comparing_labelings_or_models.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import math\n",
    "import gzip\n",
    "import pickle\n",
    "from random import gauss\n",
    "from itertools import product\n",
    "from time import sleep\n",
    "from datetime import datetime, timedelta, date\n",
    "from urllib import request\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import lightgbm as lgb\n",
    "import talib\n",
    "\n",
    "from tqdm.notebook import tqdm\n",
    "\n",
    "from sklearn.ensemble import BaggingRegressor\n",
    "from sklearn.linear_model import RidgeCV\n",
    "from sklearn.model_selection import cross_val_score, KFold, TimeSeriesSplit\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### データの準備"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>symbol</th>\n",
       "      <th>side</th>\n",
       "      <th>size</th>\n",
       "      <th>price</th>\n",
       "      <th>tickDirection</th>\n",
       "      <th>trdMatchID</th>\n",
       "      <th>grossValue</th>\n",
       "      <th>homeNotional</th>\n",
       "      <th>foreignNotional</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>timestamp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-03-01</th>\n",
       "      <td>BTCUSD</td>\n",
       "      <td>Sell</td>\n",
       "      <td>9</td>\n",
       "      <td>43179.5</td>\n",
       "      <td>MinusTick</td>\n",
       "      <td>876d94ec-9f6a-528f-9648-55c6133e19d6</td>\n",
       "      <td>20843.224215</td>\n",
       "      <td>9</td>\n",
       "      <td>0.000208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01</th>\n",
       "      <td>BTCUSD</td>\n",
       "      <td>Sell</td>\n",
       "      <td>3</td>\n",
       "      <td>43179.5</td>\n",
       "      <td>ZeroMinusTick</td>\n",
       "      <td>b94b11dc-7747-50c5-9590-a732ab859939</td>\n",
       "      <td>6947.741405</td>\n",
       "      <td>3</td>\n",
       "      <td>0.000069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01</th>\n",
       "      <td>BTCUSD</td>\n",
       "      <td>Sell</td>\n",
       "      <td>2</td>\n",
       "      <td>43179.5</td>\n",
       "      <td>ZeroMinusTick</td>\n",
       "      <td>2c833623-2598-57fa-8408-92900ffdc90f</td>\n",
       "      <td>4631.827603</td>\n",
       "      <td>2</td>\n",
       "      <td>0.000046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01</th>\n",
       "      <td>BTCUSD</td>\n",
       "      <td>Sell</td>\n",
       "      <td>1</td>\n",
       "      <td>43178.5</td>\n",
       "      <td>MinusTick</td>\n",
       "      <td>ce18bf17-3535-534d-8745-bafb1267de27</td>\n",
       "      <td>2315.967437</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01</th>\n",
       "      <td>BTCUSD</td>\n",
       "      <td>Sell</td>\n",
       "      <td>1</td>\n",
       "      <td>43178.0</td>\n",
       "      <td>MinusTick</td>\n",
       "      <td>29f8b6bf-066f-59c2-83b5-fe63a3739954</td>\n",
       "      <td>2315.994256</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000023</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            symbol  side  size    price  tickDirection  \\\n",
       "timestamp                                                \n",
       "2022-03-01  BTCUSD  Sell     9  43179.5      MinusTick   \n",
       "2022-03-01  BTCUSD  Sell     3  43179.5  ZeroMinusTick   \n",
       "2022-03-01  BTCUSD  Sell     2  43179.5  ZeroMinusTick   \n",
       "2022-03-01  BTCUSD  Sell     1  43178.5      MinusTick   \n",
       "2022-03-01  BTCUSD  Sell     1  43178.0      MinusTick   \n",
       "\n",
       "                                      trdMatchID    grossValue  homeNotional  \\\n",
       "timestamp                                                                      \n",
       "2022-03-01  876d94ec-9f6a-528f-9648-55c6133e19d6  20843.224215             9   \n",
       "2022-03-01  b94b11dc-7747-50c5-9590-a732ab859939   6947.741405             3   \n",
       "2022-03-01  2c833623-2598-57fa-8408-92900ffdc90f   4631.827603             2   \n",
       "2022-03-01  ce18bf17-3535-534d-8745-bafb1267de27   2315.967437             1   \n",
       "2022-03-01  29f8b6bf-066f-59c2-83b5-fe63a3739954   2315.994256             1   \n",
       "\n",
       "            foreignNotional  \n",
       "timestamp                    \n",
       "2022-03-01         0.000208  \n",
       "2022-03-01         0.000069  \n",
       "2022-03-01         0.000046  \n",
       "2022-03-01         0.000023  \n",
       "2022-03-01         0.000023  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open(\"./bybit_BTCUSD_2022_3.pkl\", mode=\"rb\") as f:\n",
    "    df = pickle.load(f)\n",
    "\n",
    "df[\"timestamp\"] = pd.to_datetime(df[\"timestamp\"], unit=\"s\")\n",
    "df.set_index(\"timestamp\", inplace=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>op</th>\n",
       "      <th>hi</th>\n",
       "      <th>lo</th>\n",
       "      <th>cl</th>\n",
       "      <th>volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>timestamp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:00:00</th>\n",
       "      <td>43179.5</td>\n",
       "      <td>43593.0</td>\n",
       "      <td>43177.0</td>\n",
       "      <td>43367.0</td>\n",
       "      <td>39018957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:15:00</th>\n",
       "      <td>43367.0</td>\n",
       "      <td>43384.0</td>\n",
       "      <td>43198.5</td>\n",
       "      <td>43367.0</td>\n",
       "      <td>9372331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:30:00</th>\n",
       "      <td>43367.5</td>\n",
       "      <td>43541.5</td>\n",
       "      <td>43318.5</td>\n",
       "      <td>43524.5</td>\n",
       "      <td>22522912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:45:00</th>\n",
       "      <td>43524.5</td>\n",
       "      <td>43745.0</td>\n",
       "      <td>43341.5</td>\n",
       "      <td>43549.5</td>\n",
       "      <td>27202334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 01:00:00</th>\n",
       "      <td>43549.5</td>\n",
       "      <td>43620.0</td>\n",
       "      <td>43279.0</td>\n",
       "      <td>43294.0</td>\n",
       "      <td>13692174</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          op       hi       lo       cl    volume\n",
       "timestamp                                                        \n",
       "2022-03-01 00:00:00  43179.5  43593.0  43177.0  43367.0  39018957\n",
       "2022-03-01 00:15:00  43367.0  43384.0  43198.5  43367.0   9372331\n",
       "2022-03-01 00:30:00  43367.5  43541.5  43318.5  43524.5  22522912\n",
       "2022-03-01 00:45:00  43524.5  43745.0  43341.5  43549.5  27202334\n",
       "2022-03-01 01:00:00  43549.5  43620.0  43279.0  43294.0  13692174"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FREQ='15min'\n",
    "time_bar = pd.DataFrame(columns=['op', 'hi', 'lo', 'cl', 'volume'])\n",
    "time_bar.index.name = \"timestamp\"\n",
    "\n",
    "ohlc = df[\"price\"].resample(FREQ).ohlc()\n",
    "volume = df[\"size\"].resample(FREQ).sum()\n",
    "ohlcv = pd.concat([ohlc, volume], axis=1)\n",
    "ohlcv.index.name = \"timestamp\"\n",
    "ohlcv.columns = [\"op\", \"hi\", \"lo\", \"cl\", \"volume\"]\n",
    "\n",
    "time_bar = pd.concat([time_bar, ohlcv], axis=0)\n",
    "\n",
    "time_bar.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 特徴量の作成、指定"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calc_features(df):\n",
    "    open = df['op']\n",
    "    high = df['hi']\n",
    "    low = df['lo']\n",
    "    close = df['cl']\n",
    "    volume = df['volume']\n",
    "    \n",
    "    orig_columns = df.columns\n",
    "\n",
    "    hilo = (df['hi'] + df['lo']) / 2\n",
    "    df['BBANDS_upperband'], df['BBANDS_middleband'], df['BBANDS_lowerband'] = talib.BBANDS(close, timeperiod=5, nbdevup=2, nbdevdn=2, matype=0)\n",
    "    df['BBANDS_upperband'] -= hilo\n",
    "    df['BBANDS_middleband'] -= hilo\n",
    "    df['BBANDS_lowerband'] -= hilo\n",
    "    df['DEMA'] = talib.DEMA(close, timeperiod=30) - hilo\n",
    "    df['EMA'] = talib.EMA(close, timeperiod=30) - hilo\n",
    "    df['HT_TRENDLINE'] = talib.HT_TRENDLINE(close) - hilo\n",
    "    df['KAMA'] = talib.KAMA(close, timeperiod=30) - hilo\n",
    "    df['MA'] = talib.MA(close, timeperiod=30, matype=0) - hilo\n",
    "    df['MIDPOINT'] = talib.MIDPOINT(close, timeperiod=14) - hilo\n",
    "    df['SMA'] = talib.SMA(close, timeperiod=30) - hilo\n",
    "    df['T3'] = talib.T3(close, timeperiod=5, vfactor=0) - hilo\n",
    "    df['TEMA'] = talib.TEMA(close, timeperiod=30) - hilo\n",
    "    df['TRIMA'] = talib.TRIMA(close, timeperiod=30) - hilo\n",
    "    df['WMA'] = talib.WMA(close, timeperiod=30) - hilo\n",
    "\n",
    "    df['ADX'] = talib.ADX(high, low, close, timeperiod=14)\n",
    "    df['ADXR'] = talib.ADXR(high, low, close, timeperiod=14)\n",
    "    df['APO'] = talib.APO(close, fastperiod=12, slowperiod=26, matype=0)\n",
    "    df['AROON_aroondown'], df['AROON_aroonup'] = talib.AROON(high, low, timeperiod=14)\n",
    "    df['AROONOSC'] = talib.AROONOSC(high, low, timeperiod=14)\n",
    "    df['BOP'] = talib.BOP(open, high, low, close)\n",
    "    df['CCI'] = talib.CCI(high, low, close, timeperiod=14)\n",
    "    df['DX'] = talib.DX(high, low, close, timeperiod=14)\n",
    "    df['MACD_macd'], df['MACD_macdsignal'], df['MACD_macdhist'] = talib.MACD(close, fastperiod=12, slowperiod=26, signalperiod=9)\n",
    "    # skip MACDEXT MACDFIX たぶん同じなので\n",
    "    df['MFI'] = talib.MFI(high, low, close, volume, timeperiod=14)\n",
    "    df['MINUS_DI'] = talib.MINUS_DI(high, low, close, timeperiod=14)\n",
    "    df['MINUS_DM'] = talib.MINUS_DM(high, low, timeperiod=14)\n",
    "    df['MOM'] = talib.MOM(close, timeperiod=10)\n",
    "    df['PLUS_DI'] = talib.PLUS_DI(high, low, close, timeperiod=14)\n",
    "    df['PLUS_DM'] = talib.PLUS_DM(high, low, timeperiod=14)\n",
    "    df['RSI'] = talib.RSI(close, timeperiod=14)\n",
    "    df['STOCH_slowk'], df['STOCH_slowd'] = talib.STOCH(high, low, close, fastk_period=5, slowk_period=3, slowk_matype=0, slowd_period=3, slowd_matype=0)\n",
    "    df['STOCHF_fastk'], df['STOCHF_fastd'] = talib.STOCHF(high, low, close, fastk_period=5, fastd_period=3, fastd_matype=0)\n",
    "    df['STOCHRSI_fastk'], df['STOCHRSI_fastd'] = talib.STOCHRSI(close, timeperiod=14, fastk_period=5, fastd_period=3, fastd_matype=0)\n",
    "    df['TRIX'] = talib.TRIX(close, timeperiod=30)\n",
    "    df['ULTOSC'] = talib.ULTOSC(high, low, close, timeperiod1=7, timeperiod2=14, timeperiod3=28)\n",
    "    df['WILLR'] = talib.WILLR(high, low, close, timeperiod=14)\n",
    "\n",
    "    df['AD'] = talib.AD(high, low, close, volume)\n",
    "    df['ADOSC'] = talib.ADOSC(high, low, close, volume, fastperiod=3, slowperiod=10)\n",
    "    df['OBV'] = talib.OBV(close, volume)\n",
    "\n",
    "    df['ATR'] = talib.ATR(high, low, close, timeperiod=14)\n",
    "    df['NATR'] = talib.NATR(high, low, close, timeperiod=14)\n",
    "    df['TRANGE'] = talib.TRANGE(high, low, close)\n",
    "\n",
    "    df['HT_DCPERIOD'] = talib.HT_DCPERIOD(close)\n",
    "    df['HT_DCPHASE'] = talib.HT_DCPHASE(close)\n",
    "    df['HT_PHASOR_inphase'], df['HT_PHASOR_quadrature'] = talib.HT_PHASOR(close)\n",
    "    df['HT_SINE_sine'], df['HT_SINE_leadsine'] = talib.HT_SINE(close)\n",
    "    df['HT_TRENDMODE'] = talib.HT_TRENDMODE(close)\n",
    "\n",
    "    df['BETA'] = talib.BETA(high, low, timeperiod=5)\n",
    "    df['CORREL'] = talib.CORREL(high, low, timeperiod=30)\n",
    "    df['LINEARREG'] = talib.LINEARREG(close, timeperiod=14) - close\n",
    "    df['LINEARREG_ANGLE'] = talib.LINEARREG_ANGLE(close, timeperiod=14)\n",
    "    df['LINEARREG_INTERCEPT'] = talib.LINEARREG_INTERCEPT(close, timeperiod=14) - close\n",
    "    df['LINEARREG_SLOPE'] = talib.LINEARREG_SLOPE(close, timeperiod=14)\n",
    "    df['STDDEV'] = talib.STDDEV(close, timeperiod=5, nbdev=1)\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>op</th>\n",
       "      <th>hi</th>\n",
       "      <th>lo</th>\n",
       "      <th>cl</th>\n",
       "      <th>volume</th>\n",
       "      <th>BBANDS_upperband</th>\n",
       "      <th>BBANDS_middleband</th>\n",
       "      <th>BBANDS_lowerband</th>\n",
       "      <th>DEMA</th>\n",
       "      <th>EMA</th>\n",
       "      <th>...</th>\n",
       "      <th>HT_SINE_sine</th>\n",
       "      <th>HT_SINE_leadsine</th>\n",
       "      <th>HT_TRENDMODE</th>\n",
       "      <th>BETA</th>\n",
       "      <th>CORREL</th>\n",
       "      <th>LINEARREG</th>\n",
       "      <th>LINEARREG_ANGLE</th>\n",
       "      <th>LINEARREG_INTERCEPT</th>\n",
       "      <th>LINEARREG_SLOPE</th>\n",
       "      <th>STDDEV</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>timestamp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:00:00</th>\n",
       "      <td>43179.5</td>\n",
       "      <td>43593.0</td>\n",
       "      <td>43177.0</td>\n",
       "      <td>43367.0</td>\n",
       "      <td>39018957</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:15:00</th>\n",
       "      <td>43367.0</td>\n",
       "      <td>43384.0</td>\n",
       "      <td>43198.5</td>\n",
       "      <td>43367.0</td>\n",
       "      <td>9372331</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:30:00</th>\n",
       "      <td>43367.5</td>\n",
       "      <td>43541.5</td>\n",
       "      <td>43318.5</td>\n",
       "      <td>43524.5</td>\n",
       "      <td>22522912</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 00:45:00</th>\n",
       "      <td>43524.5</td>\n",
       "      <td>43745.0</td>\n",
       "      <td>43341.5</td>\n",
       "      <td>43549.5</td>\n",
       "      <td>27202334</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 01:00:00</th>\n",
       "      <td>43549.5</td>\n",
       "      <td>43620.0</td>\n",
       "      <td>43279.0</td>\n",
       "      <td>43294.0</td>\n",
       "      <td>13692174</td>\n",
       "      <td>169.260682</td>\n",
       "      <td>-29.1</td>\n",
       "      <td>-227.460682</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>99.180341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-31 22:45:00</th>\n",
       "      <td>45690.5</td>\n",
       "      <td>45800.0</td>\n",
       "      <td>45690.0</td>\n",
       "      <td>45786.0</td>\n",
       "      <td>7734788</td>\n",
       "      <td>73.42218</td>\n",
       "      <td>1.5</td>\n",
       "      <td>-70.42218</td>\n",
       "      <td>-168.148479</td>\n",
       "      <td>193.325447</td>\n",
       "      <td>...</td>\n",
       "      <td>0.995107</td>\n",
       "      <td>0.633786</td>\n",
       "      <td>0</td>\n",
       "      <td>0.383537</td>\n",
       "      <td>0.815266</td>\n",
       "      <td>8.0</td>\n",
       "      <td>84.732303</td>\n",
       "      <td>-133.0</td>\n",
       "      <td>10.846154</td>\n",
       "      <td>35.961090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-31 23:00:00</th>\n",
       "      <td>45780.5</td>\n",
       "      <td>45780.5</td>\n",
       "      <td>45650.0</td>\n",
       "      <td>45680.5</td>\n",
       "      <td>7040102</td>\n",
       "      <td>107.737695</td>\n",
       "      <td>18.35</td>\n",
       "      <td>-71.037695</td>\n",
       "      <td>-147.272226</td>\n",
       "      <td>206.441547</td>\n",
       "      <td>...</td>\n",
       "      <td>0.707763</td>\n",
       "      <td>0.000929</td>\n",
       "      <td>0</td>\n",
       "      <td>0.857388</td>\n",
       "      <td>0.730189</td>\n",
       "      <td>86.014286</td>\n",
       "      <td>80.261024</td>\n",
       "      <td>10.271429</td>\n",
       "      <td>5.826374</td>\n",
       "      <td>44.693847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-31 23:15:00</th>\n",
       "      <td>45680.0</td>\n",
       "      <td>45684.5</td>\n",
       "      <td>45589.0</td>\n",
       "      <td>45681.5</td>\n",
       "      <td>13943959</td>\n",
       "      <td>186.497663</td>\n",
       "      <td>87.55</td>\n",
       "      <td>-11.397663</td>\n",
       "      <td>-75.944683</td>\n",
       "      <td>269.445318</td>\n",
       "      <td>...</td>\n",
       "      <td>0.065495</td>\n",
       "      <td>-0.659277</td>\n",
       "      <td>0</td>\n",
       "      <td>0.534208</td>\n",
       "      <td>0.550943</td>\n",
       "      <td>52.8</td>\n",
       "      <td>-8.930590</td>\n",
       "      <td>54.842857</td>\n",
       "      <td>-0.157143</td>\n",
       "      <td>49.473831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-31 23:30:00</th>\n",
       "      <td>45680.0</td>\n",
       "      <td>45730.0</td>\n",
       "      <td>45372.0</td>\n",
       "      <td>45417.5</td>\n",
       "      <td>36224076</td>\n",
       "      <td>346.787981</td>\n",
       "      <td>100.1</td>\n",
       "      <td>-146.587981</td>\n",
       "      <td>-28.934806</td>\n",
       "      <td>323.666588</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.840907</td>\n",
       "      <td>-0.977283</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.622943</td>\n",
       "      <td>0.604982</td>\n",
       "      <td>228.242857</td>\n",
       "      <td>-84.716867</td>\n",
       "      <td>368.828571</td>\n",
       "      <td>-10.814286</td>\n",
       "      <td>123.343991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-31 23:45:00</th>\n",
       "      <td>45420.5</td>\n",
       "      <td>45560.0</td>\n",
       "      <td>45188.0</td>\n",
       "      <td>45509.5</td>\n",
       "      <td>79193719</td>\n",
       "      <td>506.465629</td>\n",
       "      <td>241.0</td>\n",
       "      <td>-24.465629</td>\n",
       "      <td>125.215346</td>\n",
       "      <td>477.107454</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.991247</td>\n",
       "      <td>-0.607567</td>\n",
       "      <td>0</td>\n",
       "      <td>0.175338</td>\n",
       "      <td>0.760561</td>\n",
       "      <td>79.271429</td>\n",
       "      <td>-86.749270</td>\n",
       "      <td>308.157143</td>\n",
       "      <td>-17.606593</td>\n",
       "      <td>132.732814</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2976 rows × 67 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                          op       hi       lo       cl    volume  \\\n",
       "timestamp                                                           \n",
       "2022-03-01 00:00:00  43179.5  43593.0  43177.0  43367.0  39018957   \n",
       "2022-03-01 00:15:00  43367.0  43384.0  43198.5  43367.0   9372331   \n",
       "2022-03-01 00:30:00  43367.5  43541.5  43318.5  43524.5  22522912   \n",
       "2022-03-01 00:45:00  43524.5  43745.0  43341.5  43549.5  27202334   \n",
       "2022-03-01 01:00:00  43549.5  43620.0  43279.0  43294.0  13692174   \n",
       "...                      ...      ...      ...      ...       ...   \n",
       "2022-03-31 22:45:00  45690.5  45800.0  45690.0  45786.0   7734788   \n",
       "2022-03-31 23:00:00  45780.5  45780.5  45650.0  45680.5   7040102   \n",
       "2022-03-31 23:15:00  45680.0  45684.5  45589.0  45681.5  13943959   \n",
       "2022-03-31 23:30:00  45680.0  45730.0  45372.0  45417.5  36224076   \n",
       "2022-03-31 23:45:00  45420.5  45560.0  45188.0  45509.5  79193719   \n",
       "\n",
       "                    BBANDS_upperband BBANDS_middleband BBANDS_lowerband  \\\n",
       "timestamp                                                                 \n",
       "2022-03-01 00:00:00              NaN               NaN              NaN   \n",
       "2022-03-01 00:15:00              NaN               NaN              NaN   \n",
       "2022-03-01 00:30:00              NaN               NaN              NaN   \n",
       "2022-03-01 00:45:00              NaN               NaN              NaN   \n",
       "2022-03-01 01:00:00       169.260682             -29.1      -227.460682   \n",
       "...                              ...               ...              ...   \n",
       "2022-03-31 22:45:00         73.42218               1.5        -70.42218   \n",
       "2022-03-31 23:00:00       107.737695             18.35       -71.037695   \n",
       "2022-03-31 23:15:00       186.497663             87.55       -11.397663   \n",
       "2022-03-31 23:30:00       346.787981             100.1      -146.587981   \n",
       "2022-03-31 23:45:00       506.465629             241.0       -24.465629   \n",
       "\n",
       "                           DEMA         EMA  ... HT_SINE_sine  \\\n",
       "timestamp                                    ...                \n",
       "2022-03-01 00:00:00         NaN         NaN  ...          NaN   \n",
       "2022-03-01 00:15:00         NaN         NaN  ...          NaN   \n",
       "2022-03-01 00:30:00         NaN         NaN  ...          NaN   \n",
       "2022-03-01 00:45:00         NaN         NaN  ...          NaN   \n",
       "2022-03-01 01:00:00         NaN         NaN  ...          NaN   \n",
       "...                         ...         ...  ...          ...   \n",
       "2022-03-31 22:45:00 -168.148479  193.325447  ...     0.995107   \n",
       "2022-03-31 23:00:00 -147.272226  206.441547  ...     0.707763   \n",
       "2022-03-31 23:15:00  -75.944683  269.445318  ...     0.065495   \n",
       "2022-03-31 23:30:00  -28.934806  323.666588  ...    -0.840907   \n",
       "2022-03-31 23:45:00  125.215346  477.107454  ...    -0.991247   \n",
       "\n",
       "                    HT_SINE_leadsine HT_TRENDMODE      BETA    CORREL  \\\n",
       "timestamp                                                               \n",
       "2022-03-01 00:00:00              NaN            0       NaN       NaN   \n",
       "2022-03-01 00:15:00              NaN            0       NaN       NaN   \n",
       "2022-03-01 00:30:00              NaN            0       NaN       NaN   \n",
       "2022-03-01 00:45:00              NaN            0       NaN       NaN   \n",
       "2022-03-01 01:00:00              NaN            0       NaN       NaN   \n",
       "...                              ...          ...       ...       ...   \n",
       "2022-03-31 22:45:00         0.633786            0  0.383537  0.815266   \n",
       "2022-03-31 23:00:00         0.000929            0  0.857388  0.730189   \n",
       "2022-03-31 23:15:00        -0.659277            0  0.534208  0.550943   \n",
       "2022-03-31 23:30:00        -0.977283            0 -0.622943  0.604982   \n",
       "2022-03-31 23:45:00        -0.607567            0  0.175338  0.760561   \n",
       "\n",
       "                      LINEARREG LINEARREG_ANGLE LINEARREG_INTERCEPT  \\\n",
       "timestamp                                                             \n",
       "2022-03-01 00:00:00         NaN             NaN                 NaN   \n",
       "2022-03-01 00:15:00         NaN             NaN                 NaN   \n",
       "2022-03-01 00:30:00         NaN             NaN                 NaN   \n",
       "2022-03-01 00:45:00         NaN             NaN                 NaN   \n",
       "2022-03-01 01:00:00         NaN             NaN                 NaN   \n",
       "...                         ...             ...                 ...   \n",
       "2022-03-31 22:45:00         8.0       84.732303              -133.0   \n",
       "2022-03-31 23:00:00   86.014286       80.261024           10.271429   \n",
       "2022-03-31 23:15:00        52.8       -8.930590           54.842857   \n",
       "2022-03-31 23:30:00  228.242857      -84.716867          368.828571   \n",
       "2022-03-31 23:45:00   79.271429      -86.749270          308.157143   \n",
       "\n",
       "                    LINEARREG_SLOPE      STDDEV  \n",
       "timestamp                                        \n",
       "2022-03-01 00:00:00             NaN         NaN  \n",
       "2022-03-01 00:15:00             NaN         NaN  \n",
       "2022-03-01 00:30:00             NaN         NaN  \n",
       "2022-03-01 00:45:00             NaN         NaN  \n",
       "2022-03-01 01:00:00             NaN   99.180341  \n",
       "...                             ...         ...  \n",
       "2022-03-31 22:45:00       10.846154   35.961090  \n",
       "2022-03-31 23:00:00        5.826374   44.693847  \n",
       "2022-03-31 23:15:00       -0.157143   49.473831  \n",
       "2022-03-31 23:30:00      -10.814286  123.343991  \n",
       "2022-03-31 23:45:00      -17.606593  132.732814  \n",
       "\n",
       "[2976 rows x 67 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "calc_features(time_bar)\n",
    "\n",
    "time_bar.to_pickle(\"df_features.pkl\")\n",
    "display(time_bar)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['ADX', 'ADXR', 'APO', 'AROONOSC', 'AROON_aroondown', 'AROON_aroonup', 'BBANDS_lowerband', 'BBANDS_middleband', 'BBANDS_upperband', 'BETA', 'CCI', 'DEMA', 'DX', 'EMA', 'HT_DCPERIOD', 'HT_DCPHASE', 'HT_PHASOR_inphase', 'HT_PHASOR_quadrature', 'HT_TRENDLINE', 'HT_TRENDMODE', 'KAMA', 'LINEARREG', 'LINEARREG_ANGLE', 'LINEARREG_INTERCEPT', 'LINEARREG_SLOPE', 'MA', 'MACD_macd', 'MACD_macdhist', 'MACD_macdsignal', 'MFI', 'MIDPOINT', 'MOM', 'RSI', 'STDDEV', 'STOCHF_fastk', 'STOCH_slowd', 'STOCH_slowk', 'T3', 'TEMA', 'TRIMA', 'ULTOSC', 'WILLR', 'WMA']\n"
     ]
    }
   ],
   "source": [
    "#以下のコードでは学習に使う特徴量カラムを指定する。\n",
    "features =sorted([\n",
    "    \n",
    "    'ADX',\n",
    "    'ADXR',\n",
    "    'APO',\n",
    "    'AROON_aroondown',\n",
    "    'AROON_aroonup',\n",
    "    'AROONOSC',\n",
    "    'CCI',\n",
    "    'DX',\n",
    "    'MACD_macd',\n",
    "    'MACD_macdsignal',\n",
    "    'MACD_macdhist',\n",
    "    'MFI',\n",
    "#     'MINUS_DI',\n",
    "#     'MINUS_DM',\n",
    "    'MOM',\n",
    "#     'PLUS_DI',\n",
    "#     'PLUS_DM',\n",
    "    'RSI',\n",
    "    'STOCH_slowk',\n",
    "    'STOCH_slowd',\n",
    "    'STOCHF_fastk',\n",
    "#     'STOCHRSI_fastd',\n",
    "    'ULTOSC',\n",
    "    'WILLR',\n",
    "#     'ADOSC',\n",
    "#     'NATR',\n",
    "    'HT_DCPERIOD',\n",
    "    'HT_DCPHASE',\n",
    "    'HT_PHASOR_inphase',\n",
    "    'HT_PHASOR_quadrature',\n",
    "    'HT_TRENDMODE',\n",
    "    'BETA',\n",
    "    'LINEARREG',\n",
    "    'LINEARREG_ANGLE',\n",
    "    'LINEARREG_INTERCEPT',\n",
    "    'LINEARREG_SLOPE',\n",
    "    'STDDEV',\n",
    "    'BBANDS_upperband',\n",
    "    'BBANDS_middleband',\n",
    "    'BBANDS_lowerband',\n",
    "    'DEMA',\n",
    "    'EMA',\n",
    "    'HT_TRENDLINE',\n",
    "    'KAMA',\n",
    "    'MA',\n",
    "    'MIDPOINT',\n",
    "    'T3',\n",
    "    'TEMA',\n",
    "    'TRIMA',\n",
    "    'WMA',\n",
    "])\n",
    "\n",
    "print(features)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ラベリング方法での比較"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### opとclの大小比較による0か1でのラベリング"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 目的変数の作成\n",
    "y = time_bar.cl / time_bar.op\n",
    "y = y.apply(lambda x: 1 if x > 1 else 0)\n",
    "\n",
    "time_bar[\"target\"] = y\n",
    "time_bar.dropna(inplace=True)\n",
    "\n",
    "time_bar.to_pickle('df_target.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>op</th>\n",
       "      <th>hi</th>\n",
       "      <th>lo</th>\n",
       "      <th>cl</th>\n",
       "      <th>volume</th>\n",
       "      <th>BBANDS_upperband</th>\n",
       "      <th>BBANDS_middleband</th>\n",
       "      <th>BBANDS_lowerband</th>\n",
       "      <th>DEMA</th>\n",
       "      <th>EMA</th>\n",
       "      <th>...</th>\n",
       "      <th>HT_SINE_leadsine</th>\n",
       "      <th>HT_TRENDMODE</th>\n",
       "      <th>BETA</th>\n",
       "      <th>CORREL</th>\n",
       "      <th>LINEARREG</th>\n",
       "      <th>LINEARREG_ANGLE</th>\n",
       "      <th>LINEARREG_INTERCEPT</th>\n",
       "      <th>LINEARREG_SLOPE</th>\n",
       "      <th>STDDEV</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>timestamp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-03-01 22:00:00</th>\n",
       "      <td>43868.0</td>\n",
       "      <td>43868.0</td>\n",
       "      <td>43774.0</td>\n",
       "      <td>43840.0</td>\n",
       "      <td>4127124</td>\n",
       "      <td>222.605745</td>\n",
       "      <td>70.3</td>\n",
       "      <td>-82.005745</td>\n",
       "      <td>72.536391</td>\n",
       "      <td>-9.417522</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.362494</td>\n",
       "      <td>0</td>\n",
       "      <td>0.522806</td>\n",
       "      <td>0.862698</td>\n",
       "      <td>83.157143</td>\n",
       "      <td>82.122515</td>\n",
       "      <td>-10.8</td>\n",
       "      <td>7.227473</td>\n",
       "      <td>76.152873</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 22:15:00</th>\n",
       "      <td>43840.0</td>\n",
       "      <td>43960.0</td>\n",
       "      <td>43840.0</td>\n",
       "      <td>43945.5</td>\n",
       "      <td>6538919</td>\n",
       "      <td>142.120738</td>\n",
       "      <td>-9.3</td>\n",
       "      <td>-160.720738</td>\n",
       "      <td>4.971313</td>\n",
       "      <td>-79.777682</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.557421</td>\n",
       "      <td>0</td>\n",
       "      <td>0.809185</td>\n",
       "      <td>0.866785</td>\n",
       "      <td>-13.185714</td>\n",
       "      <td>82.101170</td>\n",
       "      <td>-106.885714</td>\n",
       "      <td>7.207692</td>\n",
       "      <td>75.710369</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 22:30:00</th>\n",
       "      <td>43945.5</td>\n",
       "      <td>44169.5</td>\n",
       "      <td>43939.5</td>\n",
       "      <td>44150.0</td>\n",
       "      <td>17811845</td>\n",
       "      <td>115.528242</td>\n",
       "      <td>-135.0</td>\n",
       "      <td>-385.528242</td>\n",
       "      <td>-113.817049</td>\n",
       "      <td>-213.001702</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.689043</td>\n",
       "      <td>0</td>\n",
       "      <td>0.327546</td>\n",
       "      <td>0.890181</td>\n",
       "      <td>-173.871429</td>\n",
       "      <td>83.782552</td>\n",
       "      <td>-293.2</td>\n",
       "      <td>9.179121</td>\n",
       "      <td>125.264121</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 22:45:00</th>\n",
       "      <td>44150.0</td>\n",
       "      <td>44231.0</td>\n",
       "      <td>44058.5</td>\n",
       "      <td>44158.0</td>\n",
       "      <td>9410449</td>\n",
       "      <td>120.633162</td>\n",
       "      <td>-152.55</td>\n",
       "      <td>-425.733162</td>\n",
       "      <td>-170.944514</td>\n",
       "      <td>-282.832238</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.758299</td>\n",
       "      <td>0</td>\n",
       "      <td>0.417686</td>\n",
       "      <td>0.901287</td>\n",
       "      <td>-114.042857</td>\n",
       "      <td>86.710757</td>\n",
       "      <td>-340.242857</td>\n",
       "      <td>17.400000</td>\n",
       "      <td>136.591581</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-03-01 23:00:00</th>\n",
       "      <td>44158.0</td>\n",
       "      <td>44262.5</td>\n",
       "      <td>44078.0</td>\n",
       "      <td>44125.5</td>\n",
       "      <td>7828911</td>\n",
       "      <td>129.975116</td>\n",
       "      <td>-126.45</td>\n",
       "      <td>-382.875116</td>\n",
       "      <td>-170.749582</td>\n",
       "      <td>-291.326932</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.749985</td>\n",
       "      <td>0</td>\n",
       "      <td>0.352744</td>\n",
       "      <td>0.917209</td>\n",
       "      <td>-42.685714</td>\n",
       "      <td>87.190608</td>\n",
       "      <td>-307.6</td>\n",
       "      <td>20.378022</td>\n",
       "      <td>128.212558</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 68 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                          op       hi       lo       cl    volume  \\\n",
       "timestamp                                                           \n",
       "2022-03-01 22:00:00  43868.0  43868.0  43774.0  43840.0   4127124   \n",
       "2022-03-01 22:15:00  43840.0  43960.0  43840.0  43945.5   6538919   \n",
       "2022-03-01 22:30:00  43945.5  44169.5  43939.5  44150.0  17811845   \n",
       "2022-03-01 22:45:00  44150.0  44231.0  44058.5  44158.0   9410449   \n",
       "2022-03-01 23:00:00  44158.0  44262.5  44078.0  44125.5   7828911   \n",
       "\n",
       "                    BBANDS_upperband BBANDS_middleband BBANDS_lowerband  \\\n",
       "timestamp                                                                 \n",
       "2022-03-01 22:00:00       222.605745              70.3       -82.005745   \n",
       "2022-03-01 22:15:00       142.120738              -9.3      -160.720738   \n",
       "2022-03-01 22:30:00       115.528242            -135.0      -385.528242   \n",
       "2022-03-01 22:45:00       120.633162           -152.55      -425.733162   \n",
       "2022-03-01 23:00:00       129.975116           -126.45      -382.875116   \n",
       "\n",
       "                           DEMA         EMA  ... HT_SINE_leadsine  \\\n",
       "timestamp                                    ...                    \n",
       "2022-03-01 22:00:00   72.536391   -9.417522  ...        -0.362494   \n",
       "2022-03-01 22:15:00    4.971313  -79.777682  ...        -0.557421   \n",
       "2022-03-01 22:30:00 -113.817049 -213.001702  ...        -0.689043   \n",
       "2022-03-01 22:45:00 -170.944514 -282.832238  ...        -0.758299   \n",
       "2022-03-01 23:00:00 -170.749582 -291.326932  ...        -0.749985   \n",
       "\n",
       "                    HT_TRENDMODE      BETA    CORREL   LINEARREG  \\\n",
       "timestamp                                                          \n",
       "2022-03-01 22:00:00            0  0.522806  0.862698   83.157143   \n",
       "2022-03-01 22:15:00            0  0.809185  0.866785  -13.185714   \n",
       "2022-03-01 22:30:00            0  0.327546  0.890181 -173.871429   \n",
       "2022-03-01 22:45:00            0  0.417686  0.901287 -114.042857   \n",
       "2022-03-01 23:00:00            0  0.352744  0.917209  -42.685714   \n",
       "\n",
       "                    LINEARREG_ANGLE LINEARREG_INTERCEPT LINEARREG_SLOPE  \\\n",
       "timestamp                                                                 \n",
       "2022-03-01 22:00:00       82.122515               -10.8        7.227473   \n",
       "2022-03-01 22:15:00       82.101170         -106.885714        7.207692   \n",
       "2022-03-01 22:30:00       83.782552              -293.2        9.179121   \n",
       "2022-03-01 22:45:00       86.710757         -340.242857       17.400000   \n",
       "2022-03-01 23:00:00       87.190608              -307.6       20.378022   \n",
       "\n",
       "                         STDDEV  target  \n",
       "timestamp                                \n",
       "2022-03-01 22:00:00   76.152873       0  \n",
       "2022-03-01 22:15:00   75.710369       1  \n",
       "2022-03-01 22:30:00  125.264121       1  \n",
       "2022-03-01 22:45:00  136.591581       1  \n",
       "2022-03-01 23:00:00  128.212558       0  \n",
       "\n",
       "[5 rows x 68 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "time_bar.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### returnによるラベリング"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'毎時刻、targetが1のときだけトレードした場合の累積リターン'}, xlabel='timestamp'>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.read_pickle('df_target.pkl')\n",
    "ret = df['op'].copy()\n",
    "\n",
    "for i in range(df[\"op\"].values.size):\n",
    "    ret[i] = df[\"cl\"][i]/df[\"op\"][i] -1\n",
    "    \n",
    "df['ret'] = ret\n",
    "\n",
    "df.to_pickle('df_y.pkl')\n",
    "\n",
    "plt.title('毎時刻、targetが1のときだけトレードした場合の累積リターン', fontname=\"MS Gothic\")\n",
    "df['ret'][df['target'] >0].cumsum().plot(label='買い')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_pickle('df_y.pkl')\n",
    "df.dropna()\n",
    "\n",
    "model = RidgeCV(alphas=np.logspace(-7, 7, num=20))\n",
    "cv_indicies = list(KFold().split(df))\n",
    "\n",
    "def my_cross_val_predict(estimator, X, y=None, cv=None):\n",
    "    y_pred = y.copy()\n",
    "    \n",
    "    for train_idex, val_idx in cv:\n",
    "        estimator.fit(X[train_idex], y[train_idex])\n",
    "        y_pred[val_idx] = estimator.predict(X[val_idx])\n",
    "    return y_pred\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "毎時刻、ret_predがプラスの時だけリターンした場合の累積リターン\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x20180269210>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#　retを予測\n",
    "df['ret_pred'] = my_cross_val_predict(model, df[features].values, df['ret'].values, cv_indicies)\n",
    "\n",
    "#　毎時刻ret>0のときのみトレード\n",
    "print('毎時刻、ret_predがプラスの時だけリターンした場合の累積リターン')\n",
    "df['ret'][df['ret_pred']>0].cumsum().plot(label='予測結果')\n",
    "df['ret'][df['ret']>0].cumsum().plot(label='理論値')\n",
    "plt.legend(bbox_to_anchor=(1.05, 1), prop={\"family\":\"MS Gothic\"})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "毎時刻、target_predがプラスの時だけリターンした場合の累積リターン\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x201804bccd0>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#　targetを予測(比較用にretの予測も)\n",
    "df['target_pred'] = my_cross_val_predict(model, df[features].values, df['target'].values, cv_indicies)\n",
    "df['ret_pred'] = my_cross_val_predict(model, df[features].values, df['ret'].values, cv_indicies)\n",
    "\n",
    "\n",
    "#　毎時刻target>0のときのみトレード\n",
    "print('毎時刻、target_predがプラスの時だけリターンした場合の累積リターン')\n",
    "df['ret'][df['target_pred']>0].cumsum().plot(label='target予測の結果')\n",
    "df['ret'][df['ret_pred']>0].cumsum().plot(label='ret予測の結果')\n",
    "\n",
    "df['ret'][df['ret']>0].cumsum().plot(label='理論値')\n",
    "plt.legend(bbox_to_anchor=(1.05, 1), prop={\"family\":\"MS Gothic\"})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### モデルによる比較"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_pickle('df_y.pkl')\n",
    "df.dropna()\n",
    "\n",
    "# モデル\n",
    "# model = RidgeCV(alphas=np.logspace(-7, 7, num=20))\n",
    "\n",
    "\n",
    "# アンサンブル\n",
    "# model = BaggingRegressor(model, random_state=1, n_jobs=1)\n",
    "\n",
    "cv_indicies = list(KFold().split(df))\n",
    "\n",
    "def my_cross_val_predict(estimator, X, y=None, cv=None):\n",
    "    y_pred = y.copy()\n",
    "    \n",
    "    for train_idex, val_idx in cv:\n",
    "        estimator.fit(X[train_idex], y[train_idex])\n",
    "        y_pred[val_idx] = estimator.predict(X[val_idx])\n",
    "    return y_pred\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = lgb.LGBMRegressor(n_jobs=-1, random_state=1)\n",
    "df['LGBM_pred'] = my_cross_val_predict(model, df[features].values, df['ret'].values, cv_indicies)\n",
    "\n",
    "model = RidgeCV(alphas=np.logspace(-7, 7, num=20))\n",
    "df['Ridge_pred'] = my_cross_val_predict(model, df[features].values, df['ret'].values, cv_indicies)\n",
    "\n",
    "model = lgb.LGBMRegressor(n_jobs=-1, random_state=1)\n",
    "model = BaggingRegressor(model, random_state=1, n_jobs=1)\n",
    "df['LGBM_ensemble_pred'] = my_cross_val_predict(model, df[features].values, df['ret'].values, cv_indicies)\n",
    "\n",
    "model = RidgeCV(alphas=np.logspace(-7, 7, num=20))\n",
    "model = BaggingRegressor(model, random_state=1, n_jobs=1)\n",
    "df['Ridge_ensemble_pred'] = my_cross_val_predict(model, df[features].values, df['ret'].values, cv_indicies)\n",
    "\n",
    "df.to_pickle('df_models.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "毎時刻ret_predがプラスだった時の累積リターンのモデル事の比較\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x201a0202d10>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.read_pickle('df_models.pkl')\n",
    "\n",
    "print('毎時刻ret_predがプラスだった時の累積リターンのモデル事の比較')\n",
    "df['ret'][df['LGBM_pred'] > 0].cumsum().plot(label='LGBM')\n",
    "df['ret'][df['Ridge_pred'] > 0].cumsum().plot(label='RidgeCV')\n",
    "df['ret'][df['LGBM_ensemble_pred'] > 0].cumsum().plot(label='LGBM_ensemble')\n",
    "df['ret'][df['Ridge_ensemble_pred'] > 0].cumsum().plot(label='RidgeCV_ensemble')\n",
    "df['ret'][df['ret']>0].cumsum().plot(label='理論値')\n",
    "\n",
    "plt.title('累積リターンのモデルごとの比較',  fontname=\"MS Gothic\")\n",
    "plt.legend(bbox_to_anchor=(1.05, 1), prop={\"family\":\"MS Gothic\"})\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.7 64-bit (microsoft store)",
   "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": "3.10.8"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "8dd4c1bbea936f30cb17f7d968b3a7be23d8106fb74f540a4bc6b75ca162dc08"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}