final_submission.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c64878a6",
   "metadata": {},
   "source": [
    "## 실행 환경\n",
    "category-encoders==2.2.2  \n",
    "fire==0.4.0  \n",
    "joblib==1.0.1  \n",
    "keras==2.6.0  \n",
    "numpy==1.19.5  \n",
    "pandas==1.3.3  \n",
    "scikit-learn==0.24.2  \n",
    "scipy==1.7.1  \n",
    "sklearn==0.0  \n",
    "swifter==1.0.9  \n",
    "tqdm==4.62.3  \n",
    "xgboost==1.4.2  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "da063771",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "sys.path.append('..')\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from tqdm import tqdm\n",
    "import xgboost as xgb\n",
    "import pickle\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.model_selection import train_test_split\n",
    "from datetime import datetime, timedelta\n",
    "from category_encoders import TargetEncoder\n",
    "import swifter"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb2a807b",
   "metadata": {},
   "source": [
    "## 데이터 로드\n",
    "외부 데이터 및 시간이 오래 걸리는 코드 전처리의 경우 메일로 첨부하였습니다.  \n",
    "해당 실행 파일의 위치에 data 폴더와 external 폴더를 함께 위치 시켜주어야 실행 가능합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "be829237",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_path = './data/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "fc478909",
   "metadata": {},
   "outputs": [],
   "source": [
    "external_path = './external/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "97544a25",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data_path = data_path + 'train_data(final).csv'\n",
    "test_data_path =  data_path + 'test_data(final).csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ace14ef0",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df = pd.read_csv(train_data_path)\n",
    "test_df = pd.read_csv(test_data_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e4f4232",
   "metadata": {},
   "source": [
    "## 외부 데이터 병합\n",
    "외부 데이터 병합의 경우 실행 시간 문제로 인하여 미리 전처리 하여 메일로 첨부하였습니다.  \n",
    "새로운 데이터를 돌릴 경우에만 아래 코드를 실행하면 됩니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "1e4e3e78",
   "metadata": {},
   "outputs": [],
   "source": [
    "def mergeAll(df) :\n",
    "    marcap_df = pd.read_csv(external_path + 'result/mean_marcapdata.csv')\n",
    "    marcap_df['code'] = marcap_df.apply(lambda x: 'A' + x['code'], axis=1)\n",
    "    exchange_df = pd.read_csv(external_path + 'result/exchange_rate.csv')\n",
    "    nsi_df = pd.read_csv(external_path + 'result/nsi_data.csv')\n",
    "    stk_mean_df = pd.read_csv(external_path + 'result/abs_mean_stockdata.csv')\n",
    "    \n",
    "    dtypes={'code':np.str, 'name':np.str,'open':np.int64, 'high':np.int64, 'low':np.int64, 'close':np.int64, 'volume':np.int64, 'amount':np.int64, 'changes':np.int64, 'changecode':np.str, 'changesratio':np.float64, 'marcap':np.int64, 'stocks':np.int64,'marketid':np.str, 'market':np.str, 'dept':np.str, 'rank':np.int64, 'date':np.int64}\n",
    "\n",
    "    allmarcap_df = pd.read_csv(external_path + 'result/all_marcap_data2016_2020.csv',dtype=dtypes)\n",
    "    allmarcap_df = allmarcap_df.drop(['name','market','dept','marketid'],axis=1) # 1차 드롭\n",
    "    allmarcap_df = allmarcap_df.drop(['close','changecode','changes','open','high','low','volume'],axis=1) # finance에서 들어간 것 drop\n",
    "\n",
    "    allmarcap_df['code'] = allmarcap_df.apply(lambda x : 'A' + x['code'], axis = 1)\n",
    "    allmarcap_df.columns = ['code','changesratio_d', 'amount_d', 'marcap_d', 'stocks_d', 'rank_d', 'date'] # 컬럼명 변경\n",
    "    \n",
    "\n",
    "    stk_mean_df['code'] = stk_mean_df.apply(lambda x : 'A' + x['code'], axis = 1)\n",
    "    stk_mean_df = stk_mean_df.drop(['sq_abs_change','sq_abs_change100000'], axis=1)\n",
    "    stk_mean_df.columns = ['code','open_mean', 'high_mean', 'low_mean', 'close_mean', 'volume_mean','abs_change_mean']\n",
    "\n",
    "    df = pd.merge(left = df, right = stk_mean_df, how='left', left_on='iem_cd', right_on = 'code')\n",
    "    df = df.drop('code', axis=1)\n",
    "    \n",
    "    df = MergeFinanceData(df)\n",
    "    df = MergeMarcap(df, allmarcap_df)\n",
    "    df = MergeMeanMarcap(df, marcap_df)\n",
    "    df = MergeDollar(df, exchange_df)\n",
    "    df = MergeGDP(df)\n",
    "    df = MergeNSI(df, nsi_df)\n",
    "    \n",
    "    df = df.drop_duplicates()\n",
    "    df = df.drop(['close_d', 'date', 'high_d', 'low_d'], axis = 1)\n",
    "\n",
    "    return df\n",
    "\n",
    "# ==============================================================================================================\n",
    "# 1. finance stock data - byn_dt 기준으로 open high low close volume change 추가\n",
    "def MergeFinanceData(process_df):  #\n",
    "    cd = process_df['iem_cd']\n",
    "    date = process_df['byn_dt']\n",
    "\n",
    "    di = external_path + 'crawling/stockdata'\n",
    "    file_list = os.listdir(di)\n",
    "\n",
    "    f_list = []\n",
    "    for i in cd:\n",
    "        i = i[1:]\n",
    "        f_name = list(filter(lambda x: i in x, file_list))\n",
    "        if len(f_name) != 0:\n",
    "            f_list.append(f_name)\n",
    "        else:\n",
    "            f_list.append(None)\n",
    "\n",
    "    result = pd.DataFrame(columns=['Open', 'High', 'Low', 'Close', 'Volume', 'Change'])\n",
    "    for i in range(len(f_list)):\n",
    "        if f_list[i] is not None:\n",
    "            f = f_list[i][0]\n",
    "            print(i, f)\n",
    "            filepath = '{}/{}'.format(di, f)\n",
    "            df = pd.read_csv(filepath)\n",
    "            d = str(date[i])\n",
    "            d = d[:4] + '-' + d[4:6] + '-' + d[6:]\n",
    "            temp = df[df['Date'] == d]\n",
    "            temp = temp.loc[:, ['Open', 'High', 'Low', 'Close', 'Volume', 'Change']]\n",
    "            result = result.append(temp, ignore_index=True)\n",
    "        else:\n",
    "            result = result.append({'Open': 0, 'High': 0, 'Low': 0, 'Close': 0, 'Volume': 0, 'Change': 0},\n",
    "                                   ignore_index=True)\n",
    "\n",
    "    result.columns = ['open_d', 'high_d', 'low_d', 'close_d', 'volume_d', 'change_d']  # 열이름 변경\n",
    "    process_df = pd.concat([process_df, result], axis=1)\n",
    "\n",
    "    return process_df\n",
    "\n",
    "\n",
    "# ==============================================================================================================\n",
    "# 2. 그외 marcap mean data 추가 (mean_changesratio,  mean_amount, mean_marcap, mean_stocks, mean_rank)\n",
    "def MergeMeanMarcap(process_df, marcap_df):\n",
    "    process_df = pd.merge(left=process_df, right=marcap_df[\n",
    "        {'code', 'mean_changesratio', 'mean_amount', 'mean_marcap', 'mean_stocks', 'mean_rank'}], how='left',\n",
    "                          left_on='iem_cd', right_on='code')\n",
    "    process_df = process_df.drop(['code'], axis=1)  # 중복열 제거\n",
    "\n",
    "    return process_df\n",
    "\n",
    "\n",
    "# ==============================================================================================================\n",
    "# 3. dollar 추가 - byn_dt 기준으로 환율 추가\n",
    "def MergeDollar(process_df, exchange_df):\n",
    "    process_df = pd.merge(process_df, exchange_df, how='left', left_on='byn_dt', right_on='date')\n",
    "    process_df = process_df.drop('date', axis=1)  # 중복열 제거\n",
    "\n",
    "    return process_df\n",
    "\n",
    "\n",
    "# ==============================================================================================================\n",
    "# 4. gdp 추가 - 분기별 데이터 활용, 데이터를 직업 사용\n",
    "def MergeGDP(process_df):\n",
    "    conditions = [(process_df['byn_dt'] >= 20201001), (process_df['byn_dt'] >= 20200701),\n",
    "                  (process_df['byn_dt'] >= 20200401), (process_df['byn_dt'] >= 20200101),\n",
    "                  (process_df['byn_dt'] >= 20191001), (process_df['byn_dt'] >= 20190701),\n",
    "                  (process_df['byn_dt'] >= 20190401), (process_df['byn_dt'] >= 20190101),\n",
    "                  (process_df['byn_dt'] >= 20181001), (process_df['byn_dt'] >= 20180701),\n",
    "                  (process_df['byn_dt'] >= 20180401), (process_df['byn_dt'] >= 20180101),\n",
    "                  (process_df['byn_dt'] >= 20171001), (process_df['byn_dt'] >= 20170701),\n",
    "                  (process_df['byn_dt'] >= 20170401), (process_df['byn_dt'] >= 20170101),\n",
    "                  (process_df['byn_dt'] >= 20161001), (process_df['byn_dt'] >= 20160701),\n",
    "                  (process_df['byn_dt'] >= 20160401), (process_df['byn_dt'] >= 20160101), ]\n",
    "    choices1 = [1.1, 2.2, -3.2, -1.3, 2.3, 2.0, 2.1, 1.8, 3.1, 2.4, 3.1, 3.0,\n",
    "                2.9, 3.9, 2.7, 3.1, 2.6, 2.8, 3.6, 2.8]\n",
    "    process_df['gdp_rate'] = np.select(conditions, choices1, default=0)\n",
    "\n",
    "    choices2 = [492100.10, 491181.50, 472328.10, 458202.40, 497064.80, 487177.00, 479907.00,\n",
    "                454891.00, 489800.50, 485535.70, 473018.30, 449838.10, 475092.90, 473436.20,\n",
    "                454141.70, 433027.30, 453412.40, 439421.60, 434462.90, 413482.60]\n",
    "    process_df['gdp'] = np.select(conditions, choices2, default=0)\n",
    "\n",
    "    return process_df\n",
    "\n",
    "\n",
    "# ==============================================================================================================\n",
    "# 5. nsi 추가\n",
    "def MergeNSI(process_df, nsi_df):\n",
    "    process_df = pd.merge(left=process_df, right=nsi_df, how='left', left_on='byn_dt', right_on='date')\n",
    "    process_df.drop(['date'], axis=1)\n",
    "\n",
    "    return process_df\n",
    "\n",
    "\n",
    "# ==============================================================================================================\n",
    "# 6. marcap_d추가\n",
    "def MergeMarcap(process_df, allmarcap_df):\n",
    "    process_df = pd.merge(left=process_df, right=allmarcap_df, how='left', left_on=['iem_cd', 'byn_dt'],\n",
    "                          right_on=['code', 'date'])\n",
    "    process_df = process_df.fillna(0)\n",
    "    process_df = process_df.drop(['code', 'date'], axis=1)\n",
    "    process_df = process_df.drop_duplicates()\n",
    "\n",
    "    return process_df\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b65f6c99",
   "metadata": {},
   "outputs": [],
   "source": [
    "#train_df = mergeAll(train_df)\n",
    "#test_df = mergeAll(test_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "12609de4",
   "metadata": {},
   "source": [
    "## 파일 전처리 코드"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8ae4f0c4",
   "metadata": {},
   "source": [
    "정확한 날짜 계산을 위하여 공휴일 정보 불러오기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "86261bfb",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_holiday():\n",
    "    holiday = pd.read_csv(external_path +'result/holiday_date.csv')\n",
    "\n",
    "    holidays = pd.to_datetime(holiday['date'], infer_datetime_format=True).dt.strftime(\"%Y-%m-%d\")\n",
    "    holidays = list(holidays)\n",
    "    for i in range(len(holidays)):\n",
    "        tmp = holidays[i].split('-')\n",
    "        holidays[i] = ''.join(tmp)\n",
    "    holidays = list(map(int, holidays))\n",
    "    return holidays"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "467a8b51",
   "metadata": {},
   "outputs": [],
   "source": [
    "holidays = get_holiday()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "89c56af0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def working_day(s_dt, e_dt) : # 공휴일 제외 # 공휴일이 주말인 경우 고려 \n",
    "    count = 0\n",
    "    for day in holidays:\n",
    "        if day >= s_dt and day <= e_dt:\n",
    "                count += 1\n",
    "        elif day > e_dt:\n",
    "                break\n",
    "        else:\n",
    "                continue\n",
    "    s_dt = str(s_dt).split('.')[0]\n",
    "    e_dt = str(e_dt).split('.')[0]\n",
    "    s_dt = s_dt[:4] + '-' + s_dt[4:6] + '-' + s_dt[6:]\n",
    "    e_dt = e_dt[:4] + '-' + e_dt[4:6] + '-' + e_dt[6:]\n",
    "    w_day = np.busday_count(s_dt, e_dt, weekmask='1111100') # 주말제외\n",
    "\n",
    "    result = w_day - count\n",
    "    return result"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6576ac92",
   "metadata": {},
   "source": [
    "학습 데이터에서 hist_d를 계산하기 위해서 6개월 단위로 계산"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e2526f7a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def to_integer(dt_time):\n",
    "    return 10000*dt_time.year + 100*dt_time.month + dt_time.day"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4ba2b43e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def hist_d_half(buy, hold_d) :\n",
    "    buy_d = str(buy)\n",
    "    buy_d = datetime(int(buy_d[:4]), int(buy_d[4:6]), int(buy_d[6:]))\n",
    "    sell = to_integer(buy_d+timedelta(days=hold_d))\n",
    "    base = [20160101, 20160630, 20170102, 20170630, 20180102, 20180630, \n",
    "            20190102, 20190630, 20200102, 20200630, 20210101]\n",
    "\n",
    "    for i in range(len(base)) :\n",
    "        if sell < base[i] :\n",
    "            if buy > base[i-1] :\n",
    "                return hold_d*0.6\n",
    "                #return working_day(buy, base[i])\n",
    "            else :\n",
    "                return working_day(buy, base[i-1])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58707ec9",
   "metadata": {},
   "source": [
    "종목에 따른 타겟 인코딩"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "54a74d9d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def target_encoding(df, train) :\n",
    "    if train : \n",
    "        target_encoder = TargetEncoder()\n",
    "        target_encoder.fit(df['iem_cd'], df['hold_d'])\n",
    "        df['iem_cd_te'] = target_encoder.transform(df['iem_cd'])\n",
    "        with open('./data/preprocessing/' + 'iem_te.pkl', 'wb') as f:\n",
    "            pickle.dump(target_encoder, f)\n",
    "    else :\n",
    "        with open('./data/preprocessing/' + 'iem_te.pkl', 'rb') as f:\n",
    "            target_encoder = pickle.load(f)\n",
    "        df['iem_cd_te'] = target_encoder.transform(df['iem_cd'])\n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "77be53c4",
   "metadata": {},
   "source": [
    "학습 데이터의 주식 매매 날짜 기준 past_d 계산"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "bedf17bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cal_past_d(df) : # train data\n",
    "    df['past_d'] = 0\n",
    "    df = df.sort_values(by = ['act_id', 'iem_cd', 'byn_dt'], axis = 0).reset_index(drop=True)\n",
    "    iem = ''\n",
    "    cnt = 1\n",
    "    print('past_d calculating...')\n",
    "    for idx, row in df.iterrows() :\n",
    "        if row['iem_cd'] != iem :\n",
    "            iem = row['iem_cd']\n",
    "            cnt = 1\n",
    "        else :\n",
    "            df.at[idx, 'past_d'] = (df.iloc[idx-1]['past_d']*(cnt-1) + df.iloc[idx-1]['hold_d']) / cnt\n",
    "            cnt += 1\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "334c8ea5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_test_past_d(hist, cus, test_df):\n",
    "    hist = hist\n",
    "    cus = cus\n",
    "    test_df = test_df[['act_id', 'iem_cd']]\n",
    "    \n",
    "    hist = pd.merge(hist, cus[['act_id', 'ivs_icn_cd']], how='left', on = 'act_id')\n",
    "    \n",
    "    df = hist[['act_id', 'iem_cd']]\n",
    "    id_cd_count = df.groupby('act_id')['iem_cd'].count().reset_index(name='count')\n",
    "    id_cd_list = df.groupby('act_id')['iem_cd'].apply(list).reset_index(name='iem_cd_list')\n",
    "\n",
    "    for i in range(9992):\n",
    "        cd_list = id_cd_list['iem_cd_list'][i]\n",
    "        cd_set = set(cd_list)\n",
    "        count = len(cd_set)\n",
    "        id_cd_list['iem_cd_list'][i] = cd_set\n",
    "\n",
    "    id_cd_list\n",
    "    \n",
    "    # 계좌번호 / 종목 / 날짜 별 보유기간 추가\n",
    "    for i in range(len(id_cd_list['act_id'])):\n",
    "        if i == 0:\n",
    "            act_id = id_cd_list['act_id'][i]\n",
    "            df1 = cal_duration(act_id, hist)\n",
    "        else:\n",
    "            act_id = id_cd_list['act_id'][i]\n",
    "            df2 = cal_duration(act_id, hist)\n",
    "            df1 = pd.concat([df1, df2], axis = 0)\n",
    "    df2 = df1.reset_index(drop = True)\n",
    "    df3 = df2[(df2['past_d'] != 0)]\n",
    "    df4 = df3.groupby(['act_id', 'iem_cd'])['past_d'].apply(list).reset_index(name='past_d_list')\n",
    "    df5 = df3.groupby(['act_id', 'iem_cd'])['past_d'].mean().reset_index(name='act_past_d_mean')\n",
    "    past_d_mean = df5\n",
    "    past_d_mean = past_d_mean.rename(columns = {'act_past_d_mean':'past_d'})\n",
    "    test_df3 = pd.merge(test_df, past_d_mean, how = 'left', on = ['act_id', 'iem_cd'])\n",
    "    \n",
    "    test_df3.to_csv(data_path + 'preprocessing/test_past_d_mean.csv')\n",
    "    return test_df3\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c7287076",
   "metadata": {},
   "source": [
    "테스트 데이터에서 past_d 구하기 (stk_hist 기준 모든 과거 데이터)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "ba689ba0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def stk_past_d(df) : # test data\n",
    "    if os.path.exists(data_path + 'preprocessing/test_past_d_mean.csv') :\n",
    "        print(\"test_past_d_mean file is already.\")\n",
    "        test_past_df = pd.read_csv(data_path + 'preprocessing/test_past_d_mean.csv')\n",
    "    else :\n",
    "        stk_df = pd.read_csv(data_path +'stk_bnc_hist.csv')\n",
    "        cus_df = pd.read_csv(data_path + 'cus_info.csv')\n",
    "        stk_df = stk_df.sort_values(by=['bse_dt'])\n",
    "        test_past_df = get_test_past_d(stk_df, cus_df, df)\n",
    "    df = pd.merge(df, test_past_df, how='left', on = ['act_id', 'iem_cd'])\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "8531be5d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cal_duration(act_id, hist):\n",
    "    df = hist[(hist['act_id'] == act_id)][['act_id', 'iem_cd', 'bse_dt', 'bnc_qty', 'tot_aet_amt', 'ivs_icn_cd']]\n",
    "    sorted_df = df.sort_values(by=['iem_cd', 'bse_dt'], ascending = [True, True])\n",
    "    sorted_df = sorted_df.reset_index(drop = True)\n",
    "    sorted_df['past_d'] = 0\n",
    "    cur_dur = {}\n",
    "    #     dur_list = []\n",
    "    for i in range(len(sorted_df)):    \n",
    "        # 위에서부터 하나씩 읽어가는데 딕셔너리에 종목코드가 존재하지 않는다면 해당 날짜 추가\n",
    "        if sorted_df['iem_cd'][i] not in cur_dur:\n",
    "            if sorted_df['bnc_qty'][i] == 0 :\n",
    "                sell = sorted_df['bse_dt'][i]\n",
    "                buy = 20160102\n",
    "                sorted_df['past_d'][i] = working_day(buy, sell)\n",
    "            cur_dur[sorted_df['iem_cd'][i]] = sorted_df['bse_dt'][i]\n",
    "            \n",
    "        # 이미 존재하면 해당 잔고가 0인지 확인\n",
    "        else :\n",
    "            # 일단 보유기간은 사고/팔고 개념이 있으니깐 팔지 않은(0이 아닌) 주식은 계산 X\n",
    "            # 해당 잔고가 0이면 그 날짜와 딕셔너리에 존재하는 날짜를 사용해 보유 기간 계산\n",
    "            # 어떤 종목을 다 팔고 다시 사고/팔았을 경우도 생각\n",
    "            if sorted_df['bnc_qty'][i] == 0:\n",
    "                sell = sorted_df['bse_dt'][i]\n",
    "                buy = cur_dur[sorted_df['iem_cd'][i]]\n",
    "                dur = working_day(buy, sell)\n",
    "                sorted_df['past_d'][i] = dur \n",
    "                # 다 팔았으니깐 딕셔너리에서 해당 종목코드 삭제(나중에 다시 살 경우 대비)\n",
    "                del cur_dur[sorted_df['iem_cd'][i]]\n",
    "            # 0이 없다: 다 팔지 않음 ==> 일단 보유기간 측정 안하는 걸로.\n",
    "            else:\n",
    "                continue\n",
    "    last_day = 20201231\n",
    "    for i in cur_dur.items() :\n",
    "        past_d = working_day(i[1], last_day)\n",
    "        sorted_df = sorted_df.append({'act_id' : act_id , 'iem_cd' : i[0], 'bse_dt' : last_day, 'bnc_qty' : 0, 'tot_aet_amt' : 0, 'past_d' : past_d} , ignore_index=True)\n",
    "        \n",
    "    return sorted_df\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08d28eed",
   "metadata": {},
   "source": [
    "ivs 코드에 따른 종목 past_d 평균 데이터 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "88195c26",
   "metadata": {},
   "outputs": [],
   "source": [
    "def cal_hist() : # for ivs_past_d\n",
    "    hist = pd.read_csv(data_path + 'stk_bnc_hist.csv')\n",
    "    cus = pd.read_csv(data_path + 'cus_info.csv')\n",
    "    hist = pd.merge(hist, cus[['act_id', 'ivs_icn_cd']], how='left', on = 'act_id')\n",
    "    ivs_lst = sorted(hist['ivs_icn_cd'].unique())\n",
    "    df_lst = [0] * len(ivs_lst)\n",
    "    for i in range(len(ivs_lst)) :\n",
    "        ivs_code = ivs_lst[i]\n",
    "        df = hist[hist['ivs_icn_cd'] == ivs_code]\n",
    "        id_list = df['act_id'].unique()\n",
    "        for j in range(len(id_list)):\n",
    "            if j == 0:\n",
    "                act_id = id_list[j]\n",
    "                tmp_df1 = cal_duration(act_id, hist)\n",
    "            else:\n",
    "                act_id = id_list[j]\n",
    "                tmp_df2 = cal_duration(act_id, hist)\n",
    "                tmp_df1 = pd.concat([tmp_df1, tmp_df2], axis = 0)\n",
    "        tmp_df1.reset_index(drop = True)\n",
    "        df_lst[i] = tmp_df1\n",
    "    return df_lst"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "5382f066",
   "metadata": {},
   "outputs": [],
   "source": [
    "def chk_ivs_past_d() :\n",
    "    if os.path.exists('./data/preprocessing/ivs_past_d_mean.csv') :\n",
    "        print(\"ivs_past_d_mean file is already.\")\n",
    "        return\n",
    "    print(\"make ivs_past_d mean file\")\n",
    "    ivs_lst, df_lst = cal_hist()\n",
    "    for i in range(len(ivs_lst)) :\n",
    "        if i == 0 :\n",
    "            df1 = df_lst[i].reset_index(drop=True)\n",
    "            df1 = df1[(df1['past_d']) != 0]\n",
    "            df1 = df1.groupby(['iem_cd'])['past_d'].mean().reset_index(name='ivs_past_d_mean')\n",
    "            df1['ivs_icn_cd'] = ivs_lst[i]\n",
    "        else :\n",
    "            df2 = df_lst[i].reset_index(drop=True)\n",
    "            df2 = df2[(df2['past_d']) != 0]\n",
    "            df2 = df2.groupby(['iem_cd'])['past_d'].mean().reset_index(name='ivs_past_d_mean')\n",
    "            df2['ivs_icn_cd'] = ivs_lst[i]\n",
    "            df1 = pd.concat([df1, df2], axis = 0)\n",
    "    df1.to_csv(data_path + 'preprocessing/' + 'ivs_past_d_mean.csv', index= False)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8b78264f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def apply_ivs_past_d(df) :\n",
    "    chk_ivs_past_d()\n",
    "    ivs_past_df = pd.read_csv(data_path + 'preprocessing/' + 'ivs_past_d_mean.csv')\n",
    "    df = pd.merge(df, ivs_past_df, how = 'left', on = ['iem_cd', 'ivs_icn_cd'])\n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d1841a7",
   "metadata": {},
   "source": [
    "stk hist로부터 주식 관련 정보 로드"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "685d47a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def bnc_count(hist):\n",
    "    hist = hist\n",
    "    tmp = hist.groupby(['act_id', 'iem_cd'])['bse_dt'].count().reset_index(name='bnc_count')\n",
    "    tmp = pd.merge(hist, tmp, how = 'left', on = ['act_id', 'iem_cd'])\n",
    "    return tmp\n",
    "\n",
    "def qty_change(hist):\n",
    "    hist = hist\n",
    "    sorted_hist = hist.sort_values(by=['iem_cd', 'bse_dt'], ascending = [True, True])\n",
    "    sorted_hist = sorted_hist.reset_index(drop=True)\n",
    "    sorted_hist = sorted_hist.groupby(['act_id', 'iem_cd'])['bnc_qty'].apply(list).reset_index(name = 'qty_list')\n",
    "    sorted_hist['qty_change'] = 0\n",
    "    \n",
    "    for i in range(len(sorted_hist)):\n",
    "        qty_change = []\n",
    "        tmp = sorted_hist['qty_list'][i]\n",
    "\n",
    "        for j in range(len(tmp)-1):\n",
    "            change = abs(tmp[j+1] - tmp[j])\n",
    "            qty_change.append(change)\n",
    "\n",
    "        qty_change = np.array(qty_change)\n",
    "        qty_mean = np.mean(qty_change)\n",
    "        sorted_hist['qty_change'][i] = qty_mean\n",
    "        \n",
    "    sorted_hist = sorted_hist.drop(['qty_list'], axis = 1)\n",
    "    \n",
    "    # stk_bnc_hist 데이터와 merge\n",
    "    new_hist = pd.merge(hist, sorted_hist, how = 'left', on = ['act_id', 'iem_cd'])\n",
    "    \n",
    "    \n",
    "    return new_hist\n",
    "\n",
    "def aet_amt_change(hist):\n",
    "    hist = hist\n",
    "    sorted_hist = hist.sort_values(by=['iem_cd', 'bse_dt'], ascending = [True, True])\n",
    "    sorted_hist = sorted_hist.reset_index(drop=True)\n",
    "    sorted_hist = sorted_hist.groupby(['act_id', 'iem_cd'])['tot_aet_amt'].apply(list).reset_index(name = 'aet_amt_list')\n",
    "    sorted_hist['aet_amt_change'] = 0\n",
    "    \n",
    "    for i in range(len(sorted_hist)):\n",
    "        aet_change = []\n",
    "        tmp = sorted_hist['aet_amt_list'][i]\n",
    "        for j in range(len(tmp)-1):\n",
    "            change = abs(tmp[j+1] - tmp[j])\n",
    "            aet_change.append(change)\n",
    "        aet_change = np.array(aet_change)\n",
    "        aet_mean = np.mean(aet_change)\n",
    "\n",
    "        sorted_hist['aet_amt_change'][i] = aet_mean\n",
    "        \n",
    "    sorted_hist = sorted_hist.drop(['aet_amt_list'], axis = 1)\n",
    "    \n",
    "    # stk_bnc_hist 데이터와 merge\n",
    "    new_hist = pd.merge(hist, sorted_hist, how = 'left', on = ['act_id', 'iem_cd'])\n",
    "    \n",
    "    return new_hist\n",
    "\n",
    "def end_price(hist):\n",
    "    new_hist = hist\n",
    "    new_hist['end_price'] = 0\n",
    "    \n",
    "    for i in range(len(new_hist)):\n",
    "        tot = new_hist['tot_aet_amt'][i]\n",
    "        qty = new_hist['bnc_qty'][i]\n",
    "        end_price = tot // qty\n",
    "        new_hist['end_price'][i] = end_price\n",
    "    \n",
    "    return new_hist\n",
    "\n",
    "def get_from_stk(df) :\n",
    "    hist = pd.read_csv(data_path + 'stk_bnc_hist.csv')\n",
    "    new_hist = bnc_count(hist)\n",
    "    new_hist = qty_change(new_hist)\n",
    "    new_hist = aet_amt_change(new_hist)\n",
    "    new_hist = end_price(new_hist)\n",
    "    df = pd.merge(df, new_hist, how='left', left_on = ['act_id', 'iem_cd', 'byn_dt'], right_on = ['act_id', 'iem_cd', 'bse_dt'])\n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71c4deaf",
   "metadata": {},
   "source": [
    "## 전처리"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "068bcd55",
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocessing(df, train) :\n",
    "    print('*********preprocessing...')\n",
    "\n",
    "    #cus info merge\n",
    "    print(\"*********cus info merge..\")\n",
    "    cus_df = pd.read_csv(data_path +'cus_info.csv')\n",
    "    df = pd.merge(left = df, right = cus_df, how='left', on='act_id')\n",
    "    iem_df = pd.read_csv(data_path +'iem_info_20210902.csv')\n",
    "    df = pd.merge(df, iem_df, how='left', on='iem_cd')\n",
    "\n",
    "    #iem target encoding\n",
    "    print(\"*********iem target encoding..\")\n",
    "    df = target_encoding(df, train)\n",
    "\n",
    "    # from stk hist\n",
    "    print(\"*********from stk hist..(bnc_count, qty_change, aet_amt_change, end_price\")\n",
    "    df = get_from_stk(df)\n",
    "\n",
    "    #ivs_past_d 계산\n",
    "    print(\"*********calculating ivs past_d..\")\n",
    "    df = apply_ivs_past_d(df) \n",
    "\n",
    "    # 4. past_d 계산\n",
    "    print(\"*********calculating total past_d..\")\n",
    "    if train :\n",
    "        df = cal_past_d(df)\n",
    "        df['past_d'] = df.apply(lambda x : x['ivs_past_d_mean'] if x['past_d'] == 0 else x['past_d'], axis = 1)\n",
    "    else :\n",
    "        df = stk_past_d(df)\n",
    "\n",
    "    # 5. hist_d 계산\n",
    "    if train :\n",
    "        print(\"*********calculating train hist_d..\")\n",
    "        df['hist_d'] = df.apply(lambda x : hist_d_half(x['byn_dt'], x['hold_d']), axis = 1)\n",
    "\n",
    "    print(\"**preprocessing successfully\")\n",
    "\n",
    "    df['bnc_count'] = df['bnc_count'] + 1\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "d4a03419",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "*********preprocessing...\n",
      "*********cus info merge..\n",
      "*********iem target encoding..\n",
      "*********from stk hist..(bnc_count, qty_change, aet_amt_change, end_price\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:24: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/numpy/core/fromnumeric.py:3373: RuntimeWarning: Mean of empty slice.\n",
      "  out=out, **kwargs)\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/numpy/core/_methods.py:170: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  ret = ret.dtype.type(ret / rcount)\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/pandas/core/indexing.py:1732: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self._setitem_single_block(indexer, value, name)\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:50: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:67: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:66: RuntimeWarning: invalid value encountered in double_scalars\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "*********calculating ivs past_d..\n",
      "ivs_past_d_mean file is already.\n",
      "*********calculating total past_d..\n",
      "past_d calculating...\n",
      "*********calculating train hist_d..\n",
      "**preprocessing successfully\n"
     ]
    }
   ],
   "source": [
    "train_df = preprocessing(train_df, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "a39952dd",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "*********preprocessing...\n",
      "*********cus info merge..\n",
      "*********iem target encoding..\n",
      "*********from stk hist..(bnc_count, qty_change, aet_amt_change, end_price\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:24: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:50: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:67: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/home/nyongja/anaconda3/envs/nh2021/lib/python3.7/site-packages/ipykernel_launcher.py:66: RuntimeWarning: invalid value encountered in double_scalars\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "*********calculating ivs past_d..\n",
      "ivs_past_d_mean file is already.\n",
      "*********calculating total past_d..\n",
      "test_past_d_mean file is already.\n",
      "**preprocessing successfully\n"
     ]
    }
   ],
   "source": [
    "test_df = preprocessing(test_df, False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7861515f",
   "metadata": {},
   "source": [
    "## 학습하기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "6641cd5c",
   "metadata": {},
   "outputs": [],
   "source": [
    "drop_features = ['act_id', 'iem_cd', 'bse_dt', 'iem_krl_nm']\n",
    "used_features = ['byn_dt', 'high_mean', 'change_d', 'nsi', 'mean_amount',\n",
    "       'volume_mean', 'mean_rank', 'volume_d', 'abs_change_mean', 'close_mean',\n",
    "       'open_d', 'mean_stocks', 'mean_changesratio', 'open_mean', 'dollar',\n",
    "       'gdp_rate', 'mean_marcap', 'low_mean', 'changesratio_d', 'amount_d',\n",
    "       'marcap_d', 'stocks_d', 'rank_d', 'gdp', 'sex_dit_cd', 'cus_age_stn_cd',\n",
    "       'ivs_icn_cd', 'cus_aet_stn_cd', 'mrz_pdt_tp_sgm_cd', 'lsg_sgm_cd',\n",
    "       'tco_cus_grd_cd', 'tot_ivs_te_sgm_cd', 'mrz_btp_dit_cd', 'btp_cfc_cd',\n",
    "       'mkt_pr_tal_scl_tp_cd', 'stk_dit_cd', 'iem_cd_te', 'bnc_qty',\n",
    "       'tot_aet_amt', 'stk_par_pr', 'bnc_count', 'qty_change',\n",
    "       'aet_amt_change', 'end_price', 'ivs_past_d_mean', 'past_d', 'hist_d']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "08abf743",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['act_id', 'iem_cd', 'byn_dt', 'hold_d', 'high_mean', 'change_d', 'nsi',\n",
       "       'mean_amount', 'volume_mean', 'mean_rank', 'volume_d',\n",
       "       'abs_change_mean', 'close_mean', 'open_d', 'mean_stocks',\n",
       "       'mean_changesratio', 'open_mean', 'dollar', 'gdp_rate', 'mean_marcap',\n",
       "       'low_mean', 'changesratio_d', 'amount_d', 'marcap_d', 'stocks_d',\n",
       "       'rank_d', 'gdp', 'sex_dit_cd', 'cus_age_stn_cd', 'ivs_icn_cd',\n",
       "       'cus_aet_stn_cd', 'mrz_pdt_tp_sgm_cd', 'lsg_sgm_cd', 'tco_cus_grd_cd',\n",
       "       'tot_ivs_te_sgm_cd', 'mrz_btp_dit_cd', 'iem_krl_nm', 'btp_cfc_cd',\n",
       "       'mkt_pr_tal_scl_tp_cd', 'stk_dit_cd', 'iem_cd_te', 'bse_dt', 'bnc_qty',\n",
       "       'tot_aet_amt', 'stk_par_pr', 'bnc_count', 'qty_change',\n",
       "       'aet_amt_change', 'end_price', 'ivs_past_d_mean', 'past_d', 'hist_d'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "832b3c81",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "===Training....===\n",
      "===Success===\n",
      "===model saved..====\n"
     ]
    }
   ],
   "source": [
    "train_df = train_df.drop(drop_features, axis = 1)\n",
    "train_df = train_df.fillna(0)\n",
    "col = train_df.columns.drop('hold_d')\n",
    "\n",
    "X_train, X_valid, y_train, y_valid = train_test_split(train_df[col], train_df['hold_d'], test_size = 0.2)\n",
    "\n",
    "print(\"**************\")\n",
    "model = xgb.XGBRegressor(eta=0.01, gamma=0.5, max_depth=5, min_child_weight=1, n_estimators=1000, subsample=0.7)\n",
    "\n",
    "print(\"===Training....===\")\n",
    "model.fit(X_train, y_train, verbose=True)\n",
    "print(\"===Success===\")\n",
    "model.save_model('xgboost_final.json')\n",
    "print(\"===model saved..====\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "b8454a17",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "===Predict...===\n",
      "===Success===\n"
     ]
    }
   ],
   "source": [
    "test_df = test_df[used_features]\n",
    "test_df = test_df.fillna(0)\n",
    "\n",
    "print(\"===Predict...===\")\n",
    "y_pred = model.predict(test_df)\n",
    "print(\"===Success===\")\n",
    "test_df['y_pred'] = y_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef22d7cf",
   "metadata": {},
   "source": [
    "예측 값 범위 조정 (예측 값이 hist_d + 146을 초과하지 않고, 0보다 작은 값이 나오지 않도록)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "7a7ab3c4",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = test_df.apply(lambda x : x['y_pred'] if x['y_pred'] <= x['hist_d']+146 else x['hist_d']+146, axis = 1)\n",
    "result = []\n",
    "for i in y_pred:\n",
    "    if i < 0 :\n",
    "        i = 1\n",
    "    result.append(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "f19fc157",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[260.8539733886719,\n",
       " 481.0,\n",
       " 226.87863159179688,\n",
       " 353.1401062011719,\n",
       " 15.564187049865723,\n",
       " 163.47300720214844,\n",
       " 182.1115264892578,\n",
       " 866.0,\n",
       " 147.57232666015625,\n",
       " 656.0,\n",
       " 215.48802185058594,\n",
       " 1118.005615234375,\n",
       " 613.0,\n",
       " 10.682504653930664,\n",
       " 34.16969299316406,\n",
       " 8.857787132263184,\n",
       " 196.98858642578125,\n",
       " 120.27728271484375,\n",
       " 426.0,\n",
       " 23.530668258666992,\n",
       " 19.666099548339844,\n",
       " 1066.3857421875,\n",
       " 565.0,\n",
       " 38.149513244628906,\n",
       " 11.692965507507324,\n",
       " 115.89386749267578,\n",
       " 115.53958129882812,\n",
       " 351.9452209472656,\n",
       " 612.0,\n",
       " 426.0,\n",
       " 118.84575653076172,\n",
       " 118.13790893554688,\n",
       " 1002.0,\n",
       " 235.3580322265625,\n",
       " 25.224040985107422,\n",
       " 136.1689453125,\n",
       " 232.79779052734375,\n",
       " 1121.6318359375,\n",
       " 225.8840789794922,\n",
       " 1067.05126953125,\n",
       " 1092.4423828125,\n",
       " 581.0,\n",
       " 1095.6715087890625,\n",
       " 1112.2952880859375,\n",
       " 1052.7152099609375,\n",
       " 753.0,\n",
       " 1077.8341064453125,\n",
       " 1114.8197021484375,\n",
       " 1092.3258056640625,\n",
       " 1084.60009765625,\n",
       " 1107.4056396484375,\n",
       " 1086.147216796875,\n",
       " 235.78358459472656,\n",
       " 792.0,\n",
       " 815.0,\n",
       " 252.84942626953125,\n",
       " 232.87564086914062,\n",
       " 775.0,\n",
       " 776.0,\n",
       " 868.0,\n",
       " 135.31576538085938,\n",
       " 820.0,\n",
       " 804.0,\n",
       " 885.0,\n",
       " 718.0,\n",
       " 813.0,\n",
       " 808.0,\n",
       " 129.24163818359375,\n",
       " 212.60105895996094,\n",
       " 807.0,\n",
       " 864.0,\n",
       " 890.0,\n",
       " 815.0,\n",
       " 1122.6077880859375,\n",
       " 25.16036033630371,\n",
       " 245.61708068847656,\n",
       " 45.36445999145508,\n",
       " 40.56549072265625,\n",
       " 245.4070281982422,\n",
       " 34.521705627441406,\n",
       " 6.462615966796875,\n",
       " 1093.7708740234375,\n",
       " 1103.0213623046875,\n",
       " 1029.0,\n",
       " 24.6197452545166,\n",
       " 280.3973388671875,\n",
       " 229.80470275878906,\n",
       " 46.61434555053711,\n",
       " 24.866121292114258,\n",
       " 19.160350799560547,\n",
       " 6.447112083435059,\n",
       " 320.41937255859375,\n",
       " 40.91393280029297,\n",
       " 25.144702911376953,\n",
       " 6.494322299957275,\n",
       " 15.467669486999512,\n",
       " 23.343536376953125,\n",
       " 11.635981559753418,\n",
       " 42.84465789794922,\n",
       " 185.43040466308594,\n",
       " 18.891061782836914,\n",
       " 24.376422882080078,\n",
       " 71.77727508544922,\n",
       " 610.0,\n",
       " 37.14021682739258,\n",
       " 162.9407501220703,\n",
       " 6.477539539337158,\n",
       " 11.456269264221191,\n",
       " 169.38265991210938,\n",
       " 100.06185150146484,\n",
       " 433.0,\n",
       " 181.43504333496094,\n",
       " 124.7680435180664,\n",
       " 18.900358200073242,\n",
       " 355.15496826171875,\n",
       " 31.0065975189209,\n",
       " 25.16721534729004,\n",
       " 11.602886199951172,\n",
       " 17.73186683654785,\n",
       " 170.05819702148438,\n",
       " 6.4674153327941895,\n",
       " 125.32927703857422,\n",
       " 362.1114196777344,\n",
       " 92.9709243774414,\n",
       " 613.0,\n",
       " 858.0,\n",
       " 357.60992431640625,\n",
       " 433.0,\n",
       " 782.0,\n",
       " 6.440630912780762,\n",
       " 1090.8231201171875,\n",
       " 48.82324981689453,\n",
       " 268.26226806640625,\n",
       " 843.0,\n",
       " 166.7778778076172,\n",
       " 24.516433715820312,\n",
       " 188.8962860107422,\n",
       " 924.0,\n",
       " 843.0,\n",
       " 272.58050537109375,\n",
       " 190.9864959716797,\n",
       " 23.495683670043945,\n",
       " 171.6209716796875,\n",
       " 25.125995635986328,\n",
       " 235.5101776123047,\n",
       " 44.3088493347168,\n",
       " 151.08587646484375,\n",
       " 70.28773498535156,\n",
       " 145.5068359375,\n",
       " 69.10235595703125,\n",
       " 23.434255599975586,\n",
       " 1112.9365234375,\n",
       " 1129.1915283203125,\n",
       " 1129.219482421875,\n",
       " 1120.1929931640625,\n",
       " 1113.2332763671875,\n",
       " 536.0,\n",
       " 1070.7147216796875,\n",
       " 161.45545959472656,\n",
       " 18.71437644958496,\n",
       " 62.64592361450195,\n",
       " 238.59561157226562,\n",
       " 6.496973991394043,\n",
       " 5.657010078430176,\n",
       " 942.0,\n",
       " 23.790082931518555,\n",
       " 37.9984245300293,\n",
       " 25.021846771240234,\n",
       " 5.74342155456543,\n",
       " 6.462174892425537,\n",
       " 17.448444366455078,\n",
       " 17.46187973022461,\n",
       " 17.430648803710938,\n",
       " 314.4739074707031,\n",
       " 1074.441162109375,\n",
       " 813.0,\n",
       " 11.630317687988281,\n",
       " 23.72538948059082,\n",
       " 50.74533462524414,\n",
       " 10.636312484741211,\n",
       " 10.643190383911133,\n",
       " 47.82038879394531,\n",
       " 10.553218841552734,\n",
       " 212.14341735839844,\n",
       " 1054.2208251953125,\n",
       " 569.0,\n",
       " 24.925189971923828,\n",
       " 24.872560501098633,\n",
       " 36.54374694824219,\n",
       " 41.9063606262207,\n",
       " 24.61383056640625,\n",
       " 15.527057647705078,\n",
       " 45.79690933227539,\n",
       " 30.7490234375,\n",
       " 56.646297454833984,\n",
       " 38.46485900878906,\n",
       " 103.32437133789062,\n",
       " 319.5819091796875,\n",
       " 1070.570068359375,\n",
       " 102.09153747558594,\n",
       " 136.4080810546875,\n",
       " 23.61756134033203,\n",
       " 25.365375518798828,\n",
       " 63.27880859375,\n",
       " 149.3498077392578,\n",
       " 177.0437469482422,\n",
       " 150.96334838867188,\n",
       " 1060.1446533203125,\n",
       " 101.58949279785156,\n",
       " 86.38533020019531,\n",
       " 151.1681365966797,\n",
       " 202.69677734375,\n",
       " 55.644630432128906,\n",
       " 71.9332275390625,\n",
       " 136.4049072265625,\n",
       " 130.8614501953125,\n",
       " 94.71308898925781,\n",
       " 150.2061004638672,\n",
       " 101.29119110107422,\n",
       " 18.77028465270996,\n",
       " 743.0,\n",
       " 100.0745620727539,\n",
       " 6.410455703735352,\n",
       " 36.797489166259766,\n",
       " 62.08417892456055,\n",
       " 5.422576904296875,\n",
       " 5.434065341949463,\n",
       " 759.0,\n",
       " 190.92929077148438,\n",
       " 367.8428649902344,\n",
       " 50.56853103637695,\n",
       " 803.0,\n",
       " 279.3826904296875,\n",
       " 99.30815124511719,\n",
       " 147.6590576171875,\n",
       " 568.0,\n",
       " 35.06974411010742,\n",
       " 157.0606689453125,\n",
       " 146.1087646484375,\n",
       " 132.8255157470703,\n",
       " 575.0,\n",
       " 602.0,\n",
       " 246.2715301513672,\n",
       " 100.87663269042969,\n",
       " 218.8142852783203,\n",
       " 106.5986099243164,\n",
       " 185.6434783935547,\n",
       " 101.20088958740234,\n",
       " 81.24111938476562,\n",
       " 63.149993896484375,\n",
       " 207.74664306640625,\n",
       " 62.22045135498047,\n",
       " 36.936065673828125,\n",
       " 25.45474624633789,\n",
       " 5.481751918792725,\n",
       " 161.9059600830078,\n",
       " 25.011816024780273,\n",
       " 15.681647300720215,\n",
       " 828.0,\n",
       " 537.0,\n",
       " 250.0430908203125,\n",
       " 10.507383346557617,\n",
       " 1109.103271484375,\n",
       " 208.6670379638672,\n",
       " 23.87820053100586,\n",
       " 1095.6475830078125,\n",
       " 94.51921081542969,\n",
       " 24.995309829711914,\n",
       " 118.31527709960938,\n",
       " 135.6680450439453,\n",
       " 807.0,\n",
       " 1105.8564453125,\n",
       " 1069.9488525390625,\n",
       " 1039.0,\n",
       " 806.0,\n",
       " 1036.0,\n",
       " 1088.9342041015625,\n",
       " 699.0,\n",
       " 807.0,\n",
       " 1091.7154541015625,\n",
       " 1074.671630859375,\n",
       " 1040.8756103515625,\n",
       " 24.108951568603516,\n",
       " 37.87483215332031,\n",
       " 11.589960098266602,\n",
       " 15.578505516052246,\n",
       " 30.994029998779297,\n",
       " 349.4470520019531,\n",
       " 10.57186508178711,\n",
       " 55.4785270690918,\n",
       " 24.52274513244629,\n",
       " 40.15702438354492,\n",
       " 36.601219177246094,\n",
       " 6.490869522094727,\n",
       " 33.635414123535156,\n",
       " 36.2144889831543,\n",
       " 30.80402183532715,\n",
       " 17.42304801940918,\n",
       " 716.0,\n",
       " 62.880767822265625,\n",
       " 17.600080490112305,\n",
       " 18.726913452148438,\n",
       " 6.490715026855469,\n",
       " 70.61907958984375,\n",
       " 42.82090377807617,\n",
       " 117.09239959716797,\n",
       " 56.188533782958984,\n",
       " 44.525325775146484,\n",
       " 70.3385238647461,\n",
       " 42.794437408447266,\n",
       " 6.470282077789307,\n",
       " 134.50857543945312,\n",
       " 63.356201171875,\n",
       " 239.8986053466797,\n",
       " 44.67942428588867,\n",
       " 423.0,\n",
       " 220.6714630126953,\n",
       " 222.71847534179688,\n",
       " 232.26150512695312,\n",
       " 315.6229553222656,\n",
       " 37.2070426940918,\n",
       " 229.7239532470703,\n",
       " 217.8492431640625,\n",
       " 10.604838371276855,\n",
       " 19.022371292114258,\n",
       " 37.93560028076172,\n",
       " 84.73957061767578,\n",
       " 327.0407409667969,\n",
       " 214.59768676757812,\n",
       " 56.52067565917969,\n",
       " 191.76966857910156,\n",
       " 294.4494934082031,\n",
       " 19.022369384765625,\n",
       " 193.71925354003906,\n",
       " 63.30865478515625,\n",
       " 45.772918701171875,\n",
       " 129.9706268310547,\n",
       " 180.43179321289062,\n",
       " 250.47113037109375,\n",
       " 196.37376403808594,\n",
       " 198.73899841308594,\n",
       " 188.86444091796875,\n",
       " 227.07188415527344,\n",
       " 128.07293701171875,\n",
       " 68.04000091552734,\n",
       " 64.1222152709961,\n",
       " 38.23446273803711,\n",
       " 249.0149688720703,\n",
       " 184.77796936035156,\n",
       " 360.1480407714844,\n",
       " 44.73466110229492,\n",
       " 42.11598587036133,\n",
       " 192.11819458007812,\n",
       " 35.400821685791016,\n",
       " 97.38015747070312,\n",
       " 194.30413818359375,\n",
       " 42.80925369262695,\n",
       " 44.543434143066406,\n",
       " 42.42433166503906,\n",
       " 56.90189743041992,\n",
       " 234.06527709960938,\n",
       " 243.6617431640625,\n",
       " 62.42288589477539,\n",
       " 252.9431915283203,\n",
       " 183.18589782714844,\n",
       " 203.84912109375,\n",
       " 254.77392578125,\n",
       " 161.820068359375,\n",
       " 180.32920837402344,\n",
       " 222.31930541992188,\n",
       " 152.3168487548828,\n",
       " 51.04966735839844,\n",
       " 151.60943603515625,\n",
       " 307.4386291503906,\n",
       " 118.83544921875,\n",
       " 438.0,\n",
       " 151.2200469970703,\n",
       " 254.1043701171875,\n",
       " 193.27716064453125,\n",
       " 209.8643341064453,\n",
       " 127.39640808105469,\n",
       " 191.3775177001953,\n",
       " 183.81893920898438,\n",
       " 117.75260925292969,\n",
       " 228.57273864746094,\n",
       " 23.95720863342285,\n",
       " 233.78115844726562,\n",
       " 97.21503448486328,\n",
       " 44.54368591308594,\n",
       " 164.52972412109375,\n",
       " 193.2073211669922,\n",
       " 211.43775939941406,\n",
       " 155.46507263183594,\n",
       " 152.81251525878906,\n",
       " 42.7843017578125,\n",
       " 244.82080078125,\n",
       " 237.32493591308594,\n",
       " 44.977989196777344,\n",
       " 45.713199615478516,\n",
       " 135.24327087402344,\n",
       " 319.02215576171875,\n",
       " 229.94810485839844,\n",
       " 138.15292358398438,\n",
       " 44.782569885253906,\n",
       " 356.8067932128906,\n",
       " 25.201173782348633,\n",
       " 63.22567367553711,\n",
       " 10.583666801452637,\n",
       " 326.8753662109375,\n",
       " 135.40176391601562,\n",
       " 135.1140594482422,\n",
       " 532.0,\n",
       " 211.65414428710938,\n",
       " 718.0,\n",
       " 870.0,\n",
       " 266.5294494628906,\n",
       " 127.38570404052734,\n",
       " 717.0,\n",
       " 35.17371368408203,\n",
       " 806.0,\n",
       " 24.011085510253906,\n",
       " 486.0,\n",
       " 125.32965850830078,\n",
       " 136.98638916015625,\n",
       " 268.9229431152344,\n",
       " 833.0,\n",
       " 25.187910079956055,\n",
       " 1022.0,\n",
       " 134.39276123046875,\n",
       " 133.747314453125,\n",
       " 40.57419967651367,\n",
       " 263.70404052734375,\n",
       " 118.39042663574219,\n",
       " 11.422398567199707,\n",
       " 34.028987884521484,\n",
       " 135.1907958984375,\n",
       " 134.31832885742188,\n",
       " 811.0,\n",
       " 119.01689147949219,\n",
       " 179.1837921142578,\n",
       " 630.0,\n",
       " 180.53115844726562,\n",
       " 348.6877136230469,\n",
       " 187.90469360351562,\n",
       " 179.28834533691406,\n",
       " 629.0,\n",
       " 1060.19482421875,\n",
       " 1101.6490478515625,\n",
       " 1105.2103271484375,\n",
       " 1096.1253662109375,\n",
       " 148.1708221435547,\n",
       " 1041.0,\n",
       " 751.0,\n",
       " 19.426544189453125,\n",
       " 230.66189575195312,\n",
       " 83.83952331542969,\n",
       " 25.036354064941406,\n",
       " 64.22610473632812,\n",
       " 124.74404907226562,\n",
       " 114.77019500732422,\n",
       " 23.751663208007812,\n",
       " 591.0,\n",
       " 437.0,\n",
       " 10.50645923614502,\n",
       " 6.467899322509766,\n",
       " 20.376667022705078,\n",
       " 254.5555877685547,\n",
       " 196.1385498046875,\n",
       " 209.49172973632812,\n",
       " 17.42138671875,\n",
       " 247.71351623535156,\n",
       " 125.36237335205078,\n",
       " 245.20986938476562,\n",
       " 34.469757080078125,\n",
       " 136.1219940185547,\n",
       " 209.1414794921875,\n",
       " 100.0799789428711,\n",
       " 24.02674674987793,\n",
       " 147.7456512451172,\n",
       " 115.25123596191406,\n",
       " 214.50489807128906,\n",
       " 118.65696716308594,\n",
       " 211.2019500732422,\n",
       " 275.3436279296875,\n",
       " 169.51158142089844,\n",
       " 212.0279998779297,\n",
       " 270.8732604980469,\n",
       " 162.3505096435547,\n",
       " 54.80269241333008,\n",
       " 1079.451416015625,\n",
       " 136.35533142089844,\n",
       " 250.0568084716797,\n",
       " 45.910926818847656,\n",
       " 227.25311279296875,\n",
       " 268.70635986328125,\n",
       " 1098.53564453125,\n",
       " 288.6544494628906,\n",
       " 6.490992069244385,\n",
       " 18.886688232421875,\n",
       " 17.467552185058594,\n",
       " 17.462310791015625,\n",
       " 146.80738830566406,\n",
       " 47.073917388916016,\n",
       " 5.433161735534668,\n",
       " 15.472254753112793,\n",
       " 1120.4991455078125,\n",
       " 1116.2728271484375,\n",
       " 82.43953704833984,\n",
       " 1094.5855712890625,\n",
       " 44.0869140625,\n",
       " 37.82029724121094,\n",
       " 134.97132873535156,\n",
       " 202.22470092773438,\n",
       " 31.63103485107422,\n",
       " 117.64249420166016,\n",
       " 63.09306716918945,\n",
       " 45.45387268066406,\n",
       " 44.51898193359375,\n",
       " 25.41619873046875,\n",
       " 134.51856994628906,\n",
       " 68.51323699951172,\n",
       " 50.54536437988281,\n",
       " 59.363922119140625,\n",
       " 70.30248260498047,\n",
       " 33.115333557128906,\n",
       " 42.47638702392578,\n",
       " 88.00201416015625,\n",
       " 15.538671493530273,\n",
       " 33.80820846557617,\n",
       " 17.45941162109375,\n",
       " 81.01692199707031,\n",
       " 86.60453033447266,\n",
       " 101.8827896118164,\n",
       " 85.5538101196289,\n",
       " 15.544107437133789,\n",
       " 42.745906829833984,\n",
       " 10.529255867004395,\n",
       " 99.9384536743164,\n",
       " 96.55059814453125,\n",
       " 1058.225830078125,\n",
       " 147.00942993164062,\n",
       " 23.48078727722168,\n",
       " 473.0,\n",
       " 71.40482330322266,\n",
       " 33.71083450317383,\n",
       " 31.569011688232422,\n",
       " 199.5988311767578,\n",
       " 54.162593841552734,\n",
       " 6.415030479431152,\n",
       " 600.0,\n",
       " 15.388264656066895,\n",
       " 43.854652404785156,\n",
       " 5.4175190925598145,\n",
       " 45.844974517822266,\n",
       " 10.98744010925293,\n",
       " 54.57326126098633,\n",
       " 245.63333129882812,\n",
       " 92.18717193603516,\n",
       " 6.405397891998291,\n",
       " 96.49634552001953,\n",
       " 86.28311157226562,\n",
       " 42.82544708251953,\n",
       " 5.408071994781494,\n",
       " 23.55809783935547,\n",
       " 42.96558380126953,\n",
       " 811.0,\n",
       " 77.68299865722656,\n",
       " 351.6305847167969,\n",
       " 275.1302185058594,\n",
       " 286.4667663574219,\n",
       " 71.18755340576172,\n",
       " 293.5601806640625,\n",
       " 346.1983337402344,\n",
       " 190.43772888183594,\n",
       " 360.35498046875,\n",
       " 17.441036224365234,\n",
       " 129.24282836914062,\n",
       " 6.640435695648193,\n",
       " 6.493015289306641,\n",
       " 5.499494552612305,\n",
       " 17.444257736206055,\n",
       " 6.493469715118408,\n",
       " 98.91007995605469,\n",
       " 73.55907440185547,\n",
       " 1019.0,\n",
       " 56.368202209472656,\n",
       " 1056.0,\n",
       " 845.0,\n",
       " 42.24583435058594,\n",
       " 42.020263671875,\n",
       " 794.0,\n",
       " 38.6148567199707,\n",
       " 81.028076171875,\n",
       " 35.01568603515625,\n",
       " 70.53494262695312,\n",
       " 608.0,\n",
       " 44.1951904296875,\n",
       " 181.57870483398438,\n",
       " 10.426115036010742,\n",
       " 36.20293045043945,\n",
       " 23.3571720123291,\n",
       " 1066.8607177734375,\n",
       " 15.52184009552002,\n",
       " 18.67290496826172,\n",
       " 6.497453212738037,\n",
       " 1092.983154296875,\n",
       " 6.465949535369873,\n",
       " 6.5636982917785645,\n",
       " 34.26335525512695,\n",
       " 15.404106140136719,\n",
       " 1081.971435546875,\n",
       " 1076.7022705078125,\n",
       " 1088.69970703125,\n",
       " 1040.0,\n",
       " 1077.18505859375,\n",
       " 151.85797119140625,\n",
       " 34.353736877441406,\n",
       " 25.05755615234375,\n",
       " 255.80612182617188,\n",
       " 661.0,\n",
       " 416.0,\n",
       " 661.0,\n",
       " 1071.1502685546875,\n",
       " 416.0,\n",
       " 10.488799095153809,\n",
       " 144.74908447265625,\n",
       " 134.4957275390625,\n",
       " 17.411361694335938,\n",
       " 515.0,\n",
       " 569.0,\n",
       " 6.484640121459961,\n",
       " 10.940779685974121,\n",
       " 70.78267669677734,\n",
       " 17.472511291503906,\n",
       " 49.74579620361328,\n",
       " 37.21651077270508,\n",
       " 18.99297523498535,\n",
       " 300.1605224609375,\n",
       " 46.10856628417969,\n",
       " 23.662322998046875,\n",
       " 25.306072235107422,\n",
       " 6.488840103149414,\n",
       " 63.13840866088867,\n",
       " 302.03582763671875,\n",
       " 34.112083435058594,\n",
       " 568.0,\n",
       " 191.4840087890625,\n",
       " 51.46651077270508,\n",
       " 858.0,\n",
       " 10.621603012084961,\n",
       " 354.3771667480469,\n",
       " 51.385658264160156,\n",
       " 294.952880859375,\n",
       " 101.94821166992188,\n",
       " 11.397719383239746,\n",
       " 47.83393096923828,\n",
       " 23.338523864746094,\n",
       " 26.5120849609375,\n",
       " 259.6413269042969,\n",
       " 36.47235870361328,\n",
       " 18.840669631958008,\n",
       " 15.557847023010254,\n",
       " 11.391018867492676,\n",
       " 576.0,\n",
       " 605.0,\n",
       " 38.849124908447266,\n",
       " 11.505403518676758,\n",
       " 5.481317520141602,\n",
       " 20.913528442382812,\n",
       " 46.515682220458984,\n",
       " 23.97913360595703,\n",
       " 104.87605285644531,\n",
       " 11.462348937988281,\n",
       " 146.7610626220703,\n",
       " 150.42938232421875,\n",
       " 577.0,\n",
       " 577.0,\n",
       " 844.0,\n",
       " 150.3569793701172,\n",
       " 20.276397705078125,\n",
       " 11.664509773254395,\n",
       " 23.562175750732422,\n",
       " 238.896728515625,\n",
       " 834.0,\n",
       " 528.0,\n",
       " 55.41693878173828,\n",
       " 158.88214111328125,\n",
       " 115.55783081054688,\n",
       " 44.725650787353516,\n",
       " 25.19290542602539,\n",
       " 44.69879913330078,\n",
       " 1053.9820556640625,\n",
       " 25.53036880493164,\n",
       " 224.7377166748047,\n",
       " 25.730247497558594,\n",
       " 272.3310852050781,\n",
       " 277.6676330566406,\n",
       " 135.21852111816406,\n",
       " 187.9061279296875,\n",
       " 192.54676818847656,\n",
       " 274.92120361328125,\n",
       " 168.25027465820312,\n",
       " 35.16900634765625,\n",
       " 45.83753204345703,\n",
       " 212.37107849121094,\n",
       " 868.0,\n",
       " 864.0,\n",
       " 553.0,\n",
       " 524.0,\n",
       " 582.0,\n",
       " 717.0,\n",
       " 864.0,\n",
       " 465.0,\n",
       " 543.0,\n",
       " 524.0,\n",
       " 760.0,\n",
       " 460.0,\n",
       " 703.0,\n",
       " 1072.964599609375,\n",
       " 813.0,\n",
       " 866.0,\n",
       " 720.0,\n",
       " 1071.06640625,\n",
       " 524.0,\n",
       " 621.0,\n",
       " 350.5765380859375,\n",
       " 621.0,\n",
       " 875.0,\n",
       " 808.0,\n",
       " 45.43819046020508,\n",
       " 154.03485107421875,\n",
       " 806.0,\n",
       " 181.42245483398438,\n",
       " 24.094043731689453,\n",
       " 5.462777137756348,\n",
       " 371.1501770019531,\n",
       " 97.90233612060547,\n",
       " 251.0184326171875,\n",
       " 723.0,\n",
       " 27.099342346191406,\n",
       " 33.68391799926758,\n",
       " 5.692409992218018,\n",
       " 247.1427764892578,\n",
       " 235.8647918701172,\n",
       " 233.84414672851562,\n",
       " 442.0,\n",
       " 880.0,\n",
       " 72.11955261230469,\n",
       " 812.0,\n",
       " 49.83457946777344,\n",
       " 360.2092590332031,\n",
       " 129.55050659179688,\n",
       " 17.411670684814453,\n",
       " 157.91748046875,\n",
       " 281.3768310546875,\n",
       " 64.0496826171875,\n",
       " 119.36286163330078,\n",
       " 94.02452087402344,\n",
       " 5.498402118682861,\n",
       " 6.46829891204834,\n",
       " 6.5419535636901855,\n",
       " 127.65618133544922,\n",
       " 236.17098999023438,\n",
       " 280.8716125488281,\n",
       " 810.0,\n",
       " 1070.49755859375,\n",
       " 840.0,\n",
       " 303.062255859375,\n",
       " 309.9444274902344,\n",
       " 1068.5098876953125,\n",
       " 1045.0,\n",
       " 1044.85302734375,\n",
       " 1078.0,\n",
       " 452.0,\n",
       " 451.0,\n",
       " 839.0,\n",
       " 977.0,\n",
       " 856.0,\n",
       " 353.12506103515625,\n",
       " 1040.0,\n",
       " 660.0,\n",
       " 47.245880126953125,\n",
       " 15.512797355651855,\n",
       " 25.39006233215332,\n",
       " 17.41936683654785,\n",
       " 19.281238555908203,\n",
       " 89.75773620605469,\n",
       " 354.6612854003906,\n",
       " 284.2637939453125,\n",
       " 23.679058074951172,\n",
       " 46.634193420410156,\n",
       " 42.836334228515625,\n",
       " 126.19828796386719,\n",
       " 56.14591598510742,\n",
       " 50.132938385009766,\n",
       " 37.84212112426758,\n",
       " 5.431271553039551,\n",
       " 256.69012451171875,\n",
       " 105.55119323730469,\n",
       " 41.67403030395508,\n",
       " 38.12772750854492,\n",
       " 129.00531005859375,\n",
       " 146.67800903320312,\n",
       " 25.113481521606445,\n",
       " 89.29419708251953,\n",
       " 190.8113555908203,\n",
       " 361.7833251953125,\n",
       " 10.561631202697754,\n",
       " 24.795589447021484,\n",
       " 23.809398651123047,\n",
       " 146.69070434570312,\n",
       " 784.0,\n",
       " 598.0,\n",
       " 356.36492919921875,\n",
       " 232.3662872314453,\n",
       " 438.0,\n",
       " 229.8747100830078,\n",
       " 208.28965759277344,\n",
       " 10.657631874084473,\n",
       " 6.488772869110107,\n",
       " 10.567761421203613,\n",
       " 63.06144714355469,\n",
       " 45.293304443359375,\n",
       " 45.00292205810547,\n",
       " 15.683148384094238,\n",
       " 266.5733947753906,\n",
       " 17.755573272705078,\n",
       " 10.64393424987793,\n",
       " 44.67805862426758,\n",
       " 45.324363708496094,\n",
       " 152.4630889892578,\n",
       " 11.871625900268555,\n",
       " 11.46059513092041,\n",
       " 159.29730224609375,\n",
       " 47.01197814941406,\n",
       " 12.404768943786621,\n",
       " 10.527393341064453,\n",
       " 753.0,\n",
       " 10.592676162719727,\n",
       " 25.165212631225586,\n",
       " 34.65053176879883,\n",
       " 6.504593372344971,\n",
       " 104.79914093017578,\n",
       " 18.66404914855957,\n",
       " 11.586017608642578,\n",
       " 11.512412071228027,\n",
       " 6.507274150848389,\n",
       " 15.543466567993164,\n",
       " 71.62586212158203,\n",
       " 55.99047088623047,\n",
       " 17.432172775268555,\n",
       " 147.5142364501953,\n",
       " 15.471404075622559,\n",
       " 99.97126770019531,\n",
       " 268.31243896484375,\n",
       " 455.0,\n",
       " 44.55474853515625,\n",
       " 15.687516212463379,\n",
       " 10.65594482421875,\n",
       " 23.91407012939453,\n",
       " 47.62403869628906,\n",
       " 34.532535552978516,\n",
       " 31.080507278442383,\n",
       " 102.08232116699219,\n",
       " 444.0,\n",
       " 825.0,\n",
       " 322.8018493652344,\n",
       " 446.0,\n",
       " 1092.473388671875,\n",
       " 121.01734161376953,\n",
       " 7.591175079345703,\n",
       " 665.0,\n",
       " 17.4730224609375,\n",
       " 105.22040557861328,\n",
       " 47.59669876098633,\n",
       " 544.0,\n",
       " 34.24583053588867,\n",
       " 956.0,\n",
       " 941.0,\n",
       " 1016.0,\n",
       " 1016.0,\n",
       " 1013.0,\n",
       " 260.69830322265625,\n",
       " 47.106361389160156,\n",
       " 384.9396057128906,\n",
       " 482.0,\n",
       " 545.0,\n",
       " 434.0,\n",
       " 37.604591369628906,\n",
       " 11.39938735961914,\n",
       " 32.94436264038086,\n",
       " 6.48012638092041,\n",
       " 35.25506591796875,\n",
       " 57.59650421142578,\n",
       " 38.44633102416992,\n",
       " 45.06529998779297,\n",
       " 24.959768295288086,\n",
       " 23.64072608947754,\n",
       " 61.9884147644043,\n",
       " 98.39265441894531,\n",
       " 44.475616455078125,\n",
       " 50.883941650390625,\n",
       " 40.62751770019531,\n",
       " 25.063203811645508,\n",
       " 93.28252410888672,\n",
       " 6.464985370635986,\n",
       " 44.619354248046875,\n",
       " 26.265892028808594,\n",
       " 50.769989013671875,\n",
       " 24.477886199951172,\n",
       " 25.202428817749023,\n",
       " 10.547212600708008,\n",
       " 11.440826416015625,\n",
       " 37.76613998413086,\n",
       " 49.3822135925293,\n",
       " 27.76439094543457,\n",
       " 43.408203125,\n",
       " 23.99882698059082,\n",
       " 15.53674030303955,\n",
       " 15.70608139038086,\n",
       " 24.520675659179688,\n",
       " 6.470297813415527,\n",
       " 44.043025970458984,\n",
       " 37.906578063964844,\n",
       " 864.0,\n",
       " 1063.4376220703125,\n",
       " 866.0,\n",
       " 895.0,\n",
       " 1060.8990478515625,\n",
       " 363.410888671875,\n",
       " 862.0,\n",
       " 870.0,\n",
       " 149.81707763671875,\n",
       " 1066.9058837890625,\n",
       " 353.90155029296875,\n",
       " 178.68359375,\n",
       " 169.30172729492188,\n",
       " 98.45165252685547,\n",
       " 162.38726806640625,\n",
       " 93.51689910888672,\n",
       " 226.16912841796875,\n",
       " 263.3065185546875,\n",
       " 160.37423706054688,\n",
       " 239.42149353027344,\n",
       " 269.92498779296875,\n",
       " 50.9511604309082,\n",
       " 234.8043212890625,\n",
       " 97.59868621826172,\n",
       " 107.19638061523438,\n",
       " 11.835956573486328,\n",
       " 37.941932678222656,\n",
       " 40.331180572509766,\n",
       " 833.0,\n",
       " 37.2286262512207,\n",
       " 37.194156646728516,\n",
       " 795.0,\n",
       " 45.73969650268555,\n",
       " 17.48128890991211,\n",
       " 54.519378662109375,\n",
       " 46.255455017089844,\n",
       " 17.453519821166992,\n",
       " 24.67133140563965,\n",
       " 95.84825134277344,\n",
       " 23.583820343017578,\n",
       " 171.2447052001953,\n",
       " 226.7894287109375,\n",
       " 35.712120056152344,\n",
       " 47.88555145263672,\n",
       " 24.956512451171875,\n",
       " 23.568817138671875,\n",
       " 269.23382568359375,\n",
       " 36.1091423034668,\n",
       " 96.91162109375,\n",
       " 41.98357009887695,\n",
       " 42.271087646484375,\n",
       " 38.3100700378418,\n",
       " 125.95781707763672,\n",
       " 135.97280883789062,\n",
       " 38.36560821533203,\n",
       " 11.406218528747559,\n",
       " 100.18464660644531,\n",
       " 229.2985076904297,\n",
       " 176.73727416992188,\n",
       " 223.31298828125,\n",
       " 44.23385238647461,\n",
       " 44.561309814453125,\n",
       " 38.37874984741211,\n",
       " 322.694580078125,\n",
       " 202.73228454589844,\n",
       " 10.561397552490234,\n",
       " 100.81734466552734,\n",
       " 55.242088317871094,\n",
       " 223.43148803710938,\n",
       " 25.07740592956543,\n",
       " 37.575408935546875,\n",
       " 36.516075134277344,\n",
       " 35.323638916015625,\n",
       " 25.131956100463867,\n",
       " 35.33563995361328,\n",
       " 302.10455322265625,\n",
       " ...]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "171dec72",
   "metadata": {},
   "source": [
    "## 제출 파일 만들기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "dd852ef7",
   "metadata": {},
   "outputs": [],
   "source": [
    "submission = pd.read_csv(data_path+\"sample_submission.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "20b5c167",
   "metadata": {},
   "outputs": [],
   "source": [
    "submission[\"hold_d\"] = np.round(result)\n",
    "submission.to_csv(\"stk_hld_test.csv\", index = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a7c72216",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}