0.902753\ttrain-auc:0.894211\n",
      "[26]\teval-auc:0.903336\ttrain-auc:0.894416\n",
      "[27]\teval-auc:0.905964\ttrain-auc:0.896551\n",
      "[28]\teval-auc:0.906908\ttrain-auc:0.897596\n",
      "[29]\teval-auc:0.908534\ttrain-auc:0.899164\n",
      "[30]\teval-auc:0.912367\ttrain-auc:0.903208\n",
      "[31]\teval-auc:0.913663\ttrain-auc:0.904624\n",
      "[32]\teval-auc:0.914441\ttrain-auc:0.906165\n",
      "[33]\teval-auc:0.915232\ttrain-auc:0.907098\n",
      "[34]\teval-auc:0.917216\ttrain-auc:0.90935\n",
      "[35]\teval-auc:0.918015\ttrain-auc:0.910373\n",
      "[36]\teval-auc:0.918441\ttrain-auc:0.911273\n",
      "[37]\teval-auc:0.91969\ttrain-auc:0.913049\n",
      "[38]\teval-auc:0.920284\ttrain-auc:0.913662\n",
      "[39]\teval-auc:0.920878\ttrain-auc:0.914277\n",
      "[40]\teval-auc:0.921708\ttrain-auc:0.915138\n",
      "[41]\teval-auc:0.922781\ttrain-auc:0.916155\n",
      "[42]\teval-auc:0.923299\ttrain-auc:0.916803\n",
      "[43]\teval-auc:0.923826\ttrain-auc:0.91747\n",
      "[44]\teval-auc:0.925395\ttrain-auc:0.918893\n",
      "[45]\teval-auc:0.927166\ttrain-auc:0.920836\n",
      "[46]\teval-auc:0.927484\ttrain-auc:0.921465\n",
      "[47]\teval-auc:0.927641\ttrain-auc:0.921795\n",
      "[48]\teval-auc:0.928315\ttrain-auc:0.922977\n",
      "[49]\teval-auc:0.928963\ttrain-auc:0.923631\n",
      "[50]\teval-auc:0.930242\ttrain-auc:0.925062\n",
      "[51]\teval-auc:0.930658\ttrain-auc:0.925624\n",
      "[52]\teval-auc:0.931054\ttrain-auc:0.926115\n",
      "[53]\teval-auc:0.931308\ttrain-auc:0.926495\n",
      "[54]\teval-auc:0.931621\ttrain-auc:0.926932\n",
      "[55]\teval-auc:0.932616\ttrain-auc:0.928273\n",
      "[56]\teval-auc:0.932853\ttrain-auc:0.928523\n",
      "[57]\teval-auc:0.933266\ttrain-auc:0.928977\n",
      "[58]\teval-auc:0.93389\ttrain-auc:0.929506\n",
      "[59]\teval-auc:0.934074\ttrain-auc:0.929825\n",
      "[60]\teval-auc:0.934311\ttrain-auc:0.93018\n",
      "[61]\teval-auc:0.934563\ttrain-auc:0.930606\n",
      "[62]\teval-auc:0.934852\ttrain-auc:0.931013\n",
      "[63]\teval-auc:0.935053\ttrain-auc:0.931219\n",
      "[64]\teval-auc:0.935403\ttrain-auc:0.931859\n",
      "[65]\teval-auc:0.935893\ttrain-auc:0.932426\n",
      "[66]\teval-auc:0.93623\ttrain-auc:0.933032\n",
      "[67]\teval-auc:0.936585\ttrain-auc:0.93365\n",
      "[68]\teval-auc:0.936958\ttrain-auc:0.933922\n",
      "[69]\teval-auc:0.93705\ttrain-auc:0.934009\n",
      "[70]\teval-auc:0.937169\ttrain-auc:0.934423\n",
      "[71]\teval-auc:0.937211\ttrain-auc:0.934535\n",
      "[72]\teval-auc:0.937354\ttrain-auc:0.935063\n",
      "[73]\teval-auc:0.937432\ttrain-auc:0.935314\n",
      "[74]\teval-auc:0.937674\ttrain-auc:0.935643\n",
      "[75]\teval-auc:0.937978\ttrain-auc:0.93596\n",
      "[76]\teval-auc:0.938131\ttrain-auc:0.936264\n",
      "[77]\teval-auc:0.938329\ttrain-auc:0.93655\n",
      "[78]\teval-auc:0.938424\ttrain-auc:0.936675\n",
      "[79]\teval-auc:0.93864\ttrain-auc:0.936951\n",
      "[80]\teval-auc:0.93908\ttrain-auc:0.937369\n",
      "[81]\teval-auc:0.939355\ttrain-auc:0.937577\n",
      "[82]\teval-auc:0.939513\ttrain-auc:0.937748\n",
      "[83]\teval-auc:0.939604\ttrain-auc:0.937882\n",
      "[84]\teval-auc:0.939663\ttrain-auc:0.937977\n",
      "[85]\teval-auc:0.939649\ttrain-auc:0.937974\n",
      "[86]\teval-auc:0.939801\ttrain-auc:0.938248\n",
      "[87]\teval-auc:0.939937\ttrain-auc:0.938405\n",
      "[88]\teval-auc:0.940056\ttrain-auc:0.938664\n",
      "[89]\teval-auc:0.940306\ttrain-auc:0.938912\n",
      "[90]\teval-auc:0.940391\ttrain-auc:0.939045\n",
      "[91]\teval-auc:0.940753\ttrain-auc:0.939354\n",
      "[92]\teval-auc:0.940981\ttrain-auc:0.939822\n",
      "[93]\teval-auc:0.940811\ttrain-auc:0.940089\n",
      "[94]\teval-auc:0.940945\ttrain-auc:0.940244\n",
      "[95]\teval-auc:0.941077\ttrain-auc:0.940343\n",
      "[96]\teval-auc:0.941121\ttrain-auc:0.940415\n",
      "[97]\teval-auc:0.941296\ttrain-auc:0.940836\n",
      "[98]\teval-auc:0.941408\ttrain-auc:0.941164\n",
      "[99]\teval-auc:0.941522\ttrain-auc:0.941353\n",
      "[100]\teval-auc:0.941675\ttrain-auc:0.94165\n",
      "[101]\teval-auc:0.94185\ttrain-auc:0.942008\n",
      "[102]\teval-auc:0.941809\ttrain-auc:0.94222\n",
      "[103]\teval-auc:0.941881\ttrain-auc:0.942343\n",
      "[104]\teval-auc:0.942582\ttrain-auc:0.943093\n",
      "[105]\teval-auc:0.942628\ttrain-auc:0.943148\n",
      "[106]\teval-auc:0.942683\ttrain-auc:0.943208\n",
      "[107]\teval-auc:0.942845\ttrain-auc:0.943577\n",
      "[108]\teval-auc:0.943009\ttrain-auc:0.943753\n",
      "[109]\teval-auc:0.943138\ttrain-auc:0.943928\n",
      "[110]\teval-auc:0.943113\ttrain-auc:0.943963\n",
      "[111]\teval-auc:0.943244\ttrain-auc:0.944138\n",
      "[112]\teval-auc:0.943194\ttrain-auc:0.944357\n",
      "[113]\teval-auc:0.94335\ttrain-auc:0.944588\n",
      "[114]\teval-auc:0.943437\ttrain-auc:0.944807\n",
      "[115]\teval-auc:0.94349\ttrain-auc:0.944934\n",
      "[116]\teval-auc:0.943499\ttrain-auc:0.945012\n",
      "[117]\teval-auc:0.943598\ttrain-auc:0.945109\n",
      "[118]\teval-auc:0.943595\ttrain-auc:0.94521\n",
      "[119]\teval-auc:0.943653\ttrain-auc:0.945265\n",
      "[120]\teval-auc:0.943962\ttrain-auc:0.94565\n",
      "[121]\teval-auc:0.94407\ttrain-auc:0.945818\n",
      "[122]\teval-auc:0.944029\ttrain-auc:0.945956\n",
      "[123]\teval-auc:0.944009\ttrain-auc:0.946035\n",
      "[124]\teval-auc:0.944573\ttrain-auc:0.946546\n",
      "[125]\teval-auc:0.944602\ttrain-auc:0.946679\n",
      "[126]\teval-auc:0.944923\ttrain-auc:0.946949\n",
      "[127]\teval-auc:0.944953\ttrain-auc:0.946995\n",
      "[128]\teval-auc:0.944908\ttrain-auc:0.947031\n",
      "[129]\teval-auc:0.945104\ttrain-auc:0.947325\n",
      "[130]\teval-auc:0.945115\ttrain-auc:0.947411\n",
      "[131]\teval-auc:0.945151\ttrain-auc:0.9476\n",
      "[132]\teval-auc:0.945215\ttrain-auc:0.947631\n",
      "[133]\teval-auc:0.945252\ttrain-auc:0.947791\n",
      "[134]\teval-auc:0.945234\ttrain-auc:0.947981\n",
      "[135]\teval-auc:0.945313\ttrain-auc:0.948128\n",
      "[136]\teval-auc:0.945253\ttrain-auc:0.948243\n",
      "[137]\teval-auc:0.945407\ttrain-auc:0.948423\n",
      "[138]\teval-auc:0.94547\ttrain-auc:0.948556\n",
      "[139]\teval-auc:0.945807\ttrain-auc:0.948982\n",
      "[140]\teval-auc:0.945801\ttrain-auc:0.949111\n",
      "[141]\teval-auc:0.945983\ttrain-auc:0.949394\n",
      "[142]\teval-auc:0.946011\ttrain-auc:0.949493\n",
      "[143]\teval-auc:0.946\ttrain-auc:0.94952\n",
      "[144]\teval-auc:0.946007\ttrain-auc:0.949652\n",
      "[145]\teval-auc:0.945988\ttrain-auc:0.949701\n",
      "[146]\teval-auc:0.946117\ttrain-auc:0.949898\n",
      "[147]\teval-auc:0.946162\ttrain-auc:0.950046\n",
      "[148]\teval-auc:0.946246\ttrain-auc:0.950285\n",
      "[149]\teval-auc:0.946191\ttrain-auc:0.950331\n",
      "[150]\teval-auc:0.94631\ttrain-auc:0.950473\n",
      "[151]\teval-auc:0.946324\ttrain-auc:0.950528\n",
      "[152]\teval-auc:0.946382\ttrain-auc:0.95064\n",
      "[153]\teval-auc:0.946373\ttrain-auc:0.950681\n",
      "[154]\teval-auc:0.94634\ttrain-auc:0.950729\n",
      "[155]\teval-auc:0.946344\ttrain-auc:0.950886\n",
      "[156]\teval-auc:0.946348\ttrain-auc:0.951036\n",
      "[157]\teval-auc:0.946482\ttrain-auc:0.951174\n",
      "[158]\teval-auc:0.946597\ttrain-auc:0.9514\n",
      "[159]\teval-auc:0.946583\ttrain-auc:0.951481\n",
      "[160]\teval-auc:0.94659\ttrain-auc:0.951519\n",
      "[161]\teval-auc:0.946585\ttrain-auc:0.951563\n",
      "[162]\teval-auc:0.946684\ttrain-auc:0.951693\n",
      "[163]\teval-auc:0.946714\ttrain-auc:0.951742\n",
      "[164]\teval-auc:0.946691\ttrain-auc:0.95182\n",
      "[165]\teval-auc:0.946755\ttrain-auc:0.951941\n",
      "[166]\teval-auc:0.946815\ttrain-auc:0.952116\n",
      "[167]\teval-auc:0.946855\ttrain-auc:0.95221\n",
      "[168]\teval-auc:0.946919\ttrain-auc:0.952313\n",
      "[169]\teval-auc:0.946917\ttrain-auc:0.952374\n",
      "[170]\teval-auc:0.947116\ttrain-auc:0.95259\n",
      "[171]\teval-auc:0.947139\ttrain-auc:0.952646\n",
      "[172]\teval-auc:0.947261\ttrain-auc:0.952815\n",
      "[173]\teval-auc:0.947306\ttrain-auc:0.952958\n",
      "[174]\teval-auc:0.947328\ttrain-auc:0.953018\n",
      "[175]\teval-auc:0.947302\ttrain-auc:0.953089\n",
      "[176]\teval-auc:0.947413\ttrain-auc:0.953201\n",
      "[177]\teval-auc:0.947445\ttrain-auc:0.953305\n",
      "[178]\teval-auc:0.947668\ttrain-auc:0.95341\n",
      "[179]\teval-auc:0.947673\ttrain-auc:0.953415\n",
      "[180]\teval-auc:0.947698\ttrain-auc:0.953575\n",
      "[181]\teval-auc:0.947651\ttrain-auc:0.953675\n",
      "[182]\teval-auc:0.94766\ttrain-auc:0.953713\n",
      "[183]\teval-auc:0.947644\ttrain-auc:0.953753\n",
      "[184]\teval-auc:0.947611\ttrain-auc:0.953781\n",
      "[185]\teval-auc:0.947592\ttrain-auc:0.953849\n",
      "[186]\teval-auc:0.947703\ttrain-auc:0.953957\n",
      "[187]\teval-auc:0.947782\ttrain-auc:0.954129\n",
      "[188]\teval-auc:0.94783\ttrain-auc:0.954186\n",
      "[189]\teval-auc:0.94869\ttrain-auc:0.95477\n",
      "[190]\teval-auc:0.948762\ttrain-auc:0.954918\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[191]\teval-auc:0.948751\ttrain-auc:0.954954\n",
      "[192]\teval-auc:0.948772\ttrain-auc:0.955068\n",
      "[193]\teval-auc:0.948742\ttrain-auc:0.955179\n",
      "[194]\teval-auc:0.948796\ttrain-auc:0.955268\n",
      "[195]\teval-auc:0.948839\ttrain-auc:0.95535\n",
      "[196]\teval-auc:0.948822\ttrain-auc:0.955557\n",
      "[197]\teval-auc:0.948834\ttrain-auc:0.955655\n",
      "[198]\teval-auc:0.949328\ttrain-auc:0.955977\n",
      "[199]\teval-auc:0.949364\ttrain-auc:0.956087\n",
      "[200]\teval-auc:0.949393\ttrain-auc:0.956178\n",
      "[201]\teval-auc:0.949539\ttrain-auc:0.956327\n",
      "[202]\teval-auc:0.949548\ttrain-auc:0.956349\n",
      "[203]\teval-auc:0.949574\ttrain-auc:0.956439\n",
      "[204]\teval-auc:0.949645\ttrain-auc:0.95651\n",
      "[205]\teval-auc:0.949664\ttrain-auc:0.956564\n",
      "[206]\teval-auc:0.949676\ttrain-auc:0.956668\n",
      "[207]\teval-auc:0.949723\ttrain-auc:0.956732\n",
      "[208]\teval-auc:0.949727\ttrain-auc:0.956766\n",
      "[209]\teval-auc:0.949721\ttrain-auc:0.956823\n",
      "[210]\teval-auc:0.949778\ttrain-auc:0.956907\n",
      "[211]\teval-auc:0.949802\ttrain-auc:0.956971\n",
      "[212]\teval-auc:0.949837\ttrain-auc:0.957023\n",
      "[213]\teval-auc:0.949897\ttrain-auc:0.957075\n",
      "[214]\teval-auc:0.949922\ttrain-auc:0.957179\n",
      "[215]\teval-auc:0.949896\ttrain-auc:0.957232\n",
      "[216]\teval-auc:0.950002\ttrain-auc:0.957322\n",
      "[217]\teval-auc:0.950005\ttrain-auc:0.957419\n",
      "[218]\teval-auc:0.95008\ttrain-auc:0.957553\n",
      "[219]\teval-auc:0.950072\ttrain-auc:0.957614\n",
      "[220]\teval-auc:0.950047\ttrain-auc:0.957662\n",
      "[221]\teval-auc:0.95002\ttrain-auc:0.957684\n",
      "[222]\teval-auc:0.950016\ttrain-auc:0.95773\n",
      "[223]\teval-auc:0.95006\ttrain-auc:0.957785\n",
      "[224]\teval-auc:0.950114\ttrain-auc:0.95792\n",
      "[225]\teval-auc:0.95009\ttrain-auc:0.95801\n",
      "[226]\teval-auc:0.950081\ttrain-auc:0.958094\n",
      "[227]\teval-auc:0.95011\ttrain-auc:0.958118\n",
      "[228]\teval-auc:0.950108\ttrain-auc:0.958236\n",
      "[229]\teval-auc:0.950207\ttrain-auc:0.958434\n",
      "[230]\teval-auc:0.950185\ttrain-auc:0.958521\n",
      "[231]\teval-auc:0.950194\ttrain-auc:0.95858\n",
      "[232]\teval-auc:0.950203\ttrain-auc:0.958671\n",
      "[233]\teval-auc:0.950236\ttrain-auc:0.958723\n",
      "[234]\teval-auc:0.950312\ttrain-auc:0.958778\n",
      "[235]\teval-auc:0.950323\ttrain-auc:0.958915\n",
      "[236]\teval-auc:0.950317\ttrain-auc:0.958982\n",
      "[237]\teval-auc:0.950304\ttrain-auc:0.959063\n",
      "[238]\teval-auc:0.950324\ttrain-auc:0.959093\n",
      "[239]\teval-auc:0.950333\ttrain-auc:0.959105\n",
      "[240]\teval-auc:0.950381\ttrain-auc:0.959194\n",
      "[241]\teval-auc:0.950327\ttrain-auc:0.959221\n",
      "[242]\teval-auc:0.950414\ttrain-auc:0.959323\n",
      "[243]\teval-auc:0.950498\ttrain-auc:0.959434\n",
      "[244]\teval-auc:0.950598\ttrain-auc:0.959624\n",
      "[245]\teval-auc:0.950549\ttrain-auc:0.959663\n",
      "[246]\teval-auc:0.950588\ttrain-auc:0.959742\n",
      "[247]\teval-auc:0.950637\ttrain-auc:0.959831\n",
      "[248]\teval-auc:0.950629\ttrain-auc:0.959902\n",
      "[249]\teval-auc:0.950653\ttrain-auc:0.959959\n",
      "[250]\teval-auc:0.95067\ttrain-auc:0.960087\n",
      "[251]\teval-auc:0.950639\ttrain-auc:0.960135\n",
      "[252]\teval-auc:0.950623\ttrain-auc:0.960163\n",
      "[253]\teval-auc:0.950668\ttrain-auc:0.960255\n",
      "[254]\teval-auc:0.950676\ttrain-auc:0.96033\n",
      "[255]\teval-auc:0.950801\ttrain-auc:0.960382\n",
      "[256]\teval-auc:0.950938\ttrain-auc:0.96044\n",
      "[257]\teval-auc:0.951025\ttrain-auc:0.960528\n",
      "[258]\teval-auc:0.951171\ttrain-auc:0.960607\n",
      "[259]\teval-auc:0.951259\ttrain-auc:0.960688\n",
      "[260]\teval-auc:0.951231\ttrain-auc:0.960722\n",
      "[261]\teval-auc:0.951285\ttrain-auc:0.960875\n",
      "[262]\teval-auc:0.951289\ttrain-auc:0.960923\n",
      "[263]\teval-auc:0.951292\ttrain-auc:0.960994\n",
      "[264]\teval-auc:0.951359\ttrain-auc:0.96107\n",
      "[265]\teval-auc:0.951416\ttrain-auc:0.961209\n",
      "[266]\teval-auc:0.951452\ttrain-auc:0.961284\n",
      "[267]\teval-auc:0.951483\ttrain-auc:0.961387\n",
      "[268]\teval-auc:0.951522\ttrain-auc:0.961434\n",
      "[269]\teval-auc:0.951514\ttrain-auc:0.961551\n",
      "[270]\teval-auc:0.951478\ttrain-auc:0.961682\n",
      "[271]\teval-auc:0.951512\ttrain-auc:0.961853\n",
      "[272]\teval-auc:0.951559\ttrain-auc:0.961993\n",
      "[273]\teval-auc:0.95149\ttrain-auc:0.96207\n",
      "[274]\teval-auc:0.951464\ttrain-auc:0.962082\n",
      "[275]\teval-auc:0.951498\ttrain-auc:0.962134\n",
      "[276]\teval-auc:0.951485\ttrain-auc:0.962148\n",
      "[277]\teval-auc:0.951571\ttrain-auc:0.962234\n",
      "[278]\teval-auc:0.951531\ttrain-auc:0.962322\n",
      "[279]\teval-auc:0.95149\ttrain-auc:0.962442\n",
      "[280]\teval-auc:0.951601\ttrain-auc:0.962595\n",
      "[281]\teval-auc:0.951537\ttrain-auc:0.962672\n",
      "[282]\teval-auc:0.95157\ttrain-auc:0.962777\n",
      "[283]\teval-auc:0.951543\ttrain-auc:0.962806\n",
      "[284]\teval-auc:0.951568\ttrain-auc:0.962869\n",
      "[285]\teval-auc:0.951574\ttrain-auc:0.962987\n",
      "[286]\teval-auc:0.951554\ttrain-auc:0.963056\n",
      "[287]\teval-auc:0.951547\ttrain-auc:0.963148\n",
      "[288]\teval-auc:0.951599\ttrain-auc:0.963224\n",
      "[289]\teval-auc:0.951684\ttrain-auc:0.963371\n",
      "[290]\teval-auc:0.951626\ttrain-auc:0.963452\n",
      "[291]\teval-auc:0.951765\ttrain-auc:0.963609\n",
      "[292]\teval-auc:0.951774\ttrain-auc:0.96363\n",
      "[293]\teval-auc:0.951747\ttrain-auc:0.963729\n",
      "[294]\teval-auc:0.951766\ttrain-auc:0.963757\n",
      "[295]\teval-auc:0.95176\ttrain-auc:0.963791\n",
      "[296]\teval-auc:0.9518\ttrain-auc:0.963865\n",
      "[297]\teval-auc:0.951871\ttrain-auc:0.963961\n",
      "[298]\teval-auc:0.95188\ttrain-auc:0.963988\n",
      "[299]\teval-auc:0.951871\ttrain-auc:0.964064\n",
      "[300]\teval-auc:0.951889\ttrain-auc:0.964113\n",
      "[301]\teval-auc:0.951894\ttrain-auc:0.964229\n",
      "[302]\teval-auc:0.951972\ttrain-auc:0.964364\n",
      "[303]\teval-auc:0.952022\ttrain-auc:0.964419\n",
      "[304]\teval-auc:0.952054\ttrain-auc:0.964469\n",
      "[305]\teval-auc:0.952018\ttrain-auc:0.964513\n",
      "[306]\teval-auc:0.952116\ttrain-auc:0.964635\n",
      "[307]\teval-auc:0.952117\ttrain-auc:0.964651\n",
      "[308]\teval-auc:0.952147\ttrain-auc:0.964717\n",
      "[309]\teval-auc:0.95214\ttrain-auc:0.964775\n",
      "[310]\teval-auc:0.952148\ttrain-auc:0.964815\n",
      "[311]\teval-auc:0.952196\ttrain-auc:0.96488\n",
      "[312]\teval-auc:0.95222\ttrain-auc:0.964938\n",
      "[313]\teval-auc:0.952213\ttrain-auc:0.96498\n",
      "[314]\teval-auc:0.95223\ttrain-auc:0.964999\n",
      "[315]\teval-auc:0.952356\ttrain-auc:0.965075\n",
      "[316]\teval-auc:0.952347\ttrain-auc:0.965084\n",
      "[317]\teval-auc:0.952759\ttrain-auc:0.965304\n",
      "[318]\teval-auc:0.95273\ttrain-auc:0.965375\n",
      "[319]\teval-auc:0.952739\ttrain-auc:0.965427\n",
      "[320]\teval-auc:0.952743\ttrain-auc:0.965442\n",
      "[321]\teval-auc:0.952771\ttrain-auc:0.965515\n",
      "[322]\teval-auc:0.952782\ttrain-auc:0.965564\n",
      "[323]\teval-auc:0.952815\ttrain-auc:0.965629\n",
      "[324]\teval-auc:0.952863\ttrain-auc:0.965708\n",
      "[325]\teval-auc:0.952814\ttrain-auc:0.965735\n",
      "[326]\teval-auc:0.952883\ttrain-auc:0.96583\n",
      "[327]\teval-auc:0.95289\ttrain-auc:0.9659\n",
      "[328]\teval-auc:0.952908\ttrain-auc:0.965924\n",
      "[329]\teval-auc:0.952864\ttrain-auc:0.966048\n",
      "[330]\teval-auc:0.952906\ttrain-auc:0.966127\n",
      "[331]\teval-auc:0.952871\ttrain-auc:0.966177\n",
      "[332]\teval-auc:0.952852\ttrain-auc:0.966265\n",
      "[333]\teval-auc:0.952902\ttrain-auc:0.966285\n",
      "[334]\teval-auc:0.952884\ttrain-auc:0.966298\n",
      "[335]\teval-auc:0.95291\ttrain-auc:0.966395\n",
      "[336]\teval-auc:0.952908\ttrain-auc:0.966506\n",
      "[337]\teval-auc:0.952943\ttrain-auc:0.966569\n",
      "[338]\teval-auc:0.952968\ttrain-auc:0.966639\n",
      "[339]\teval-auc:0.952965\ttrain-auc:0.966691\n",
      "[340]\teval-auc:0.952968\ttrain-auc:0.966754\n",
      "[341]\teval-auc:0.952999\ttrain-auc:0.966795\n",
      "[342]\teval-auc:0.953008\ttrain-auc:0.966887\n",
      "[343]\teval-auc:0.952993\ttrain-auc:0.966961\n",
      "[344]\teval-auc:0.952969\ttrain-auc:0.967061\n",
      "[345]\teval-auc:0.952971\ttrain-auc:0.967091\n",
      "[346]\teval-auc:0.95297\ttrain-auc:0.967127\n",
      "[347]\teval-auc:0.95299\ttrain-auc:0.967191\n",
      "[348]\teval-auc:0.953037\ttrain-auc:0.967245\n",
      "[349]\teval-auc:0.953029\ttrain-auc:0.967279\n",
      "[350]\teval-auc:0.953044\ttrain-auc:0.967319\n",
      "[351]\teval-auc:0.953036\ttrain-auc:0.967331\n",
      "[352]\teval-auc:0.953024\ttrain-auc:0.967396\n",
      "[353]\teval-auc:0.953077\ttrain-auc:0.967443\n",
      "[354]\teval-auc:0.953089\ttrain-auc:0.967503\n",
      "[355]\teval-auc:0.953133\ttrain-auc:0.967577\n",
      "[356]\teval-auc:0.953117\ttrain-auc:0.967689\n",
      "[357]\teval-auc:0.953126\ttrain-auc:0.967725\n",
      "[358]\teval-auc:0.953079\ttrain-auc:0.967807\n",
      "[359]\teval-auc:0.953065\ttrain-auc:0.967883\n",
      "[360]\teval-auc:0.953077\ttrain-auc:0.967964\n",
      "[361]\teval-auc:0.953078\ttrain-auc:0.968054\n",
      "[362]\teval-auc:0.953068\ttrain-auc:0.968052\n",
      "[363]\teval-auc:0.953018\ttrain-auc:0.968132\n",
      "[364]\teval-auc:0.95302\ttrain-auc:0.968188\n",
      "[365]\teval-auc:0.953014\ttrain-auc:0.968248\n",
      "[366]\teval-auc:0.953013\ttrain-auc:0.968297\n",
      "[367]\teval-auc:0.953026\ttrain-auc:0.968355\n",
      "[368]\teval-auc:0.953037\ttrain-auc:0.9684\n",
      "[369]\teval-auc:0.952988\ttrain-auc:0.968505\n",
      "[370]\teval-auc:0.95296\ttrain-auc:0.968524\n",
      "[371]\teval-auc:0.952936\ttrain-auc:0.968569\n",
      "[372]\teval-auc:0.953019\ttrain-auc:0.968651\n",
      "[373]\teval-auc:0.952998\ttrain-auc:0.968679\n",
      "[374]\teval-auc:0.953011\ttrain-auc:0.96869\n",
      "[375]\teval-auc:0.953024\ttrain-auc:0.968731\n",
      "[376]\teval-auc:0.953047\ttrain-auc:0.968809\n",
      "[377]\teval-auc:0.953065\ttrain-auc:0.968876\n",
      "[378]\teval-auc:0.953054\ttrain-auc:0.968931\n",
      "[379]\teval-auc:0.953117\ttrain-auc:0.969032\n",
      "[380]\teval-auc:0.95312\ttrain-auc:0.969081\n",
      "[381]\teval-auc:0.953262\ttrain-auc:0.969238\n",
      "[382]\teval-auc:0.953259\ttrain-auc:0.969294\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[383]\teval-auc:0.953228\ttrain-auc:0.969319\n",
      "[384]\teval-auc:0.953198\ttrain-auc:0.9694\n",
      "[385]\teval-auc:0.953225\ttrain-auc:0.969429\n",
      "[386]\teval-auc:0.953301\ttrain-auc:0.969552\n",
      "[387]\teval-auc:0.95325\ttrain-auc:0.969585\n",
      "[388]\teval-auc:0.953165\ttrain-auc:0.969644\n",
      "[389]\teval-auc:0.953188\ttrain-auc:0.969689\n",
      "[390]\teval-auc:0.953208\ttrain-auc:0.969778\n",
      "[391]\teval-auc:0.953237\ttrain-auc:0.969806\n",
      "[392]\teval-auc:0.95324\ttrain-auc:0.969896\n",
      "[393]\teval-auc:0.953243\ttrain-auc:0.96994\n",
      "[394]\teval-auc:0.953275\ttrain-auc:0.97\n",
      "[395]\teval-auc:0.953286\ttrain-auc:0.970028\n",
      "[396]\teval-auc:0.953256\ttrain-auc:0.97012\n",
      "[397]\teval-auc:0.953262\ttrain-auc:0.970133\n",
      "[398]\teval-auc:0.953219\ttrain-auc:0.970164\n",
      "[399]\teval-auc:0.95324\ttrain-auc:0.970183\n",
      "[400]\teval-auc:0.953258\ttrain-auc:0.970215\n",
      "[401]\teval-auc:0.953234\ttrain-auc:0.970222\n",
      "[402]\teval-auc:0.953209\ttrain-auc:0.9703\n",
      "[403]\teval-auc:0.95318\ttrain-auc:0.970418\n",
      "[404]\teval-auc:0.953231\ttrain-auc:0.970495\n",
      "[405]\teval-auc:0.953268\ttrain-auc:0.970548\n",
      "[406]\teval-auc:0.953263\ttrain-auc:0.97058\n",
      "[407]\teval-auc:0.95327\ttrain-auc:0.970599\n",
      "[408]\teval-auc:0.953246\ttrain-auc:0.970649\n",
      "[409]\teval-auc:0.953217\ttrain-auc:0.970772\n",
      "[410]\teval-auc:0.953167\ttrain-auc:0.970852\n",
      "[411]\teval-auc:0.953183\ttrain-auc:0.970945\n",
      "[412]\teval-auc:0.95317\ttrain-auc:0.97096\n",
      "[413]\teval-auc:0.9532\ttrain-auc:0.970998\n",
      "[414]\teval-auc:0.953163\ttrain-auc:0.971015\n",
      "[415]\teval-auc:0.953173\ttrain-auc:0.971097\n",
      "[416]\teval-auc:0.953198\ttrain-auc:0.971134\n",
      "[417]\teval-auc:0.953256\ttrain-auc:0.971163\n",
      "[418]\teval-auc:0.953287\ttrain-auc:0.971223\n",
      "[419]\teval-auc:0.953279\ttrain-auc:0.971232\n",
      "[420]\teval-auc:0.953234\ttrain-auc:0.971256\n",
      "[421]\teval-auc:0.95323\ttrain-auc:0.971281\n",
      "[422]\teval-auc:0.953206\ttrain-auc:0.971323\n",
      "[423]\teval-auc:0.953223\ttrain-auc:0.971373\n",
      "[424]\teval-auc:0.953228\ttrain-auc:0.971418\n",
      "[425]\teval-auc:0.953275\ttrain-auc:0.971467\n",
      "[426]\teval-auc:0.95329\ttrain-auc:0.971473\n",
      "[427]\teval-auc:0.953322\ttrain-auc:0.971512\n",
      "[428]\teval-auc:0.953283\ttrain-auc:0.971545\n",
      "[429]\teval-auc:0.953271\ttrain-auc:0.971567\n",
      "[430]\teval-auc:0.953248\ttrain-auc:0.971593\n",
      "[431]\teval-auc:0.953257\ttrain-auc:0.97162\n",
      "[432]\teval-auc:0.95323\ttrain-auc:0.971648\n",
      "[433]\teval-auc:0.953253\ttrain-auc:0.971689\n",
      "[434]\teval-auc:0.953242\ttrain-auc:0.971743\n",
      "[435]\teval-auc:0.953271\ttrain-auc:0.97178\n",
      "[436]\teval-auc:0.953514\ttrain-auc:0.971908\n",
      "[437]\teval-auc:0.953468\ttrain-auc:0.971953\n",
      "[438]\teval-auc:0.953499\ttrain-auc:0.971967\n",
      "[439]\teval-auc:0.95367\ttrain-auc:0.972069\n",
      "[440]\teval-auc:0.953664\ttrain-auc:0.97212\n",
      "[441]\teval-auc:0.953637\ttrain-auc:0.972147\n",
      "[442]\teval-auc:0.953622\ttrain-auc:0.972158\n",
      "[443]\teval-auc:0.953697\ttrain-auc:0.972267\n",
      "[444]\teval-auc:0.953689\ttrain-auc:0.972303\n",
      "[445]\teval-auc:0.953824\ttrain-auc:0.972451\n",
      "[446]\teval-auc:0.953891\ttrain-auc:0.972536\n",
      "[447]\teval-auc:0.95388\ttrain-auc:0.972547\n",
      "[448]\teval-auc:0.95387\ttrain-auc:0.97257\n",
      "[449]\teval-auc:0.953877\ttrain-auc:0.972613\n",
      "[450]\teval-auc:0.953836\ttrain-auc:0.972654\n",
      "[451]\teval-auc:0.953842\ttrain-auc:0.972693\n",
      "[452]\teval-auc:0.95387\ttrain-auc:0.972733\n",
      "[453]\teval-auc:0.95388\ttrain-auc:0.972788\n",
      "[454]\teval-auc:0.953859\ttrain-auc:0.972817\n",
      "[455]\teval-auc:0.953833\ttrain-auc:0.972893\n",
      "[456]\teval-auc:0.953836\ttrain-auc:0.972943\n",
      "[457]\teval-auc:0.953832\ttrain-auc:0.973009\n",
      "[458]\teval-auc:0.953799\ttrain-auc:0.973048\n",
      "[459]\teval-auc:0.953821\ttrain-auc:0.973085\n",
      "[460]\teval-auc:0.953876\ttrain-auc:0.973123\n",
      "[461]\teval-auc:0.953873\ttrain-auc:0.973172\n",
      "[462]\teval-auc:0.953886\ttrain-auc:0.973179\n",
      "[463]\teval-auc:0.953901\ttrain-auc:0.973225\n",
      "[464]\teval-auc:0.953937\ttrain-auc:0.973251\n",
      "[465]\teval-auc:0.953905\ttrain-auc:0.973267\n",
      "[466]\teval-auc:0.953879\ttrain-auc:0.973279\n",
      "[467]\teval-auc:0.953923\ttrain-auc:0.973341\n",
      "[468]\teval-auc:0.953873\ttrain-auc:0.973387\n",
      "[469]\teval-auc:0.953854\ttrain-auc:0.973405\n",
      "[470]\teval-auc:0.953817\ttrain-auc:0.973441\n",
      "[471]\teval-auc:0.953758\ttrain-auc:0.973493\n",
      "[472]\teval-auc:0.953774\ttrain-auc:0.973497\n",
      "[473]\teval-auc:0.953796\ttrain-auc:0.973557\n",
      "[474]\teval-auc:0.953788\ttrain-auc:0.973559\n",
      "[475]\teval-auc:0.953802\ttrain-auc:0.973579\n",
      "[476]\teval-auc:0.953927\ttrain-auc:0.973647\n",
      "[477]\teval-auc:0.953895\ttrain-auc:0.973666\n",
      "[478]\teval-auc:0.953948\ttrain-auc:0.9737\n",
      "[479]\teval-auc:0.953935\ttrain-auc:0.973746\n",
      "[480]\teval-auc:0.953949\ttrain-auc:0.973779\n",
      "[481]\teval-auc:0.953934\ttrain-auc:0.973822\n",
      "[482]\teval-auc:0.953895\ttrain-auc:0.973859\n",
      "[483]\teval-auc:0.953868\ttrain-auc:0.973882\n",
      "[484]\teval-auc:0.953865\ttrain-auc:0.973925\n",
      "[485]\teval-auc:0.95382\ttrain-auc:0.973957\n",
      "[486]\teval-auc:0.953825\ttrain-auc:0.974\n",
      "[487]\teval-auc:0.953817\ttrain-auc:0.974037\n",
      "[488]\teval-auc:0.953792\ttrain-auc:0.974051\n",
      "[489]\teval-auc:0.953804\ttrain-auc:0.974057\n",
      "[490]\teval-auc:0.953777\ttrain-auc:0.974082\n",
      "[491]\teval-auc:0.95378\ttrain-auc:0.974112\n",
      "[492]\teval-auc:0.953785\ttrain-auc:0.974132\n",
      "[493]\teval-auc:0.953779\ttrain-auc:0.974188\n",
      "[494]\teval-auc:0.953775\ttrain-auc:0.974226\n",
      "[495]\teval-auc:0.9538\ttrain-auc:0.974269\n",
      "[496]\teval-auc:0.953827\ttrain-auc:0.974335\n",
      "[497]\teval-auc:0.953818\ttrain-auc:0.974344\n",
      "[498]\teval-auc:0.953797\ttrain-auc:0.974365\n",
      "[499]\teval-auc:0.953797\ttrain-auc:0.974434\n",
      "[500]\teval-auc:0.953762\ttrain-auc:0.974462\n",
      "[501]\teval-auc:0.95375\ttrain-auc:0.974492\n",
      "[502]\teval-auc:0.953747\ttrain-auc:0.974508\n",
      "[503]\teval-auc:0.953757\ttrain-auc:0.974542\n",
      "[504]\teval-auc:0.953764\ttrain-auc:0.974557\n",
      "[505]\teval-auc:0.953802\ttrain-auc:0.974614\n",
      "[506]\teval-auc:0.953838\ttrain-auc:0.974641\n",
      "[507]\teval-auc:0.953834\ttrain-auc:0.974702\n",
      "[508]\teval-auc:0.953818\ttrain-auc:0.974734\n",
      "[509]\teval-auc:0.953831\ttrain-auc:0.974797\n",
      "[510]\teval-auc:0.953798\ttrain-auc:0.974832\n",
      "[511]\teval-auc:0.953794\ttrain-auc:0.974839\n",
      "[512]\teval-auc:0.953836\ttrain-auc:0.97486\n",
      "[513]\teval-auc:0.953823\ttrain-auc:0.974908\n",
      "[514]\teval-auc:0.953805\ttrain-auc:0.974921\n",
      "[515]\teval-auc:0.953817\ttrain-auc:0.975011\n",
      "[516]\teval-auc:0.953831\ttrain-auc:0.975043\n",
      "[517]\teval-auc:0.953796\ttrain-auc:0.975073\n",
      "[518]\teval-auc:0.953759\ttrain-auc:0.975103\n",
      "[519]\teval-auc:0.953833\ttrain-auc:0.975173\n",
      "[520]\teval-auc:0.953834\ttrain-auc:0.975223\n",
      "[521]\teval-auc:0.953939\ttrain-auc:0.975237\n",
      "[522]\teval-auc:0.953916\ttrain-auc:0.975272\n",
      "[523]\teval-auc:0.953911\ttrain-auc:0.975287\n",
      "[524]\teval-auc:0.953905\ttrain-auc:0.975286\n",
      "[525]\teval-auc:0.953895\ttrain-auc:0.975325\n",
      "[526]\teval-auc:0.953872\ttrain-auc:0.97536\n",
      "[527]\teval-auc:0.953847\ttrain-auc:0.975376\n",
      "[528]\teval-auc:0.953807\ttrain-auc:0.975392\n",
      "[529]\teval-auc:0.953794\ttrain-auc:0.975442\n",
      "[530]\teval-auc:0.953772\ttrain-auc:0.975533\n",
      "[531]\teval-auc:0.953768\ttrain-auc:0.975538\n",
      "[532]\teval-auc:0.953771\ttrain-auc:0.975556\n",
      "[533]\teval-auc:0.953778\ttrain-auc:0.975595\n",
      "[534]\teval-auc:0.953771\ttrain-auc:0.975603\n",
      "[535]\teval-auc:0.953812\ttrain-auc:0.975673\n",
      "[536]\teval-auc:0.953816\ttrain-auc:0.97568\n",
      "[537]\teval-auc:0.953838\ttrain-auc:0.975715\n",
      "[538]\teval-auc:0.953823\ttrain-auc:0.975756\n",
      "[539]\teval-auc:0.953824\ttrain-auc:0.975768\n",
      "[540]\teval-auc:0.953844\ttrain-auc:0.975819\n",
      "[541]\teval-auc:0.953847\ttrain-auc:0.975844\n",
      "[542]\teval-auc:0.953833\ttrain-auc:0.975872\n",
      "[543]\teval-auc:0.953797\ttrain-auc:0.975896\n",
      "[544]\teval-auc:0.953811\ttrain-auc:0.975927\n",
      "[545]\teval-auc:0.95383\ttrain-auc:0.97596\n",
      "[546]\teval-auc:0.95384\ttrain-auc:0.976002\n",
      "[547]\teval-auc:0.953808\ttrain-auc:0.976052\n",
      "[548]\teval-auc:0.95378\ttrain-auc:0.976112\n",
      "[549]\teval-auc:0.95377\ttrain-auc:0.976135\n",
      "[550]\teval-auc:0.953737\ttrain-auc:0.97616\n",
      "[551]\teval-auc:0.95373\ttrain-auc:0.97618\n",
      "[552]\teval-auc:0.953743\ttrain-auc:0.976234\n",
      "[553]\teval-auc:0.953752\ttrain-auc:0.976245\n",
      "[554]\teval-auc:0.953735\ttrain-auc:0.9763\n",
      "[555]\teval-auc:0.953784\ttrain-auc:0.976355\n",
      "[556]\teval-auc:0.95376\ttrain-auc:0.976365\n",
      "[557]\teval-auc:0.953798\ttrain-auc:0.976429\n",
      "[558]\teval-auc:0.95377\ttrain-auc:0.976476\n",
      "[559]\teval-auc:0.953796\ttrain-auc:0.976498\n",
      "[560]\teval-auc:0.953752\ttrain-auc:0.976529\n",
      "[561]\teval-auc:0.953741\ttrain-auc:0.976564\n",
      "[562]\teval-auc:0.953742\ttrain-auc:0.976563\n",
      "[563]\teval-auc:0.953733\ttrain-auc:0.976576\n",
      "[564]\teval-auc:0.953739\ttrain-auc:0.976619\n",
      "[565]\teval-auc:0.953728\ttrain-auc:0.976647\n",
      "[566]\teval-auc:0.95374\ttrain-auc:0.976671\n",
      "[567]\teval-auc:0.95375\ttrain-auc:0.976721\n",
      "[568]\teval-auc:0.95373\ttrain-auc:0.976747\n",
      "[569]\teval-auc:0.953703\ttrain-auc:0.976822\n",
      "[570]\teval-auc:0.953686\ttrain-auc:0.976837\n",
      "[571]\teval-auc:0.953635\ttrain-auc:0.97688\n",
      "[572]\teval-auc:0.953628\ttrain-auc:0.976892\n",
      "[573]\teval-auc:0.953603\ttrain-auc:0.976929\n",
      "[574]\teval-auc:0.953591\ttrain-auc:0.976955\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[575]\teval-auc:0.953565\ttrain-auc:0.976971\n",
      "[576]\teval-auc:0.953596\ttrain-auc:0.977006\n",
      "[577]\teval-auc:0.953559\ttrain-auc:0.977043\n",
      "[578]\teval-auc:0.953518\ttrain-auc:0.977078\n",
      "[579]\teval-auc:0.953514\ttrain-auc:0.977138\n",
      "[580]\teval-auc:0.953505\ttrain-auc:0.977145\n",
      "[581]\teval-auc:0.953483\ttrain-auc:0.977189\n",
      "[582]\teval-auc:0.953478\ttrain-auc:0.977199\n",
      "[583]\teval-auc:0.95347\ttrain-auc:0.977206\n",
      "[584]\teval-auc:0.953409\ttrain-auc:0.977248\n",
      "[585]\teval-auc:0.953408\ttrain-auc:0.977291\n",
      "[586]\teval-auc:0.953398\ttrain-auc:0.977317\n",
      "[587]\teval-auc:0.953358\ttrain-auc:0.977393\n",
      "[588]\teval-auc:0.95334\ttrain-auc:0.977408\n",
      "[589]\teval-auc:0.953346\ttrain-auc:0.977448\n",
      "[590]\teval-auc:0.953385\ttrain-auc:0.977472\n",
      "[591]\teval-auc:0.953398\ttrain-auc:0.977503\n",
      "[592]\teval-auc:0.953386\ttrain-auc:0.97753\n",
      "[593]\teval-auc:0.953374\ttrain-auc:0.977566\n",
      "[594]\teval-auc:0.953362\ttrain-auc:0.977577\n",
      "[595]\teval-auc:0.953312\ttrain-auc:0.977625\n",
      "[596]\teval-auc:0.95336\ttrain-auc:0.977647\n",
      "[597]\teval-auc:0.953364\ttrain-auc:0.977655\n",
      "[598]\teval-auc:0.953397\ttrain-auc:0.977659\n",
      "[599]\teval-auc:0.953408\ttrain-auc:0.977693\n",
      "[600]\teval-auc:0.953408\ttrain-auc:0.977734\n",
      "[601]\teval-auc:0.953401\ttrain-auc:0.977797\n",
      "[602]\teval-auc:0.95343\ttrain-auc:0.977845\n",
      "[603]\teval-auc:0.953427\ttrain-auc:0.97788\n",
      "[604]\teval-auc:0.953417\ttrain-auc:0.977925\n",
      "[605]\teval-auc:0.953379\ttrain-auc:0.977953\n",
      "[606]\teval-auc:0.953358\ttrain-auc:0.977975\n",
      "[607]\teval-auc:0.953364\ttrain-auc:0.978004\n",
      "[608]\teval-auc:0.953383\ttrain-auc:0.978036\n",
      "[609]\teval-auc:0.953334\ttrain-auc:0.978087\n",
      "[610]\teval-auc:0.953317\ttrain-auc:0.978144\n",
      "[611]\teval-auc:0.953307\ttrain-auc:0.978212\n",
      "[612]\teval-auc:0.953418\ttrain-auc:0.978235\n",
      "[613]\teval-auc:0.953438\ttrain-auc:0.978247\n",
      "[614]\teval-auc:0.95345\ttrain-auc:0.978293\n",
      "[615]\teval-auc:0.953428\ttrain-auc:0.978313\n",
      "[616]\teval-auc:0.953404\ttrain-auc:0.978348\n",
      "[617]\teval-auc:0.953416\ttrain-auc:0.978376\n",
      "[618]\teval-auc:0.953451\ttrain-auc:0.978427\n",
      "[619]\teval-auc:0.953398\ttrain-auc:0.978455\n",
      "[620]\teval-auc:0.953412\ttrain-auc:0.978475\n",
      "[621]\teval-auc:0.953418\ttrain-auc:0.9785\n",
      "[622]\teval-auc:0.953415\ttrain-auc:0.978552\n",
      "[623]\teval-auc:0.953422\ttrain-auc:0.978577\n",
      "[624]\teval-auc:0.953407\ttrain-auc:0.978621\n",
      "[625]\teval-auc:0.953462\ttrain-auc:0.978668\n",
      "[626]\teval-auc:0.953445\ttrain-auc:0.978697\n",
      "[627]\teval-auc:0.953488\ttrain-auc:0.978684\n",
      "[628]\teval-auc:0.953484\ttrain-auc:0.97872\n",
      "[629]\teval-auc:0.953497\ttrain-auc:0.978729\n",
      "[630]\teval-auc:0.953503\ttrain-auc:0.978754\n",
      "[631]\teval-auc:0.953538\ttrain-auc:0.978776\n",
      "[632]\teval-auc:0.953555\ttrain-auc:0.978815\n",
      "[633]\teval-auc:0.953536\ttrain-auc:0.978868\n",
      "[634]\teval-auc:0.953527\ttrain-auc:0.978888\n",
      "[635]\teval-auc:0.953489\ttrain-auc:0.978921\n",
      "[636]\teval-auc:0.953478\ttrain-auc:0.978923\n",
      "[637]\teval-auc:0.953462\ttrain-auc:0.978955\n",
      "[638]\teval-auc:0.953428\ttrain-auc:0.978997\n",
      "[639]\teval-auc:0.953433\ttrain-auc:0.979038\n",
      "[640]\teval-auc:0.953429\ttrain-auc:0.979088\n",
      "[641]\teval-auc:0.953404\ttrain-auc:0.979123\n",
      "[642]\teval-auc:0.953402\ttrain-auc:0.979136\n",
      "[643]\teval-auc:0.953381\ttrain-auc:0.979151\n",
      "[644]\teval-auc:0.953367\ttrain-auc:0.979183\n",
      "[645]\teval-auc:0.953383\ttrain-auc:0.979211\n",
      "[646]\teval-auc:0.953471\ttrain-auc:0.979245\n",
      "[647]\teval-auc:0.953463\ttrain-auc:0.979277\n",
      "[648]\teval-auc:0.953466\ttrain-auc:0.979292\n",
      "[649]\teval-auc:0.953477\ttrain-auc:0.979305\n",
      "[650]\teval-auc:0.953488\ttrain-auc:0.979332\n",
      "[651]\teval-auc:0.953481\ttrain-auc:0.979352\n",
      "[652]\teval-auc:0.953504\ttrain-auc:0.979364\n",
      "[653]\teval-auc:0.953504\ttrain-auc:0.979395\n",
      "[654]\teval-auc:0.953531\ttrain-auc:0.979399\n",
      "[655]\teval-auc:0.953524\ttrain-auc:0.979456\n",
      "[656]\teval-auc:0.953524\ttrain-auc:0.979468\n",
      "[657]\teval-auc:0.953481\ttrain-auc:0.979493\n",
      "[658]\teval-auc:0.953365\ttrain-auc:0.979542\n",
      "[659]\teval-auc:0.953329\ttrain-auc:0.979557\n",
      "[660]\teval-auc:0.953367\ttrain-auc:0.979606\n",
      "[661]\teval-auc:0.953392\ttrain-auc:0.97965\n",
      "[662]\teval-auc:0.953417\ttrain-auc:0.979675\n",
      "[663]\teval-auc:0.953425\ttrain-auc:0.979702\n",
      "[664]\teval-auc:0.953424\ttrain-auc:0.979736\n",
      "[665]\teval-auc:0.953413\ttrain-auc:0.979748\n",
      "[666]\teval-auc:0.953463\ttrain-auc:0.979784\n",
      "[667]\teval-auc:0.953538\ttrain-auc:0.979827\n",
      "[668]\teval-auc:0.953491\ttrain-auc:0.979855\n",
      "[669]\teval-auc:0.953508\ttrain-auc:0.979888\n",
      "[670]\teval-auc:0.953469\ttrain-auc:0.979929\n",
      "[671]\teval-auc:0.953443\ttrain-auc:0.979964\n",
      "[672]\teval-auc:0.953447\ttrain-auc:0.979977\n",
      "[673]\teval-auc:0.953438\ttrain-auc:0.980006\n",
      "[674]\teval-auc:0.953494\ttrain-auc:0.980032\n",
      "[675]\teval-auc:0.953479\ttrain-auc:0.980048\n",
      "[676]\teval-auc:0.953461\ttrain-auc:0.980063\n",
      "[677]\teval-auc:0.953466\ttrain-auc:0.980086\n",
      "[678]\teval-auc:0.953454\ttrain-auc:0.98011\n",
      "[679]\teval-auc:0.953446\ttrain-auc:0.980148\n",
      "[680]\teval-auc:0.953488\ttrain-auc:0.980185\n",
      "[681]\teval-auc:0.95349\ttrain-auc:0.980236\n",
      "[682]\teval-auc:0.953462\ttrain-auc:0.980258\n",
      "[683]\teval-auc:0.953428\ttrain-auc:0.980301\n",
      "[684]\teval-auc:0.953429\ttrain-auc:0.980326\n",
      "[685]\teval-auc:0.953473\ttrain-auc:0.980359\n",
      "[686]\teval-auc:0.953397\ttrain-auc:0.980394\n",
      "[687]\teval-auc:0.953394\ttrain-auc:0.980403\n",
      "[688]\teval-auc:0.953399\ttrain-auc:0.980411\n",
      "[689]\teval-auc:0.953394\ttrain-auc:0.980465\n",
      "[690]\teval-auc:0.953361\ttrain-auc:0.980499\n",
      "[691]\teval-auc:0.953321\ttrain-auc:0.98052\n",
      "[692]\teval-auc:0.953228\ttrain-auc:0.980571\n",
      "[693]\teval-auc:0.95322\ttrain-auc:0.980598\n",
      "[694]\teval-auc:0.953207\ttrain-auc:0.980604\n",
      "[695]\teval-auc:0.9532\ttrain-auc:0.980644\n",
      "[696]\teval-auc:0.953196\ttrain-auc:0.980667\n",
      "[697]\teval-auc:0.953188\ttrain-auc:0.980696\n",
      "[698]\teval-auc:0.953192\ttrain-auc:0.980711\n",
      "[699]\teval-auc:0.953193\ttrain-auc:0.980723\n",
      "[700]\teval-auc:0.953166\ttrain-auc:0.980766\n",
      "[701]\teval-auc:0.953157\ttrain-auc:0.980806\n",
      "[702]\teval-auc:0.953156\ttrain-auc:0.980838\n",
      "[703]\teval-auc:0.953141\ttrain-auc:0.980881\n",
      "[704]\teval-auc:0.95312\ttrain-auc:0.980904\n",
      "[705]\teval-auc:0.953111\ttrain-auc:0.980909\n",
      "[706]\teval-auc:0.953118\ttrain-auc:0.980932\n",
      "[707]\teval-auc:0.953148\ttrain-auc:0.980968\n",
      "[708]\teval-auc:0.953134\ttrain-auc:0.981001\n",
      "[709]\teval-auc:0.953173\ttrain-auc:0.981032\n",
      "[710]\teval-auc:0.953165\ttrain-auc:0.981047\n",
      "[711]\teval-auc:0.953192\ttrain-auc:0.981062\n",
      "[712]\teval-auc:0.953159\ttrain-auc:0.981082\n",
      "[713]\teval-auc:0.953154\ttrain-auc:0.981089\n",
      "[714]\teval-auc:0.953175\ttrain-auc:0.981104\n",
      "[715]\teval-auc:0.95317\ttrain-auc:0.981111\n",
      "[716]\teval-auc:0.953184\ttrain-auc:0.981125\n",
      "[717]\teval-auc:0.953174\ttrain-auc:0.981176\n",
      "[718]\teval-auc:0.953183\ttrain-auc:0.98119\n",
      "[719]\teval-auc:0.953194\ttrain-auc:0.981226\n"
     ]
    }
   ],
   "source": [
    "import xgboost as xgb\n",
    "print ('start running ....')\n",
    "dtrain = xgb.DMatrix(x_train,label=y_train)\n",
    "dval = xgb.DMatrix(x_val,label=y_val)\n",
    "param = {'learning_rate' : 0.1,\n",
    "        'n_estimators': 1000,\n",
    "        'max_depth': 4,\n",
    "        'min_child_weight': 7,\n",
    "        'gamma': 0,\n",
    "        'subsample': 0.94,\n",
    "        'colsample_bytree': 0.75,\n",
    "        'eta': 0.05,\n",
    "        'silent': 1,\n",
    "        'objective':\n",
    "        'binary:logistic',\n",
    "        'scale_pos_weight':1}\n",
    "\n",
    "num_round =666\n",
    "plst = list(param.items())\n",
    "plst += [('eval_metric', 'auc')]\n",
    "evallist = [(dval, 'eval'), (dtrain, 'train')]\n",
    "bst=xgb.train(plst,dtrain,num_round,evallist,early_stopping_rounds=100)\n",
    "dtest = xgb.DMatrix(test_data)\n",
    "y = bst.predict(dtest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\ttrain-auc:0.76467\n",
      "Will train until train-auc hasn't improved in 100 rounds.\n",
      "[1]\ttrain-auc:0.818515\n",
      "[2]\ttrain-auc:0.822543\n",
      "[3]\ttrain-auc:0.824663\n",
      "[4]\ttrain-auc:0.83195\n",
      "[5]\ttrain-auc:0.840617\n",
      "[6]\ttrain-auc:0.847946\n",
      "[7]\ttrain-auc:0.848859\n",
      "[8]\ttrain-auc:0.858868\n",
      "[9]\ttrain-auc:0.860146\n",
      "[10]\ttrain-auc:0.864873\n",
      "[11]\ttrain-auc:0.8643\n",
      "[12]\ttrain-auc:0.870293\n",
      "[13]\ttrain-auc:0.871633\n",
      "[14]\ttrain-auc:0.870347\n",
      "[15]\ttrain-auc:0.871358\n",
      "[16]\ttrain-auc:0.876579\n",
      "[17]\ttrain-auc:0.876169\n",
      "[18]\ttrain-auc:0.878922\n",
      "[19]\ttrain-auc:0.882755\n",
      "[20]\ttrain-auc:0.884615\n",
      "[21]\ttrain-auc:0.889024\n",
      "[22]\ttrain-auc:0.890116\n",
      "[23]\ttrain-auc:0.891941\n",
      "[24]\ttrain-auc:0.895097\n",
      "[25]\ttrain-auc:0.896698\n",
      "[26]\ttrain-auc:0.898044\n",
      "[27]\ttrain-auc:0.898656\n",
      "[28]\ttrain-auc:0.899306\n",
      "[29]\ttrain-auc:0.900659\n",
      "[30]\ttrain-auc:0.9045\n",
      "[31]\ttrain-auc:0.904845\n",
      "[32]\ttrain-auc:0.907877\n",
      "[33]\ttrain-auc:0.909377\n",
      "[34]\ttrain-auc:0.910258\n",
      "[35]\ttrain-auc:0.911349\n",
      "[36]\ttrain-auc:0.912024\n",
      "[37]\ttrain-auc:0.913673\n",
      "[38]\ttrain-auc:0.914259\n",
      "[39]\ttrain-auc:0.914563\n",
      "[40]\ttrain-auc:0.91541\n",
      "[41]\ttrain-auc:0.916355\n",
      "[42]\ttrain-auc:0.917361\n",
      "[43]\ttrain-auc:0.918064\n",
      "[44]\ttrain-auc:0.918534\n",
      "[45]\ttrain-auc:0.919923\n",
      "[46]\ttrain-auc:0.920197\n",
      "[47]\ttrain-auc:0.920905\n",
      "[48]\ttrain-auc:0.922221\n",
      "[49]\ttrain-auc:0.922777\n",
      "[50]\ttrain-auc:0.923441\n",
      "[51]\ttrain-auc:0.924025\n",
      "[52]\ttrain-auc:0.924365\n",
      "[53]\ttrain-auc:0.925345\n",
      "[54]\ttrain-auc:0.925955\n",
      "[55]\ttrain-auc:0.926814\n",
      "[56]\ttrain-auc:0.927296\n",
      "[57]\ttrain-auc:0.927984\n",
      "[58]\ttrain-auc:0.928255\n",
      "[59]\ttrain-auc:0.929625\n",
      "[60]\ttrain-auc:0.929931\n",
      "[61]\ttrain-auc:0.93036\n",
      "[62]\ttrain-auc:0.930334\n",
      "[63]\ttrain-auc:0.930493\n",
      "[64]\ttrain-auc:0.931016\n",
      "[65]\ttrain-auc:0.93156\n",
      "[66]\ttrain-auc:0.93182\n",
      "[67]\ttrain-auc:0.932017\n",
      "[68]\ttrain-auc:0.933355\n",
      "[69]\ttrain-auc:0.933568\n",
      "[70]\ttrain-auc:0.934116\n",
      "[71]\ttrain-auc:0.934697\n",
      "[72]\ttrain-auc:0.934855\n",
      "[73]\ttrain-auc:0.93504\n",
      "[74]\ttrain-auc:0.93528\n",
      "[75]\ttrain-auc:0.935443\n",
      "[76]\ttrain-auc:0.935537\n",
      "[77]\ttrain-auc:0.935834\n",
      "[78]\ttrain-auc:0.935975\n",
      "[79]\ttrain-auc:0.935932\n",
      "[80]\ttrain-auc:0.93614\n",
      "[81]\ttrain-auc:0.936589\n",
      "[82]\ttrain-auc:0.937126\n",
      "[83]\ttrain-auc:0.937299\n",
      "[84]\ttrain-auc:0.937541\n",
      "[85]\ttrain-auc:0.937703\n",
      "[86]\ttrain-auc:0.938795\n",
      "[87]\ttrain-auc:0.938841\n",
      "[88]\ttrain-auc:0.939272\n",
      "[89]\ttrain-auc:0.939632\n",
      "[90]\ttrain-auc:0.939712\n",
      "[91]\ttrain-auc:0.939927\n",
      "[92]\ttrain-auc:0.940192\n",
      "[93]\ttrain-auc:0.940314\n",
      "[94]\ttrain-auc:0.940852\n",
      "[95]\ttrain-auc:0.941195\n",
      "[96]\ttrain-auc:0.941411\n",
      "[97]\ttrain-auc:0.941521\n",
      "[98]\ttrain-auc:0.941519\n",
      "[99]\ttrain-auc:0.941586\n",
      "[100]\ttrain-auc:0.941771\n",
      "[101]\ttrain-auc:0.942037\n",
      "[102]\ttrain-auc:0.942241\n",
      "[103]\ttrain-auc:0.942445\n",
      "[104]\ttrain-auc:0.942616\n",
      "[105]\ttrain-auc:0.942671\n",
      "[106]\ttrain-auc:0.942762\n",
      "[107]\ttrain-auc:0.942921\n",
      "[108]\ttrain-auc:0.942992\n",
      "[109]\ttrain-auc:0.943036\n",
      "[110]\ttrain-auc:0.943178\n",
      "[111]\ttrain-auc:0.943274\n",
      "[112]\ttrain-auc:0.94349\n",
      "[113]\ttrain-auc:0.943645\n",
      "[114]\ttrain-auc:0.943813\n",
      "[115]\ttrain-auc:0.944192\n",
      "[116]\ttrain-auc:0.944261\n",
      "[117]\ttrain-auc:0.944311\n",
      "[118]\ttrain-auc:0.944696\n",
      "[119]\ttrain-auc:0.944989\n",
      "[120]\ttrain-auc:0.945178\n",
      "[121]\ttrain-auc:0.945395\n",
      "[122]\ttrain-auc:0.945461\n",
      "[123]\ttrain-auc:0.945714\n",
      "[124]\ttrain-auc:0.945765\n",
      "[125]\ttrain-auc:0.945944\n",
      "[126]\ttrain-auc:0.946055\n",
      "[127]\ttrain-auc:0.94612\n",
      "[128]\ttrain-auc:0.946317\n",
      "[129]\ttrain-auc:0.946434\n",
      "[130]\ttrain-auc:0.946614\n",
      "[131]\ttrain-auc:0.947112\n",
      "[132]\ttrain-auc:0.947226\n",
      "[133]\ttrain-auc:0.947428\n",
      "[134]\ttrain-auc:0.947496\n",
      "[135]\ttrain-auc:0.947652\n",
      "[136]\ttrain-auc:0.947683\n",
      "[137]\ttrain-auc:0.947831\n",
      "[138]\ttrain-auc:0.947926\n",
      "[139]\ttrain-auc:0.947948\n",
      "[140]\ttrain-auc:0.948066\n",
      "[141]\ttrain-auc:0.948077\n",
      "[142]\ttrain-auc:0.948127\n",
      "[143]\ttrain-auc:0.94824\n",
      "[144]\ttrain-auc:0.948382\n",
      "[145]\ttrain-auc:0.948448\n",
      "[146]\ttrain-auc:0.948801\n",
      "[147]\ttrain-auc:0.94903\n",
      "[148]\ttrain-auc:0.949111\n",
      "[149]\ttrain-auc:0.949178\n",
      "[150]\ttrain-auc:0.94922\n",
      "[151]\ttrain-auc:0.949314\n",
      "[152]\ttrain-auc:0.949442\n",
      "[153]\ttrain-auc:0.949549\n",
      "[154]\ttrain-auc:0.949929\n",
      "[155]\ttrain-auc:0.949991\n",
      "[156]\ttrain-auc:0.950228\n",
      "[157]\ttrain-auc:0.950279\n",
      "[158]\ttrain-auc:0.950338\n",
      "[159]\ttrain-auc:0.950415\n",
      "[160]\ttrain-auc:0.950481\n",
      "[161]\ttrain-auc:0.950532\n",
      "[162]\ttrain-auc:0.950641\n",
      "[163]\ttrain-auc:0.950751\n",
      "[164]\ttrain-auc:0.950802\n",
      "[165]\ttrain-auc:0.950914\n",
      "[166]\ttrain-auc:0.951053\n",
      "[167]\ttrain-auc:0.951219\n",
      "[168]\ttrain-auc:0.951263\n",
      "[169]\ttrain-auc:0.951359\n",
      "[170]\ttrain-auc:0.951487\n",
      "[171]\ttrain-auc:0.95166\n",
      "[172]\ttrain-auc:0.951804\n",
      "[173]\ttrain-auc:0.95193\n",
      "[174]\ttrain-auc:0.952026\n",
      "[175]\ttrain-auc:0.952153\n",
      "[176]\ttrain-auc:0.952309\n",
      "[177]\ttrain-auc:0.952422\n",
      "[178]\ttrain-auc:0.952598\n",
      "[179]\ttrain-auc:0.952737\n",
      "[180]\ttrain-auc:0.952784\n",
      "[181]\ttrain-auc:0.952881\n",
      "[182]\ttrain-auc:0.952958\n",
      "[183]\ttrain-auc:0.953142\n",
      "[184]\ttrain-auc:0.953253\n",
      "[185]\ttrain-auc:0.953286\n",
      "[186]\ttrain-auc:0.953369\n",
      "[187]\ttrain-auc:0.953454\n",
      "[188]\ttrain-auc:0.953498\n",
      "[189]\ttrain-auc:0.953673\n",
      "[190]\ttrain-auc:0.953828\n",
      "[191]\ttrain-auc:0.953998\n",
      "[192]\ttrain-auc:0.95405\n",
      "[193]\ttrain-auc:0.954133\n",
      "[194]\ttrain-auc:0.9542\n",
      "[195]\ttrain-auc:0.954301\n",
      "[196]\ttrain-auc:0.954449\n",
      "[197]\ttrain-auc:0.954499\n",
      "[198]\ttrain-auc:0.954593\n",
      "[199]\ttrain-auc:0.954727\n",
      "[200]\ttrain-auc:0.954885\n",
      "[201]\ttrain-auc:0.954918\n",
      "[202]\ttrain-auc:0.954956\n",
      "[203]\ttrain-auc:0.955087\n",
      "[204]\ttrain-auc:0.955111\n",
      "[205]\ttrain-auc:0.955291\n",
      "[206]\ttrain-auc:0.95529\n",
      "[207]\ttrain-auc:0.955343\n",
      "[208]\ttrain-auc:0.955377\n",
      "[209]\ttrain-auc:0.955474\n",
      "[210]\ttrain-auc:0.955598\n",
      "[211]\ttrain-auc:0.955689\n",
      "[212]\ttrain-auc:0.955812\n",
      "[213]\ttrain-auc:0.955902\n",
      "[214]\ttrain-auc:0.956003\n",
      "[215]\ttrain-auc:0.956102\n",
      "[216]\ttrain-auc:0.956167\n",
      "[217]\ttrain-auc:0.956246\n",
      "[218]\ttrain-auc:0.956267\n",
      "[219]\ttrain-auc:0.956413\n",
      "[220]\ttrain-auc:0.956573\n",
      "[221]\ttrain-auc:0.956772\n",
      "[222]\ttrain-auc:0.95689\n",
      "[223]\ttrain-auc:0.956941\n",
      "[224]\ttrain-auc:0.95702\n",
      "[225]\ttrain-auc:0.957041\n",
      "[226]\ttrain-auc:0.957106\n",
      "[227]\ttrain-auc:0.957196\n",
      "[228]\ttrain-auc:0.957232\n",
      "[229]\ttrain-auc:0.957304\n",
      "[230]\ttrain-auc:0.957343\n",
      "[231]\ttrain-auc:0.957355\n",
      "[232]\ttrain-auc:0.957425\n",
      "[233]\ttrain-auc:0.957512\n",
      "[234]\ttrain-auc:0.957563\n",
      "[235]\ttrain-auc:0.957723\n",
      "[236]\ttrain-auc:0.957825\n",
      "[237]\ttrain-auc:0.957937\n",
      "[238]\ttrain-auc:0.957966\n",
      "[239]\ttrain-auc:0.957993\n",
      "[240]\ttrain-auc:0.958073\n",
      "[241]\ttrain-auc:0.958128\n",
      "[242]\ttrain-auc:0.958201\n",
      "[243]\ttrain-auc:0.958299\n",
      "[244]\ttrain-auc:0.958309\n",
      "[245]\ttrain-auc:0.958351\n",
      "[246]\ttrain-auc:0.958432\n",
      "[247]\ttrain-auc:0.958543\n",
      "[248]\ttrain-auc:0.958712\n",
      "[249]\ttrain-auc:0.958778\n",
      "[250]\ttrain-auc:0.958848\n",
      "[251]\ttrain-auc:0.958879\n",
      "[252]\ttrain-auc:0.958948\n",
      "[253]\ttrain-auc:0.958983\n",
      "[254]\ttrain-auc:0.959102\n",
      "[255]\ttrain-auc:0.959156\n",
      "[256]\ttrain-auc:0.959223\n",
      "[257]\ttrain-auc:0.959265\n",
      "[258]\ttrain-auc:0.959399\n",
      "[259]\ttrain-auc:0.959445\n",
      "[260]\ttrain-auc:0.959583\n",
      "[261]\ttrain-auc:0.959627\n",
      "[262]\ttrain-auc:0.959654\n",
      "[263]\ttrain-auc:0.959684\n",
      "[264]\ttrain-auc:0.959775\n",
      "[265]\ttrain-auc:0.959847\n",
      "[266]\ttrain-auc:0.959865\n",
      "[267]\ttrain-auc:0.959932\n",
      "[268]\ttrain-auc:0.960025\n",
      "[269]\ttrain-auc:0.960052\n",
      "[270]\ttrain-auc:0.960115\n",
      "[271]\ttrain-auc:0.960131\n",
      "[272]\ttrain-auc:0.960234\n",
      "[273]\ttrain-auc:0.960245\n",
      "[274]\ttrain-auc:0.96032\n",
      "[275]\ttrain-auc:0.960366\n",
      "[276]\ttrain-auc:0.96042\n",
      "[277]\ttrain-auc:0.960501\n",
      "[278]\ttrain-auc:0.960572\n",
      "[279]\ttrain-auc:0.960684\n",
      "[280]\ttrain-auc:0.960859\n",
      "[281]\ttrain-auc:0.96087\n",
      "[282]\ttrain-auc:0.960928\n",
      "[283]\ttrain-auc:0.960993\n",
      "[284]\ttrain-auc:0.961031\n",
      "[285]\ttrain-auc:0.961122\n",
      "[286]\ttrain-auc:0.961238\n",
      "[287]\ttrain-auc:0.961281\n",
      "[288]\ttrain-auc:0.961367\n",
      "[289]\ttrain-auc:0.961464\n",
      "[290]\ttrain-auc:0.961545\n",
      "[291]\ttrain-auc:0.961619\n",
      "[292]\ttrain-auc:0.961689\n",
      "[293]\ttrain-auc:0.961798\n",
      "[294]\ttrain-auc:0.961828\n",
      "[295]\ttrain-auc:0.961878\n",
      "[296]\ttrain-auc:0.961958\n",
      "[297]\ttrain-auc:0.961992\n",
      "[298]\ttrain-auc:0.962085\n",
      "[299]\ttrain-auc:0.962164\n",
      "[300]\ttrain-auc:0.962175\n",
      "[301]\ttrain-auc:0.962216\n",
      "[302]\ttrain-auc:0.962312\n",
      "[303]\ttrain-auc:0.962373\n",
      "[304]\ttrain-auc:0.962434\n",
      "[305]\ttrain-auc:0.962473\n",
      "[306]\ttrain-auc:0.962493\n",
      "[307]\ttrain-auc:0.962528\n",
      "[308]\ttrain-auc:0.962603\n",
      "[309]\ttrain-auc:0.962658\n",
      "[310]\ttrain-auc:0.962713\n",
      "[311]\ttrain-auc:0.962759\n",
      "[312]\ttrain-auc:0.962845\n",
      "[313]\ttrain-auc:0.962955\n",
      "[314]\ttrain-auc:0.963015\n",
      "[315]\ttrain-auc:0.96311\n",
      "[316]\ttrain-auc:0.963158\n",
      "[317]\ttrain-auc:0.963212\n",
      "[318]\ttrain-auc:0.963274\n",
      "[319]\ttrain-auc:0.963361\n",
      "[320]\ttrain-auc:0.96342\n",
      "[321]\ttrain-auc:0.963437\n",
      "[322]\ttrain-auc:0.963505\n",
      "[323]\ttrain-auc:0.963604\n",
      "[324]\ttrain-auc:0.963639\n",
      "[325]\ttrain-auc:0.963751\n",
      "[326]\ttrain-auc:0.963978\n",
      "[327]\ttrain-auc:0.964006\n",
      "[328]\ttrain-auc:0.964032\n",
      "[329]\ttrain-auc:0.964063\n",
      "[330]\ttrain-auc:0.964079\n",
      "[331]\ttrain-auc:0.964117\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[332]\ttrain-auc:0.964125\n",
      "[333]\ttrain-auc:0.964246\n",
      "[334]\ttrain-auc:0.964282\n",
      "[335]\ttrain-auc:0.964313\n",
      "[336]\ttrain-auc:0.964416\n",
      "[337]\ttrain-auc:0.964471\n",
      "[338]\ttrain-auc:0.964531\n",
      "[339]\ttrain-auc:0.964576\n",
      "[340]\ttrain-auc:0.964629\n",
      "[341]\ttrain-auc:0.964721\n",
      "[342]\ttrain-auc:0.964804\n",
      "[343]\ttrain-auc:0.964876\n",
      "[344]\ttrain-auc:0.964884\n",
      "[345]\ttrain-auc:0.964973\n",
      "[346]\ttrain-auc:0.965083\n",
      "[347]\ttrain-auc:0.96515\n",
      "[348]\ttrain-auc:0.965194\n",
      "[349]\ttrain-auc:0.965208\n",
      "[350]\ttrain-auc:0.965252\n",
      "[351]\ttrain-auc:0.965306\n",
      "[352]\ttrain-auc:0.965365\n",
      "[353]\ttrain-auc:0.965376\n",
      "[354]\ttrain-auc:0.965395\n",
      "[355]\ttrain-auc:0.965475\n",
      "[356]\ttrain-auc:0.965557\n",
      "[357]\ttrain-auc:0.965589\n",
      "[358]\ttrain-auc:0.965645\n",
      "[359]\ttrain-auc:0.965679\n",
      "[360]\ttrain-auc:0.965701\n",
      "[361]\ttrain-auc:0.965748\n",
      "[362]\ttrain-auc:0.965811\n",
      "[363]\ttrain-auc:0.965887\n",
      "[364]\ttrain-auc:0.965904\n",
      "[365]\ttrain-auc:0.965965\n",
      "[366]\ttrain-auc:0.966003\n",
      "[367]\ttrain-auc:0.966025\n",
      "[368]\ttrain-auc:0.966086\n",
      "[369]\ttrain-auc:0.966124\n",
      "[370]\ttrain-auc:0.966198\n",
      "[371]\ttrain-auc:0.966205\n",
      "[372]\ttrain-auc:0.96626\n",
      "[373]\ttrain-auc:0.966278\n",
      "[374]\ttrain-auc:0.966302\n",
      "[375]\ttrain-auc:0.966314\n",
      "[376]\ttrain-auc:0.966366\n",
      "[377]\ttrain-auc:0.966401\n",
      "[378]\ttrain-auc:0.966414\n",
      "[379]\ttrain-auc:0.966517\n",
      "[380]\ttrain-auc:0.96658\n",
      "[381]\ttrain-auc:0.96662\n",
      "[382]\ttrain-auc:0.966676\n",
      "[383]\ttrain-auc:0.966718\n",
      "[384]\ttrain-auc:0.966775\n",
      "[385]\ttrain-auc:0.966813\n",
      "[386]\ttrain-auc:0.966881\n",
      "[387]\ttrain-auc:0.966926\n",
      "[388]\ttrain-auc:0.966975\n",
      "[389]\ttrain-auc:0.967027\n",
      "[390]\ttrain-auc:0.967043\n",
      "[391]\ttrain-auc:0.967138\n",
      "[392]\ttrain-auc:0.96717\n",
      "[393]\ttrain-auc:0.967221\n",
      "[394]\ttrain-auc:0.967301\n",
      "[395]\ttrain-auc:0.967343\n",
      "[396]\ttrain-auc:0.967422\n",
      "[397]\ttrain-auc:0.967464\n",
      "[398]\ttrain-auc:0.967506\n",
      "[399]\ttrain-auc:0.967553\n",
      "[400]\ttrain-auc:0.967591\n",
      "[401]\ttrain-auc:0.967614\n",
      "[402]\ttrain-auc:0.96768\n",
      "[403]\ttrain-auc:0.967725\n",
      "[404]\ttrain-auc:0.967788\n",
      "[405]\ttrain-auc:0.967798\n",
      "[406]\ttrain-auc:0.967822\n",
      "[407]\ttrain-auc:0.967885\n",
      "[408]\ttrain-auc:0.967942\n",
      "[409]\ttrain-auc:0.967953\n",
      "[410]\ttrain-auc:0.968016\n",
      "[411]\ttrain-auc:0.968109\n",
      "[412]\ttrain-auc:0.968127\n",
      "[413]\ttrain-auc:0.968171\n",
      "[414]\ttrain-auc:0.968198\n",
      "[415]\ttrain-auc:0.968242\n",
      "[416]\ttrain-auc:0.96829\n",
      "[417]\ttrain-auc:0.968329\n",
      "[418]\ttrain-auc:0.968366\n",
      "[419]\ttrain-auc:0.968429\n",
      "[420]\ttrain-auc:0.968477\n",
      "[421]\ttrain-auc:0.968508\n",
      "[422]\ttrain-auc:0.968534\n",
      "[423]\ttrain-auc:0.968549\n",
      "[424]\ttrain-auc:0.968609\n",
      "[425]\ttrain-auc:0.968664\n",
      "[426]\ttrain-auc:0.96868\n",
      "[427]\ttrain-auc:0.968731\n",
      "[428]\ttrain-auc:0.968764\n",
      "[429]\ttrain-auc:0.968803\n",
      "[430]\ttrain-auc:0.968843\n",
      "[431]\ttrain-auc:0.968857\n",
      "[432]\ttrain-auc:0.968918\n",
      "[433]\ttrain-auc:0.968971\n",
      "[434]\ttrain-auc:0.969028\n",
      "[435]\ttrain-auc:0.96912\n",
      "[436]\ttrain-auc:0.969127\n",
      "[437]\ttrain-auc:0.969188\n",
      "[438]\ttrain-auc:0.969207\n",
      "[439]\ttrain-auc:0.969256\n",
      "[440]\ttrain-auc:0.969275\n",
      "[441]\ttrain-auc:0.969332\n",
      "[442]\ttrain-auc:0.969355\n",
      "[443]\ttrain-auc:0.969378\n",
      "[444]\ttrain-auc:0.969399\n",
      "[445]\ttrain-auc:0.969461\n",
      "[446]\ttrain-auc:0.969515\n",
      "[447]\ttrain-auc:0.969555\n",
      "[448]\ttrain-auc:0.969563\n",
      "[449]\ttrain-auc:0.969641\n",
      "[450]\ttrain-auc:0.96967\n",
      "[451]\ttrain-auc:0.969715\n",
      "[452]\ttrain-auc:0.969738\n",
      "[453]\ttrain-auc:0.969766\n",
      "[454]\ttrain-auc:0.969801\n",
      "[455]\ttrain-auc:0.969879\n",
      "[456]\ttrain-auc:0.969948\n",
      "[457]\ttrain-auc:0.969957\n",
      "[458]\ttrain-auc:0.970021\n",
      "[459]\ttrain-auc:0.970073\n",
      "[460]\ttrain-auc:0.970078\n",
      "[461]\ttrain-auc:0.970095\n",
      "[462]\ttrain-auc:0.970138\n",
      "[463]\ttrain-auc:0.970164\n",
      "[464]\ttrain-auc:0.970249\n",
      "[465]\ttrain-auc:0.970257\n",
      "[466]\ttrain-auc:0.970299\n",
      "[467]\ttrain-auc:0.970333\n",
      "[468]\ttrain-auc:0.970368\n",
      "[469]\ttrain-auc:0.970408\n",
      "[470]\ttrain-auc:0.970458\n",
      "[471]\ttrain-auc:0.970498\n",
      "[472]\ttrain-auc:0.970552\n",
      "[473]\ttrain-auc:0.970582\n",
      "[474]\ttrain-auc:0.970588\n",
      "[475]\ttrain-auc:0.970665\n",
      "[476]\ttrain-auc:0.970693\n",
      "[477]\ttrain-auc:0.970752\n",
      "[478]\ttrain-auc:0.970787\n",
      "[479]\ttrain-auc:0.970845\n",
      "[480]\ttrain-auc:0.970887\n",
      "[481]\ttrain-auc:0.970894\n",
      "[482]\ttrain-auc:0.970913\n",
      "[483]\ttrain-auc:0.970941\n",
      "[484]\ttrain-auc:0.970967\n",
      "[485]\ttrain-auc:0.970997\n",
      "[486]\ttrain-auc:0.971006\n",
      "[487]\ttrain-auc:0.971027\n",
      "[488]\ttrain-auc:0.97106\n",
      "[489]\ttrain-auc:0.971071\n",
      "[490]\ttrain-auc:0.971082\n",
      "[491]\ttrain-auc:0.97112\n",
      "[492]\ttrain-auc:0.971188\n",
      "[493]\ttrain-auc:0.971183\n",
      "[494]\ttrain-auc:0.97125\n",
      "[495]\ttrain-auc:0.971284\n",
      "[496]\ttrain-auc:0.971327\n",
      "[497]\ttrain-auc:0.971371\n",
      "[498]\ttrain-auc:0.971382\n",
      "[499]\ttrain-auc:0.971448\n",
      "[500]\ttrain-auc:0.971501\n",
      "[501]\ttrain-auc:0.97153\n",
      "[502]\ttrain-auc:0.971559\n",
      "[503]\ttrain-auc:0.971597\n",
      "[504]\ttrain-auc:0.971604\n",
      "[505]\ttrain-auc:0.971666\n",
      "[506]\ttrain-auc:0.97171\n",
      "[507]\ttrain-auc:0.971774\n",
      "[508]\ttrain-auc:0.971788\n",
      "[509]\ttrain-auc:0.971808\n",
      "[510]\ttrain-auc:0.971849\n",
      "[511]\ttrain-auc:0.971902\n",
      "[512]\ttrain-auc:0.971912\n",
      "[513]\ttrain-auc:0.971953\n",
      "[514]\ttrain-auc:0.972003\n",
      "[515]\ttrain-auc:0.972017\n",
      "[516]\ttrain-auc:0.972051\n",
      "[517]\ttrain-auc:0.972101\n",
      "[518]\ttrain-auc:0.972143\n",
      "[519]\ttrain-auc:0.972158\n",
      "[520]\ttrain-auc:0.972198\n",
      "[521]\ttrain-auc:0.972241\n",
      "[522]\ttrain-auc:0.972281\n",
      "[523]\ttrain-auc:0.972316\n",
      "[524]\ttrain-auc:0.972354\n",
      "[525]\ttrain-auc:0.972404\n",
      "[526]\ttrain-auc:0.972434\n",
      "[527]\ttrain-auc:0.972467\n",
      "[528]\ttrain-auc:0.9725\n",
      "[529]\ttrain-auc:0.97257\n",
      "[530]\ttrain-auc:0.972591\n",
      "[531]\ttrain-auc:0.972645\n",
      "[532]\ttrain-auc:0.972674\n",
      "[533]\ttrain-auc:0.972766\n",
      "[534]\ttrain-auc:0.972787\n",
      "[535]\ttrain-auc:0.972813\n",
      "[536]\ttrain-auc:0.972819\n",
      "[537]\ttrain-auc:0.972854\n",
      "[538]\ttrain-auc:0.97288\n",
      "[539]\ttrain-auc:0.9729\n",
      "[540]\ttrain-auc:0.972934\n",
      "[541]\ttrain-auc:0.972959\n",
      "[542]\ttrain-auc:0.972979\n",
      "[543]\ttrain-auc:0.973036\n",
      "[544]\ttrain-auc:0.97304\n",
      "[545]\ttrain-auc:0.973056\n",
      "[546]\ttrain-auc:0.973093\n",
      "[547]\ttrain-auc:0.973168\n",
      "[548]\ttrain-auc:0.973221\n",
      "[549]\ttrain-auc:0.973252\n",
      "[550]\ttrain-auc:0.973321\n",
      "[551]\ttrain-auc:0.973329\n",
      "[552]\ttrain-auc:0.973396\n",
      "[553]\ttrain-auc:0.973433\n",
      "[554]\ttrain-auc:0.973471\n",
      "[555]\ttrain-auc:0.973499\n",
      "[556]\ttrain-auc:0.973506\n",
      "[557]\ttrain-auc:0.973527\n",
      "[558]\ttrain-auc:0.973567\n",
      "[559]\ttrain-auc:0.973598\n",
      "[560]\ttrain-auc:0.973604\n",
      "[561]\ttrain-auc:0.973626\n",
      "[562]\ttrain-auc:0.973652\n",
      "[563]\ttrain-auc:0.973683\n",
      "[564]\ttrain-auc:0.973713\n",
      "[565]\ttrain-auc:0.973758\n",
      "[566]\ttrain-auc:0.973782\n",
      "[567]\ttrain-auc:0.973815\n",
      "[568]\ttrain-auc:0.973857\n",
      "[569]\ttrain-auc:0.973861\n",
      "[570]\ttrain-auc:0.97392\n",
      "[571]\ttrain-auc:0.973943\n",
      "[572]\ttrain-auc:0.974034\n",
      "[573]\ttrain-auc:0.974083\n",
      "[574]\ttrain-auc:0.97412\n",
      "[575]\ttrain-auc:0.974128\n",
      "[576]\ttrain-auc:0.974174\n",
      "[577]\ttrain-auc:0.974314\n",
      "[578]\ttrain-auc:0.974358\n",
      "[579]\ttrain-auc:0.974364\n",
      "[580]\ttrain-auc:0.974383\n",
      "[581]\ttrain-auc:0.974408\n",
      "[582]\ttrain-auc:0.97441\n",
      "[583]\ttrain-auc:0.974463\n",
      "[584]\ttrain-auc:0.974495\n",
      "[585]\ttrain-auc:0.9745\n",
      "[586]\ttrain-auc:0.974518\n",
      "[587]\ttrain-auc:0.974546\n",
      "[588]\ttrain-auc:0.974567\n",
      "[589]\ttrain-auc:0.974568\n",
      "[590]\ttrain-auc:0.974595\n",
      "[591]\ttrain-auc:0.974606\n",
      "[592]\ttrain-auc:0.974639\n",
      "[593]\ttrain-auc:0.974683\n",
      "[594]\ttrain-auc:0.974727\n",
      "[595]\ttrain-auc:0.974781\n",
      "[596]\ttrain-auc:0.974806\n",
      "[597]\ttrain-auc:0.97483\n",
      "[598]\ttrain-auc:0.974863\n",
      "[599]\ttrain-auc:0.974881\n",
      "[600]\ttrain-auc:0.974908\n",
      "[601]\ttrain-auc:0.974921\n",
      "[602]\ttrain-auc:0.974935\n",
      "[603]\ttrain-auc:0.974961\n",
      "[604]\ttrain-auc:0.974982\n",
      "[605]\ttrain-auc:0.975018\n",
      "[606]\ttrain-auc:0.975026\n",
      "[607]\ttrain-auc:0.975068\n",
      "[608]\ttrain-auc:0.97509\n",
      "[609]\ttrain-auc:0.975134\n",
      "[610]\ttrain-auc:0.975205\n",
      "[611]\ttrain-auc:0.975222\n",
      "[612]\ttrain-auc:0.975263\n",
      "[613]\ttrain-auc:0.975267\n",
      "[614]\ttrain-auc:0.975284\n",
      "[615]\ttrain-auc:0.975291\n",
      "[616]\ttrain-auc:0.975332\n",
      "[617]\ttrain-auc:0.97536\n",
      "[618]\ttrain-auc:0.975445\n",
      "[619]\ttrain-auc:0.975468\n",
      "[620]\ttrain-auc:0.975471\n",
      "[621]\ttrain-auc:0.975521\n",
      "[622]\ttrain-auc:0.975538\n",
      "[623]\ttrain-auc:0.975582\n",
      "[624]\ttrain-auc:0.975612\n",
      "[625]\ttrain-auc:0.975629\n",
      "[626]\ttrain-auc:0.975643\n",
      "[627]\ttrain-auc:0.975666\n",
      "[628]\ttrain-auc:0.975698\n",
      "[629]\ttrain-auc:0.975746\n",
      "[630]\ttrain-auc:0.975773\n",
      "[631]\ttrain-auc:0.975788\n",
      "[632]\ttrain-auc:0.975846\n",
      "[633]\ttrain-auc:0.97587\n",
      "[634]\ttrain-auc:0.975902\n",
      "[635]\ttrain-auc:0.975949\n",
      "[636]\ttrain-auc:0.975965\n",
      "[637]\ttrain-auc:0.976009\n",
      "[638]\ttrain-auc:0.976029\n",
      "[639]\ttrain-auc:0.976039\n",
      "[640]\ttrain-auc:0.976072\n",
      "[641]\ttrain-auc:0.976114\n",
      "[642]\ttrain-auc:0.976131\n",
      "[643]\ttrain-auc:0.976155\n",
      "[644]\ttrain-auc:0.976169\n",
      "[645]\ttrain-auc:0.976182\n",
      "[646]\ttrain-auc:0.976209\n",
      "[647]\ttrain-auc:0.976234\n",
      "[648]\ttrain-auc:0.976269\n",
      "[649]\ttrain-auc:0.976277\n",
      "[650]\ttrain-auc:0.976327\n",
      "[651]\ttrain-auc:0.976353\n",
      "[652]\ttrain-auc:0.976372\n",
      "[653]\ttrain-auc:0.976397\n",
      "[654]\ttrain-auc:0.976429\n",
      "[655]\ttrain-auc:0.976462\n",
      "[656]\ttrain-auc:0.976487\n",
      "[657]\ttrain-auc:0.976544\n",
      "[658]\ttrain-auc:0.976551\n",
      "[659]\ttrain-auc:0.976585\n",
      "[660]\ttrain-auc:0.976605\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[661]\ttrain-auc:0.97663\n",
      "[662]\ttrain-auc:0.976655\n",
      "[663]\ttrain-auc:0.976694\n",
      "[664]\ttrain-auc:0.976728\n",
      "[665]\ttrain-auc:0.976753\n",
      "[666]\ttrain-auc:0.976783\n",
      "[667]\ttrain-auc:0.976795\n",
      "[668]\ttrain-auc:0.976818\n",
      "[669]\ttrain-auc:0.976851\n",
      "[670]\ttrain-auc:0.976857\n",
      "[671]\ttrain-auc:0.97689\n",
      "[672]\ttrain-auc:0.976899\n",
      "[673]\ttrain-auc:0.976945\n",
      "[674]\ttrain-auc:0.976981\n",
      "[675]\ttrain-auc:0.977008\n",
      "[676]\ttrain-auc:0.97705\n",
      "[677]\ttrain-auc:0.977083\n",
      "[678]\ttrain-auc:0.977119\n",
      "[679]\ttrain-auc:0.977145\n",
      "[680]\ttrain-auc:0.977171\n",
      "[681]\ttrain-auc:0.9772\n",
      "[682]\ttrain-auc:0.977246\n",
      "[683]\ttrain-auc:0.977265\n",
      "[684]\ttrain-auc:0.977295\n",
      "[685]\ttrain-auc:0.977331\n",
      "[686]\ttrain-auc:0.977361\n",
      "[687]\ttrain-auc:0.977379\n",
      "[688]\ttrain-auc:0.977385\n",
      "[689]\ttrain-auc:0.977414\n",
      "[690]\ttrain-auc:0.977427\n",
      "[691]\ttrain-auc:0.977444\n",
      "[692]\ttrain-auc:0.977465\n",
      "[693]\ttrain-auc:0.977493\n",
      "[694]\ttrain-auc:0.977504\n",
      "[695]\ttrain-auc:0.97752\n",
      "[696]\ttrain-auc:0.977525\n",
      "[697]\ttrain-auc:0.977551\n",
      "[698]\ttrain-auc:0.977591\n",
      "[699]\ttrain-auc:0.977601\n",
      "[700]\ttrain-auc:0.977623\n",
      "[701]\ttrain-auc:0.977643\n",
      "[702]\ttrain-auc:0.977664\n",
      "[703]\ttrain-auc:0.977686\n",
      "[704]\ttrain-auc:0.977731\n",
      "[705]\ttrain-auc:0.977753\n",
      "[706]\ttrain-auc:0.977771\n",
      "[707]\ttrain-auc:0.977803\n",
      "[708]\ttrain-auc:0.977856\n",
      "[709]\ttrain-auc:0.977879\n",
      "[710]\ttrain-auc:0.977904\n",
      "[711]\ttrain-auc:0.977922\n",
      "[712]\ttrain-auc:0.977953\n",
      "[713]\ttrain-auc:0.978011\n",
      "[714]\ttrain-auc:0.978017\n",
      "[715]\ttrain-auc:0.978035\n",
      "[716]\ttrain-auc:0.978062\n",
      "[717]\ttrain-auc:0.978065\n",
      "[718]\ttrain-auc:0.978099\n",
      "[719]\ttrain-auc:0.978114\n",
      "[720]\ttrain-auc:0.978153\n",
      "[721]\ttrain-auc:0.978227\n",
      "[722]\ttrain-auc:0.978249\n",
      "[723]\ttrain-auc:0.97831\n",
      "[724]\ttrain-auc:0.978315\n",
      "[725]\ttrain-auc:0.978343\n",
      "[726]\ttrain-auc:0.978369\n",
      "[727]\ttrain-auc:0.978402\n",
      "[728]\ttrain-auc:0.978415\n",
      "[729]\ttrain-auc:0.97845\n",
      "[730]\ttrain-auc:0.978492\n",
      "[731]\ttrain-auc:0.978527\n",
      "[732]\ttrain-auc:0.978546\n",
      "[733]\ttrain-auc:0.978571\n",
      "[734]\ttrain-auc:0.978593\n",
      "[735]\ttrain-auc:0.978635\n",
      "[736]\ttrain-auc:0.978654\n",
      "[737]\ttrain-auc:0.978689\n",
      "[738]\ttrain-auc:0.97871\n",
      "[739]\ttrain-auc:0.978714\n",
      "[740]\ttrain-auc:0.978785\n",
      "[741]\ttrain-auc:0.978819\n",
      "[742]\ttrain-auc:0.978843\n",
      "[743]\ttrain-auc:0.9789\n",
      "[744]\ttrain-auc:0.978918\n",
      "[745]\ttrain-auc:0.978923\n",
      "[746]\ttrain-auc:0.978938\n",
      "[747]\ttrain-auc:0.978955\n",
      "[748]\ttrain-auc:0.978974\n",
      "[749]\ttrain-auc:0.978996\n",
      "[750]\ttrain-auc:0.979019\n",
      "[751]\ttrain-auc:0.979053\n",
      "[752]\ttrain-auc:0.979082\n",
      "[753]\ttrain-auc:0.979086\n",
      "[754]\ttrain-auc:0.979094\n",
      "[755]\ttrain-auc:0.979123\n",
      "[756]\ttrain-auc:0.979153\n",
      "[757]\ttrain-auc:0.979163\n",
      "[758]\ttrain-auc:0.979186\n",
      "[759]\ttrain-auc:0.979198\n",
      "[760]\ttrain-auc:0.979218\n",
      "[761]\ttrain-auc:0.97925\n",
      "[762]\ttrain-auc:0.979264\n",
      "[763]\ttrain-auc:0.979289\n",
      "[764]\ttrain-auc:0.979319\n",
      "[765]\ttrain-auc:0.979326\n",
      "[766]\ttrain-auc:0.979337\n",
      "[767]\ttrain-auc:0.979369\n",
      "[768]\ttrain-auc:0.979389\n",
      "[769]\ttrain-auc:0.979402\n",
      "[770]\ttrain-auc:0.97943\n",
      "[771]\ttrain-auc:0.979437\n",
      "[772]\ttrain-auc:0.979452\n",
      "[773]\ttrain-auc:0.979494\n",
      "[774]\ttrain-auc:0.979508\n",
      "[775]\ttrain-auc:0.979547\n",
      "[776]\ttrain-auc:0.979556\n",
      "[777]\ttrain-auc:0.979567\n",
      "[778]\ttrain-auc:0.979601\n",
      "[779]\ttrain-auc:0.979604\n",
      "[780]\ttrain-auc:0.979612\n",
      "[781]\ttrain-auc:0.979635\n",
      "[782]\ttrain-auc:0.979655\n",
      "[783]\ttrain-auc:0.979664\n",
      "[784]\ttrain-auc:0.979674\n",
      "[785]\ttrain-auc:0.979704\n",
      "[786]\ttrain-auc:0.979717\n",
      "[787]\ttrain-auc:0.979756\n",
      "[788]\ttrain-auc:0.979782\n",
      "[789]\ttrain-auc:0.979782\n",
      "[790]\ttrain-auc:0.979786\n",
      "[791]\ttrain-auc:0.97979\n",
      "[792]\ttrain-auc:0.979808\n",
      "[793]\ttrain-auc:0.979829\n",
      "[794]\ttrain-auc:0.979856\n",
      "[795]\ttrain-auc:0.979878\n",
      "[796]\ttrain-auc:0.979926\n",
      "[797]\ttrain-auc:0.979946\n",
      "[798]\ttrain-auc:0.979956\n",
      "[799]\ttrain-auc:0.979961\n",
      "[800]\ttrain-auc:0.979986\n",
      "[801]\ttrain-auc:0.980005\n",
      "[802]\ttrain-auc:0.980025\n",
      "[803]\ttrain-auc:0.980059\n",
      "[804]\ttrain-auc:0.980066\n",
      "[805]\ttrain-auc:0.98009\n",
      "[806]\ttrain-auc:0.980094\n",
      "[807]\ttrain-auc:0.98014\n",
      "[808]\ttrain-auc:0.980147\n",
      "[809]\ttrain-auc:0.98016\n",
      "[810]\ttrain-auc:0.9802\n",
      "[811]\ttrain-auc:0.98022\n",
      "[812]\ttrain-auc:0.980253\n",
      "[813]\ttrain-auc:0.980283\n",
      "[814]\ttrain-auc:0.980324\n",
      "[815]\ttrain-auc:0.980362\n",
      "[816]\ttrain-auc:0.980412\n",
      "[817]\ttrain-auc:0.980436\n",
      "[818]\ttrain-auc:0.980452\n",
      "[819]\ttrain-auc:0.980485\n",
      "[820]\ttrain-auc:0.980502\n",
      "[821]\ttrain-auc:0.980513\n",
      "[822]\ttrain-auc:0.980552\n",
      "[823]\ttrain-auc:0.980571\n",
      "[824]\ttrain-auc:0.98061\n",
      "[825]\ttrain-auc:0.980631\n",
      "[826]\ttrain-auc:0.980675\n",
      "[827]\ttrain-auc:0.980688\n",
      "[828]\ttrain-auc:0.980718\n",
      "[829]\ttrain-auc:0.980744\n",
      "[830]\ttrain-auc:0.980774\n",
      "[831]\ttrain-auc:0.9808\n",
      "[832]\ttrain-auc:0.980802\n",
      "[833]\ttrain-auc:0.980843\n",
      "[834]\ttrain-auc:0.980852\n",
      "[835]\ttrain-auc:0.980863\n",
      "[836]\ttrain-auc:0.980872\n",
      "[837]\ttrain-auc:0.980885\n",
      "[838]\ttrain-auc:0.980893\n",
      "[839]\ttrain-auc:0.980915\n",
      "[840]\ttrain-auc:0.980925\n",
      "[841]\ttrain-auc:0.980931\n",
      "[842]\ttrain-auc:0.98096\n",
      "[843]\ttrain-auc:0.980988\n",
      "[844]\ttrain-auc:0.981021\n",
      "[845]\ttrain-auc:0.981032\n",
      "[846]\ttrain-auc:0.981049\n",
      "[847]\ttrain-auc:0.981055\n",
      "[848]\ttrain-auc:0.981076\n",
      "[849]\ttrain-auc:0.981091\n",
      "[850]\ttrain-auc:0.981115\n",
      "[851]\ttrain-auc:0.981125\n",
      "[852]\ttrain-auc:0.981148\n",
      "[853]\ttrain-auc:0.981161\n",
      "[854]\ttrain-auc:0.981182\n",
      "[855]\ttrain-auc:0.981195\n",
      "[856]\ttrain-auc:0.981207\n",
      "[857]\ttrain-auc:0.981239\n",
      "[858]\ttrain-auc:0.981259\n",
      "[859]\ttrain-auc:0.981266\n",
      "[860]\ttrain-auc:0.981281\n",
      "[861]\ttrain-auc:0.981299\n",
      "[862]\ttrain-auc:0.98134\n",
      "[863]\ttrain-auc:0.981359\n",
      "[864]\ttrain-auc:0.981369\n",
      "[865]\ttrain-auc:0.981374\n",
      "[866]\ttrain-auc:0.981377\n",
      "[867]\ttrain-auc:0.981409\n",
      "[868]\ttrain-auc:0.981448\n",
      "[869]\ttrain-auc:0.981471\n",
      "[870]\ttrain-auc:0.981479\n",
      "[871]\ttrain-auc:0.981504\n",
      "[872]\ttrain-auc:0.981528\n",
      "[873]\ttrain-auc:0.981545\n",
      "[874]\ttrain-auc:0.981589\n",
      "[875]\ttrain-auc:0.981637\n",
      "[876]\ttrain-auc:0.981657\n",
      "[877]\ttrain-auc:0.98169\n",
      "[878]\ttrain-auc:0.981727\n",
      "[879]\ttrain-auc:0.981746\n",
      "[880]\ttrain-auc:0.981748\n",
      "[881]\ttrain-auc:0.981754\n",
      "[882]\ttrain-auc:0.981773\n",
      "[883]\ttrain-auc:0.981782\n",
      "[884]\ttrain-auc:0.981802\n",
      "[885]\ttrain-auc:0.981824\n",
      "[886]\ttrain-auc:0.981836\n",
      "[887]\ttrain-auc:0.981839\n",
      "[888]\ttrain-auc:0.981867\n",
      "[889]\ttrain-auc:0.981876\n",
      "[890]\ttrain-auc:0.981878\n",
      "[891]\ttrain-auc:0.981895\n",
      "[892]\ttrain-auc:0.981905\n",
      "[893]\ttrain-auc:0.981914\n",
      "[894]\ttrain-auc:0.981919\n",
      "[895]\ttrain-auc:0.981934\n",
      "[896]\ttrain-auc:0.98194\n",
      "[897]\ttrain-auc:0.98196\n",
      "[898]\ttrain-auc:0.982008\n",
      "[899]\ttrain-auc:0.982026\n",
      "[900]\ttrain-auc:0.982057\n",
      "[901]\ttrain-auc:0.982092\n",
      "[902]\ttrain-auc:0.982105\n",
      "[903]\ttrain-auc:0.982122\n",
      "[904]\ttrain-auc:0.982132\n",
      "[905]\ttrain-auc:0.982152\n",
      "[906]\ttrain-auc:0.98217\n",
      "[907]\ttrain-auc:0.982187\n",
      "[908]\ttrain-auc:0.982193\n",
      "[909]\ttrain-auc:0.982196\n",
      "[910]\ttrain-auc:0.982201\n",
      "[911]\ttrain-auc:0.982218\n",
      "[912]\ttrain-auc:0.982231\n",
      "[913]\ttrain-auc:0.982264\n",
      "[914]\ttrain-auc:0.982308\n",
      "[915]\ttrain-auc:0.982317\n",
      "[916]\ttrain-auc:0.982325\n",
      "[917]\ttrain-auc:0.982338\n",
      "[918]\ttrain-auc:0.982378\n",
      "[919]\ttrain-auc:0.982425\n",
      "[920]\ttrain-auc:0.982432\n",
      "[921]\ttrain-auc:0.982444\n",
      "[922]\ttrain-auc:0.982457\n",
      "[923]\ttrain-auc:0.982477\n",
      "[924]\ttrain-auc:0.982517\n",
      "[925]\ttrain-auc:0.982518\n",
      "[926]\ttrain-auc:0.98253\n",
      "[927]\ttrain-auc:0.982544\n",
      "[928]\ttrain-auc:0.982551\n",
      "[929]\ttrain-auc:0.982557\n",
      "[930]\ttrain-auc:0.982558\n",
      "[931]\ttrain-auc:0.982561\n",
      "[932]\ttrain-auc:0.98257\n",
      "[933]\ttrain-auc:0.982581\n",
      "[934]\ttrain-auc:0.982586\n",
      "[935]\ttrain-auc:0.982618\n",
      "[936]\ttrain-auc:0.982638\n",
      "[937]\ttrain-auc:0.982649\n",
      "[938]\ttrain-auc:0.982672\n",
      "[939]\ttrain-auc:0.982677\n",
      "[940]\ttrain-auc:0.982704\n",
      "[941]\ttrain-auc:0.982722\n",
      "[942]\ttrain-auc:0.982747\n",
      "[943]\ttrain-auc:0.982765\n",
      "[944]\ttrain-auc:0.982776\n",
      "[945]\ttrain-auc:0.982802\n",
      "[946]\ttrain-auc:0.982813\n",
      "[947]\ttrain-auc:0.982822\n",
      "[948]\ttrain-auc:0.982847\n",
      "[949]\ttrain-auc:0.982873\n",
      "[950]\ttrain-auc:0.98289\n",
      "[951]\ttrain-auc:0.982898\n",
      "[952]\ttrain-auc:0.982909\n",
      "[953]\ttrain-auc:0.982911\n",
      "[954]\ttrain-auc:0.98293\n",
      "[955]\ttrain-auc:0.98297\n",
      "[956]\ttrain-auc:0.983011\n",
      "[957]\ttrain-auc:0.983032\n",
      "[958]\ttrain-auc:0.98304\n",
      "[959]\ttrain-auc:0.983052\n",
      "[960]\ttrain-auc:0.983062\n",
      "[961]\ttrain-auc:0.983075\n",
      "[962]\ttrain-auc:0.983106\n",
      "[963]\ttrain-auc:0.983124\n",
      "[964]\ttrain-auc:0.983138\n",
      "[965]\ttrain-auc:0.983151\n",
      "[966]\ttrain-auc:0.983156\n",
      "[967]\ttrain-auc:0.983178\n",
      "[968]\ttrain-auc:0.983199\n",
      "[969]\ttrain-auc:0.983216\n",
      "[970]\ttrain-auc:0.983235\n",
      "[971]\ttrain-auc:0.98327\n",
      "[972]\ttrain-auc:0.983281\n",
      "[973]\ttrain-auc:0.983307\n",
      "[974]\ttrain-auc:0.983326\n",
      "[975]\ttrain-auc:0.983355\n",
      "[976]\ttrain-auc:0.983373\n",
      "[977]\ttrain-auc:0.983389\n",
      "[978]\ttrain-auc:0.983402\n",
      "[979]\ttrain-auc:0.98343\n",
      "[980]\ttrain-auc:0.983439\n",
      "[981]\ttrain-auc:0.983451\n",
      "[982]\ttrain-auc:0.983463\n",
      "[983]\ttrain-auc:0.983467\n",
      "[984]\ttrain-auc:0.983469\n",
      "[985]\ttrain-auc:0.983476\n",
      "[986]\ttrain-auc:0.983494\n",
      "[987]\ttrain-auc:0.983502\n",
      "[988]\ttrain-auc:0.983518\n",
      "[989]\ttrain-auc:0.983538\n",
      "[990]\ttrain-auc:0.983558\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[991]\ttrain-auc:0.983579\n",
      "[992]\ttrain-auc:0.983595\n",
      "[993]\ttrain-auc:0.983593\n",
      "[994]\ttrain-auc:0.983621\n",
      "[995]\ttrain-auc:0.983621\n",
      "[996]\ttrain-auc:0.983637\n",
      "[997]\ttrain-auc:0.983654\n",
      "[998]\ttrain-auc:0.983666\n",
      "[999]\ttrain-auc:0.983685\n"
     ]
    }
   ],
   "source": [
    "dtrain = xgb.DMatrix(train_data,label=train_label)\n",
    "# dval = xgb.DMatrix(x_val,label=y_val)\n",
    "param = {'learning_rate' : 0.1,\n",
    "        'n_estimators': 1000,\n",
    "        'max_depth': 4,\n",
    "        'min_child_weight': 7,\n",
    "        'gamma': 0.9,\n",
    "        'subsample': 0.94,\n",
    "        'colsample_bytree': 0.75,\n",
    "        'eta': 0.05,\n",
    "        'silent': 1,\n",
    "        'objective':\n",
    "        'binary:logistic',\n",
    "        'scale_pos_weight':1}\n",
    "\n",
    "num_round = 1000\n",
    "plst = list(param.items())\n",
    "plst += [('eval_metric', 'auc')]\n",
    "evallist = [(dtrain, 'train')]\n",
    "bst=xgb.train(plst,dtrain,num_round,evallist,early_stopping_rounds=100)\n",
    "dtest = xgb.DMatrix(test_data)\n",
    "y = bst.predict(dtest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "mysubmit = order_test\n",
    "mysubmit['orderType'] = y\n",
    "mysubmit.to_csv('submit/xgboost14.csv',index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'actionType_1': 357,\n",
       " 'actionType_2': 137,\n",
       " 'actionType_3': 115,\n",
       " 'actionType_4': 124,\n",
       " 'actionType_5': 380,\n",
       " 'actionType_6': 390,\n",
       " 'actionType_7': 131,\n",
       " 'actionType_8': 134,\n",
       " 'actionType_9': 92,\n",
       " 'age': 63,\n",
       " 'buy_flags': 17,\n",
       " 'city': 230,\n",
       " 'continent': 73,\n",
       " 'count': 211,\n",
       " 'flag_five': 57,\n",
       " 'flag_seven': 4,\n",
       " 'flag_six': 93,\n",
       " 'gender': 61,\n",
       " 'last_5dis': 660,\n",
       " 'last_6dis': 652,\n",
       " 'last_7dis': 187,\n",
       " 'last_8dis': 230,\n",
       " 'last_9dis': 92,\n",
       " 'last_two_day': 74,\n",
       " 'last_two_second': 521,\n",
       " 'last_type': 243,\n",
       " 'orderHistory_feat_sum': 65,\n",
       " 'province': 178,\n",
       " 'rating': 80,\n",
       " 'three_type': 217,\n",
       " 'two_three_day': 71,\n",
       " 'two_three_second': 698,\n",
       " 'two_type': 174,\n",
       " 'userid': 666}"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bst.get_fscore()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "temp = ac[ac.userid == 100000000637]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "order[order_his.userid == 100000000637]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "temp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "temp.loc[1121, 'actionType']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderid</th>\n",
       "      <th>rating</th>\n",
       "      <th>tags</th>\n",
       "      <th>commentsKeyWords</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000013</td>\n",
       "      <td>1000015</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['很','简陋','太','随便']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000000231</td>\n",
       "      <td>1000024</td>\n",
       "      <td>5.0</td>\n",
       "      <td>提前联系|耐心等候</td>\n",
       "      <td>['很','细心']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000000471</td>\n",
       "      <td>1000038</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000000637</td>\n",
       "      <td>1000040</td>\n",
       "      <td>5.0</td>\n",
       "      <td>主动热情|提前联系|举牌迎接|主动搬运行李</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000000755</td>\n",
       "      <td>1000045</td>\n",
       "      <td>1.0</td>\n",
       "      <td>未举牌服务</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderid  rating                   tags     commentsKeyWords\n",
       "0  100000000013  1000015     4.0                    NaN  ['很','简陋','太','随便']\n",
       "1  100000000231  1000024     5.0              提前联系|耐心等候           ['很','细心']\n",
       "2  100000000471  1000038     5.0                    NaN                  NaN\n",
       "3  100000000637  1000040     5.0  主动热情|提前联系|举牌迎接|主动搬运行李                  NaN\n",
       "4  100000000755  1000045     1.0                  未举牌服务                  NaN"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "comment.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderid</th>\n",
       "      <th>orderTime</th>\n",
       "      <th>orderType</th>\n",
       "      <th>city</th>\n",
       "      <th>country</th>\n",
       "      <th>continent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [userid, orderid, orderTime, orderType, city, country, continent]\n",
       "Index: []"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "order_his[order_his.orderid == 1000045]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderType</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>100000000637</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>100000000695</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>100000000867</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>100000000975</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>100000001023</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>100000001505</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>100000002199</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>100000002215</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>100000003091</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>100000003461</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>100000003611</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>100000004297</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>100000005007</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>100000005649</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>100000005739</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>100000006539</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>100000006605</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>100000007745</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>100000008119</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>100000008425</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>100000009017</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>100000009249</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>100000009481</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>100000009491</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>100000009773</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>100000009925</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>100000010101</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>100000010199</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>100000011845</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>100000012867</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40130</th>\n",
       "      <td>114862767040</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40139</th>\n",
       "      <td>114862969140</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40149</th>\n",
       "      <td>114864169646</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40159</th>\n",
       "      <td>114864367449</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40162</th>\n",
       "      <td>114864467347</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40169</th>\n",
       "      <td>114864667143</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40171</th>\n",
       "      <td>114864667440</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40189</th>\n",
       "      <td>114865153947</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40191</th>\n",
       "      <td>114865167546</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40197</th>\n",
       "      <td>114865267049</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200</th>\n",
       "      <td>114865267842</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40201</th>\n",
       "      <td>114865267940</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40207</th>\n",
       "      <td>114865367843</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40211</th>\n",
       "      <td>114865453949</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40212</th>\n",
       "      <td>114865467046</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40221</th>\n",
       "      <td>114865653740</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40225</th>\n",
       "      <td>114865667845</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40246</th>\n",
       "      <td>114868267642</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40250</th>\n",
       "      <td>114868367541</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40253</th>\n",
       "      <td>114868467442</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40254</th>\n",
       "      <td>114868467547</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40256</th>\n",
       "      <td>114868567347</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40257</th>\n",
       "      <td>114868567441</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40275</th>\n",
       "      <td>114869067447</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40277</th>\n",
       "      <td>114869267343</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40278</th>\n",
       "      <td>114869267443</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40283</th>\n",
       "      <td>114869567242</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40291</th>\n",
       "      <td>114869867143</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40292</th>\n",
       "      <td>114869867443</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40302</th>\n",
       "      <td>114999280232</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>6625 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             userid  orderType\n",
       "10     100000000637          1\n",
       "11     100000000695          1\n",
       "14     100000000867          1\n",
       "16     100000000975          1\n",
       "17     100000001023          1\n",
       "22     100000001505          1\n",
       "32     100000002199          1\n",
       "33     100000002215          1\n",
       "46     100000003091          1\n",
       "49     100000003461          1\n",
       "52     100000003611          1\n",
       "62     100000004297          1\n",
       "71     100000005007          1\n",
       "75     100000005649          1\n",
       "77     100000005739          1\n",
       "87     100000006539          1\n",
       "89     100000006605          1\n",
       "104    100000007745          1\n",
       "107    100000008119          1\n",
       "109    100000008425          1\n",
       "115    100000009017          1\n",
       "116    100000009249          1\n",
       "119    100000009481          1\n",
       "120    100000009491          1\n",
       "125    100000009773          1\n",
       "127    100000009925          1\n",
       "128    100000010101          1\n",
       "131    100000010199          1\n",
       "149    100000011845          1\n",
       "168    100000012867          1\n",
       "...             ...        ...\n",
       "40130  114862767040          1\n",
       "40139  114862969140          1\n",
       "40149  114864169646          1\n",
       "40159  114864367449          1\n",
       "40162  114864467347          1\n",
       "40169  114864667143          1\n",
       "40171  114864667440          1\n",
       "40189  114865153947          1\n",
       "40191  114865167546          1\n",
       "40197  114865267049          1\n",
       "40200  114865267842          1\n",
       "40201  114865267940          1\n",
       "40207  114865367843          1\n",
       "40211  114865453949          1\n",
       "40212  114865467046          1\n",
       "40221  114865653740          1\n",
       "40225  114865667845          1\n",
       "40246  114868267642          1\n",
       "40250  114868367541          1\n",
       "40253  114868467442          1\n",
       "40254  114868467547          1\n",
       "40256  114868567347          1\n",
       "40257  114868567441          1\n",
       "40275  114869067447          1\n",
       "40277  114869267343          1\n",
       "40278  114869267443          1\n",
       "40283  114869567242          1\n",
       "40291  114869867143          1\n",
       "40292  114869867443          1\n",
       "40302  114999280232          1\n",
       "\n",
       "[6625 rows x 2 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "order[order.orderType == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderType</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000371</td>\n",
       "      <td>0.075764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000001215</td>\n",
       "      <td>0.003399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000001445</td>\n",
       "      <td>0.011033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000001575</td>\n",
       "      <td>0.998092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000001601</td>\n",
       "      <td>0.000687</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderType\n",
       "0  100000000371   0.075764\n",
       "1  100000001215   0.003399\n",
       "2  100000001445   0.011033\n",
       "3  100000001575   0.998092\n",
       "4  100000001601   0.000687"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# rank\n",
    "\n",
    "test = pd.read_csv('submit/xgboost8.csv')\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "m = np.array(test.sort_values(by='orderType').index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10076"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "test['rank'] = m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderType</th>\n",
       "      <th>rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000371</td>\n",
       "      <td>0.075764</td>\n",
       "      <td>7665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000001215</td>\n",
       "      <td>0.003399</td>\n",
       "      <td>5854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000001445</td>\n",
       "      <td>0.011033</td>\n",
       "      <td>8279</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000001575</td>\n",
       "      <td>0.998092</td>\n",
       "      <td>7592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000001601</td>\n",
       "      <td>0.000687</td>\n",
       "      <td>6938</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderType  rank\n",
       "0  100000000371   0.075764  7665\n",
       "1  100000001215   0.003399  5854\n",
       "2  100000001445   0.011033  8279\n",
       "3  100000001575   0.998092  7592\n",
       "4  100000001601   0.000687  6938"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>orderType</th>\n",
       "      <th>rank</th>\n",
       "      <th>new</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000371</td>\n",
       "      <td>0.075764</td>\n",
       "      <td>7665</td>\n",
       "      <td>0.760794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000001215</td>\n",
       "      <td>0.003399</td>\n",
       "      <td>5854</td>\n",
       "      <td>0.581042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000001445</td>\n",
       "      <td>0.011033</td>\n",
       "      <td>8279</td>\n",
       "      <td>0.821737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000001575</td>\n",
       "      <td>0.998092</td>\n",
       "      <td>7592</td>\n",
       "      <td>0.753548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000001601</td>\n",
       "      <td>0.000687</td>\n",
       "      <td>6938</td>\n",
       "      <td>0.688635</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid  orderType  rank       new\n",
       "0  100000000371   0.075764  7665  0.760794\n",
       "1  100000001215   0.003399  5854  0.581042\n",
       "2  100000001445   0.011033  8279  0.821737\n",
       "3  100000001575   0.998092  7592  0.753548\n",
       "4  100000001601   0.000687  6938  0.688635"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test['new'] = test['rank'] / 10075\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "del test['orderType']\n",
    "del test['rank']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userid</th>\n",
       "      <th>new</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000000371</td>\n",
       "      <td>0.760794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100000001215</td>\n",
       "      <td>0.581042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000001445</td>\n",
       "      <td>0.821737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100000001575</td>\n",
       "      <td>0.753548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100000001601</td>\n",
       "      <td>0.688635</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         userid       new\n",
       "0  100000000371  0.760794\n",
       "1  100000001215  0.581042\n",
       "2  100000001445  0.821737\n",
       "3  100000001575  0.753548\n",
       "4  100000001601  0.688635"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "test.rename(columns={'new':'orderType'}, inplace = True)\n",
    "test.to_csv('submit/xg8-rank.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}