5.TEXT_SIM/5.3.1.XGboost.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# XGboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "\n",
    "import json\n",
    "\n",
    "from sklearn.model_selection  import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 설정된 값들"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_IN_PATH = './data_in/'\n",
    "DATA_OUT_PATH = './data_out/'\n",
    "\n",
    "TRAIN_Q1_DATA_FILE = 'train_q1.npy'\n",
    "TRAIN_Q2_DATA_FILE = 'train_q2.npy'\n",
    "TRAIN_LABEL_DATA_FILE = 'train_label.npy'\n",
    "\n",
    "\n",
    "# 훈련 데이터 가져오는 부분이다.\n",
    "train_q1_data = np.load(open(DATA_IN_PATH + TRAIN_Q1_DATA_FILE, 'rb'))\n",
    "train_q2_data = np.load(open(DATA_IN_PATH + TRAIN_Q2_DATA_FILE, 'rb'))\n",
    "train_labels = np.load(open(DATA_IN_PATH + TRAIN_LABEL_DATA_FILE, 'rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_input = np.stack((train_q1_data, train_q2_data), axis=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(train_input.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 훈련 셋과 평가 셋 나누기¶"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_input, eval_input, train_label, eval_label = train_test_split(train_input, train_labels, test_size=0.2, random_state=4242)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 모델 구성"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = xgb.DMatrix(train_input.sum(axis=1), label=train_label) # 학습 데이터 읽어 오기\n",
    "eval_data = xgb.DMatrix(eval_input.sum(axis=1), label=eval_label) # 평가 데이터 읽어 오기\n",
    "\n",
    "data_list = [(train_data, 'train'), (eval_data, 'valid')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {} # 인자를 통해 XGB모델에 넣어 주자 \n",
    "params['objective'] = 'binary:logistic' # 로지스틱 예측을 통해서 \n",
    "params['eval_metric'] = 'rmse' # root mean square error를 사용  \n",
    "\n",
    "bst = xgb.train(params, train_data, num_boost_round = 1000, evals = data_list, early_stopping_rounds=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 테스트 데이터 가져오기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "TEST_Q1_DATA_FILE = 'test_q1.npy'\n",
    "TEST_Q2_DATA_FILE = 'test_q2.npy'\n",
    "TEST_ID_DATA_FILE = 'test_id.npy'\n",
    "\n",
    "test_q1_data = np.load(open(DATA_IN_PATH + TEST_Q1_DATA_FILE, 'rb'))\n",
    "test_q2_data = np.load(open(DATA_IN_PATH + TEST_Q2_DATA_FILE, 'rb'))\n",
    "test_id_data = np.load(open(DATA_IN_PATH + TEST_ID_DATA_FILE, 'rb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 예측하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_input = np.stack((test_q1_data, test_q2_data), axis=1) \n",
    "test_data = xgb.DMatrix(test_input.sum(axis=1))\n",
    "test_predict = bst.predict(test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "if not os.path.exists(DATA_OUT_PATH):\n",
    "    os.makedirs(DATA_OUT_PATH)\n",
    "    \n",
    "output = pd.DataFrame({'test_id': test_id_data, 'is_duplicate': test_predict})\n",
    "output.to_csv(DATA_OUT_PATH + 'simple_xgb.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}