20180823 세미나 진명훈.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import matplotlib\n",
    "from matplotlib import font_manager, rc\n",
    "font_name = font_manager.FontProperties(fname=\"c:/Windows/Fonts/malgun.ttf\").get_name()\n",
    "rc('font', family=font_name)\n",
    "matplotlib.rcParams['axes.unicode_minus']=False\n",
    "\n",
    "plt.rcParams['axes.labelsize'] = 14\n",
    "plt.rcParams['xtick.labelsize'] = 12\n",
    "plt.rcParams['ytick.labelsize'] = 12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_test_split(df, train_ratio, X, y, random = False, dtrain = False):\n",
    "    \"\"\"\n",
    "    Argument\n",
    "    df : DataFrame object\n",
    "    train_ratio : float, in range(0,1)\n",
    "    X : sequence data, such as list, tuple (Train Features)\n",
    "    y : str\n",
    "    random : boolean, you need \"numpy\" library cause we use np.random.permutation function.\n",
    "    \n",
    "    Return\n",
    "    train : 학습시킬 Feature data\n",
    "    train_idx : 지도학습의 Y value\n",
    "    test : test 검증할 Feature data\n",
    "    test_idx : test set의 Y value\n",
    "    \"\"\"\n",
    "    train_size = int(len(df)*train_ratio)\n",
    "    if random:\n",
    "        shuffle_indicies = np.random.permutation(len(df))\n",
    "        train_indicies = shuffle_indicies[:train_size]\n",
    "        test_indicies = shuffle_indicies[train_size:]\n",
    "    else:\n",
    "        normal_indicies = np.arange(len(df))\n",
    "        train_indicies = normal_indicies[:train_size]\n",
    "        test_indicies = normal_indicies[train_size:]\n",
    "    if dtrain:\n",
    "        train = df.iloc[:train_size]\n",
    "        train_idx = df.iloc[:train_size]\n",
    "        test = df.iloc[train_size:]\n",
    "        test_idx = df.iloc[train_size:]\n",
    "    else:\n",
    "        train = df.iloc[:train_size][X]\n",
    "        train_idx = df.iloc[:train_size][y]\n",
    "        test = df.iloc[train_size:][X]\n",
    "        test_idx = df.iloc[train_size:][y]\n",
    "    return train, train_idx, test, test_idx\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pandas_datareader import data as pdr\n",
    "import fix_yahoo_finance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "kospi200 = pd.read_csv(\"./data/data.csv\", encoding=\"cp949\")\n",
    "kospi200 = kospi200[kospi200.columns[:2]]\n",
    "kospi200[\"종목코드\"] = kospi200[\"종목코드\"].map('{:06d}'.format)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "code_dict = { i : j for i, j in zip(kospi200[\"종목명\"], kospi200[\"종목코드\"]) }\n",
    "name_dict = { j : i for i, j in zip(kospi200[\"종목명\"], kospi200[\"종목코드\"]) }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n",
      "[*********************100%***********************]  1 of 1 downloaded\n"
     ]
    }
   ],
   "source": [
    "raw_data = {}\n",
    "for code in [i+\".KS\" for i in kospi200[\"종목코드\"]]:\n",
    "    raw_data[code[:6]] = fix_yahoo_finance.download(code, start=\"2008-01-01\", end=\"2018-08-01\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "del_list = []\n",
    "for key, df in raw_data.items():\n",
    "    if len(df) < 252*2:\n",
    "        del_list.append(key)\n",
    "for code in del_list:\n",
    "    del raw_data[code]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "mod = sys.modules[__name__]\n",
    "for code in raw_data.keys():\n",
    "    \"\"\"\n",
    "    X\n",
    "    \"\"\"\n",
    "    raw_data[code][\"pct_change\"] = raw_data[code].Close.pct_change()\n",
    "    for i in range(3, 120):\n",
    "        raw_data[code][\"close_ma_{}\".format(i)] = raw_data[code].Close.rolling(window = i).mean()\n",
    "        raw_data[code][\"volume_ma_{}\".format(i)] =  raw_data[code].Volume.rolling(window = i).mean()\n",
    "        raw_data[code][\"close_ewm_{}\".format(i)] = raw_data[code].Close.ewm(span = i).mean()\n",
    "        raw_data[code][\"volume_ewm_{}\".format(i)] = raw_data[code].Volume.ewm(span = i).mean()\n",
    "        if i in [7, 21, 30, 60, 90, 91, 100, 120]:\n",
    "            raw_data[code][\"his_vol_{}\".format(i)] = raw_data[code][\"pct_change\"].fillna(0).rolling(window = i).std()/(252**0.5)\n",
    "    \"\"\"\n",
    "    Y\n",
    "    \"\"\"\n",
    "    \n",
    "    raw_data[code] = raw_data[code].assign(Y1 = raw_data[code].Close - raw_data[code].close_ma_7.shift(1))\n",
    "    raw_data[code] = raw_data[code].assign(Y2 =\n",
    "                                           raw_data[code][\"pct_change\"] - raw_data[code][\"pct_change\"].rolling(window = 7).mean().shift(1))\n",
    "    raw_data[code] = raw_data[code].assign(Y3 = raw_data[code].Close.rolling(window=7).sum().pct_change())\n",
    "    \n",
    "    raw_data[code] = raw_data[code].assign(Y1 = np.where(raw_data[code].Y1>=0, 1, 0))\n",
    "    raw_data[code] = raw_data[code].assign(Y2 = np.where(raw_data[code].Y2>=0, 1, 0))\n",
    "    raw_data[code] = raw_data[code].assign(Y3 = np.where(raw_data[code].Y3>=0, 1, 0))\n",
    "    \"\"\"\n",
    "    Drop NaN\n",
    "    \"\"\"\n",
    "    raw_data[code] = raw_data[code].dropna(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = [x for x in df.columns if x not in ['Y1','Y2','Y3']]\n",
    "target = \"Y2\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = {}\n",
    "for code in raw_data.keys():\n",
    "    data[code] = raw_data[code].loc['2012-01-01':]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = {}\n",
    "train_y = {}\n",
    "test = {}\n",
    "test_y = {}\n",
    "for code, df in data.items():\n",
    "    X_train, y_train, X_test, y_test = train_test_split(df, 0.8, predictors, target, dtrain=True)\n",
    "    train[code] = X_train\n",
    "    train_y[code] = y_train\n",
    "    test[code] = X_test\n",
    "    test_y[code] = y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "std_scaler = StandardScaler()\n",
    "for code in raw_data.keys():\n",
    "    train[code] = std_scaler.fit_transform(train[code].values)\n",
    "    test[code] = std_scaler.fit_transform(test[code].values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost.sklearn import XGBClassifier\n",
    "#from sklearn.model_selection import cross_validation, metrics   #Additional scklearn functions\n",
    "from sklearn.model_selection import GridSearchCV   #Perforing grid search\n",
    "\n",
    "xgb_clf = XGBClassifier(random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def modelfit(alg, dtrain, predictors, target, useTrainCV = True, cv_folds = 5, early_stopping_rounds = 50):\n",
    "    \n",
    "    start = time.time()\n",
    "    \n",
    "    if useTrainCV:\n",
    "        xgb_param = alg.get_xgb_params()\n",
    "        xgtrain = xgb.DMatrix(dtrain[predictors].values, label=dtrain[target].values)\n",
    "        cvresult = xgb.cv(xgb_param, xgtrain, num_boost_round=alg.get_params()['n_estimators'], nfold=cv_folds,\n",
    "            metrics='auc', early_stopping_rounds=early_stopping_rounds, show_stdv=False)\n",
    "        alg.set_params(n_estimators=cvresult.shape[0])\n",
    "    \n",
    "    #Fit the algorithm on the data\n",
    "    alg.fit(dtrain[predictors], dtrain[target], eval_metric='auc')\n",
    "        \n",
    "    #Predict training set:\n",
    "    dtrain_predictions = alg.predict(dtrain[predictors])\n",
    "    dtrain_predprob = alg.predict_proba(dtrain[predictors])[:,1]\n",
    "        \n",
    "    #Print model report:\n",
    "    print(\"Model Report\")\n",
    "    print(\"Accuracy : %.4g\" % metrics.accuracy_score(dtrain[target].values, dtrain_predictions))\n",
    "    print(\"AUC Score (Train): %f\" % metrics.roc_auc_score(dtrain[target], dtrain_predprob))\n",
    "                    \n",
    "    imp_fear_name = pd.Series(predictors)[alg.feature_importances_ > 0.01]\n",
    "    imp_fear_name.index = range(len(imp_fear_name))\n",
    "    imp_fear = pd.concat((imp_fear_name, \n",
    "               pd.Series(alg.feature_importances_[alg.feature_importances_ > 0.01])),axis=1).sort_values(by=1,ascending=True)\n",
    "    imp_fear.index = range(len(imp_fear))\n",
    "    imp_fear.columns = [\"feature_name\", \"feature_importance\"]\n",
    "#     {i:j for i,j in zip(imp_fear.index, imp_fear[\"feature_name\"])}\n",
    "    imp_fear.plot(y='feature_importance', x='feature_name', kind='barh', legend=False, figsize=(8,5))\n",
    "    \n",
    "#     feat_imp = pd.Series(alg.feature_importances_).sort_values(ascending=False)\n",
    "#     feat_imp.plot(kind='bar', title='Feature Importances', figsize=(200,5))\n",
    "#     plt.ylabel('Feature Importance Score')\n",
    "    \n",
    "    elapse = time.time() - start\n",
    "    print(\"elapse time : {}\".format(round(elapse,2)))\n",
    "    \n",
    "    return alg, imp_fear"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "alg = XGBClassifier(\n",
    " learning_rate =0.1,\n",
    " n_estimators=1000,\n",
    " max_depth=5,\n",
    " min_child_weight=1,\n",
    " gamma=0,\n",
    " subsample=0.8,\n",
    " colsample_bytree=0.8,\n",
    " objective= 'binary:logistic',\n",
    " nthread=4,\n",
    " scale_pos_weight=1,\n",
    " seed=27)\n",
    "top10 = ['삼성전자', 'SK하이닉스', '셀트리온', 'POSCO', \n",
    "         '현대차', 'LG화학', 'NAVER', '삼성물산', \n",
    "         '현대모비스']\n",
    "predictors = [x for x in df.columns if x not in ['Y1','Y2','Y3']]\n",
    "target = \"Y2\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Report\n",
      "Accuracy : 0.9977\n",
      "AUC Score (Train): 0.999960\n",
      "elapse time : 7.26\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a129a97ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model, fear_importance = modelfit(alg, train[\"005930\"], predictors, target, cv_folds=5, early_stopping_rounds=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    }
   ],
   "source": [
    "pred = model.predict(test[\"005930\"][predictors])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8950617283950617"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics.accuracy_score(pred, test[\"005930\"][target])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictors.index(\"pct_change\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'pct_change'"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictors.pop(6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Report\n",
      "Accuracy : 0.8094\n",
      "AUC Score (Train): 0.891980\n",
      "elapse time : 1.74\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1297ff9b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model, fear_importance = modelfit(alg, train[\"005930\"], predictors, target, cv_folds=5, early_stopping_rounds=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.49382716049382713"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = model.predict(test[\"005930\"][predictors])\n",
    "metrics.accuracy_score(pred, test[\"005930\"][target])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = ['Open', 'his_vol_7', 'Volume', 'volume_ma_4']\n",
    "target = 'Y2'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Report\n",
      "Accuracy : 0.6644\n",
      "AUC Score (Train): 0.741534\n",
      "elapse time : 0.28\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a129a3ba58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model, fear_importance = modelfit(alg, train[\"005930\"], predictors, target, cv_folds=5, early_stopping_rounds=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.49691358024691357"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = model.predict(test[\"005930\"][predictors])\n",
    "metrics.accuracy_score(pred, test[\"005930\"][target])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = ['Open', 'close_ma_3', 'close_ma_5']\n",
    "target = 'Y3'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Report\n",
      "Accuracy : 0.6829\n",
      "AUC Score (Train): 0.772316\n",
      "elapse time : 0.49\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a10645de48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model, fear_importance = modelfit(alg, train[\"005930\"], predictors, target, cv_folds=5, early_stopping_rounds=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\JinMyungHun\\Anaconda3\\envs\\ds\\lib\\site-packages\\sklearn\\preprocessing\\label.py:151: DeprecationWarning: The truth value of an empty array is ambiguous. Returning False, but in future this will result in an error. Use `array.size > 0` to check that an array is not empty.\n",
      "  if diff:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.5061728395061729"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = model.predict(test[\"005930\"][predictors])\n",
    "metrics.accuracy_score(pred, test[\"005930\"][target])"
   ]
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}