neural-network-hourly-v2.ipynb

{
 "cells": [
  {
   "metadata": {
    "collapsed": true,
    "ExecuteTime": {
     "end_time": "2024-08-24T06:41:46.629056Z",
     "start_time": "2024-08-24T06:41:45.432278Z"
    }
   },
   "cell_type": "code",
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import yfinance as yf"
   ],
   "id": "initial_id",
   "outputs": [],
   "execution_count": 1
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "### DATA    ",
   "id": "87c1920adf33ef94"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:41:46.657477Z",
     "start_time": "2024-08-24T06:41:46.653111Z"
    }
   },
   "cell_type": "code",
   "source": [
    "symbol = \"EURUSD=X\"\n",
    "start = \"2022-08-27\"\n",
    "end = \"2024-08-20\"\n",
    "interval = '1h'"
   ],
   "id": "22e1b7311c9a17e9",
   "outputs": [],
   "execution_count": 2
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "### Variables",
   "id": "7797cc32ab1e7262"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:41:46.701270Z",
     "start_time": "2024-08-24T06:41:46.697715Z"
    }
   },
   "cell_type": "code",
   "source": [
    "sma_short =20\n",
    "sma_long=50\n"
   ],
   "id": "71a2b1fafa012d31",
   "outputs": [],
   "execution_count": 3
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## Downloading Data",
   "id": "a5ee5d9e40360776"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:41:48.916777Z",
     "start_time": "2024-08-24T06:41:47.911662Z"
    }
   },
   "cell_type": "code",
   "source": [
    "df = yf.download(symbol, start, end, interval=interval)\n",
    "#df.rename(columns={\"Close\": \"price\"}, inplace=True)\n",
    "df.drop(columns=['Adj Close', 'Volume'], inplace=True)\n",
    "df.index = df.index.tz_convert('utc')\n",
    "df['returns'] = np.log(df['Close'].div(df['Close'].shift(1)))\n",
    "df"
   ],
   "id": "b083fe3843223ade",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[*********************100%%**********************]  1 of 1 completed\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "                               Open      High       Low     Close   returns\n",
       "Datetime                                                                   \n",
       "2022-08-28 23:00:00+00:00  0.995520  0.995520  0.993838  0.993838       NaN\n",
       "2022-08-29 00:00:00+00:00  0.994233  0.995124  0.993246  0.993345 -0.000497\n",
       "2022-08-29 01:00:00+00:00  0.993345  0.993937  0.992162  0.993443  0.000099\n",
       "2022-08-29 02:00:00+00:00  0.993147  0.993937  0.992753  0.993049 -0.000397\n",
       "2022-08-29 03:00:00+00:00  0.992950  0.993345  0.992457  0.992654 -0.000397\n",
       "...                             ...       ...       ...       ...       ...\n",
       "2024-08-19 18:00:00+00:00  1.108156  1.108770  1.108033  1.108525  0.000554\n",
       "2024-08-19 19:00:00+00:00  1.108402  1.108893  1.108279  1.108647  0.000111\n",
       "2024-08-19 20:00:00+00:00  1.108647  1.108893  1.108525  1.108893  0.000222\n",
       "2024-08-19 21:00:00+00:00  1.108525  1.109139  1.108279  1.108647 -0.000222\n",
       "2024-08-19 22:00:00+00:00  1.109016  1.109016  1.108647  1.108770  0.000111\n",
       "\n",
       "[12247 rows x 5 columns]"
      ],
      "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>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>returns</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-28 23:00:00+00:00</th>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 00:00:00+00:00</th>\n",
       "      <td>0.994233</td>\n",
       "      <td>0.995124</td>\n",
       "      <td>0.993246</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>-0.000497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 01:00:00+00:00</th>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992162</td>\n",
       "      <td>0.993443</td>\n",
       "      <td>0.000099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 02:00:00+00:00</th>\n",
       "      <td>0.993147</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992753</td>\n",
       "      <td>0.993049</td>\n",
       "      <td>-0.000397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 03:00:00+00:00</th>\n",
       "      <td>0.992950</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.992457</td>\n",
       "      <td>0.992654</td>\n",
       "      <td>-0.000397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>1.108156</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>1.108033</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>0.000554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>1.108402</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>0.000111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>0.000222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.109139</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>-0.000222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>0.000111</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12247 rows × 5 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 4
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:41:51.802235Z",
     "start_time": "2024-08-24T06:41:51.784191Z"
    }
   },
   "cell_type": "code",
   "source": [
    "df['day_of_week'] = df.index.dayofweek\n",
    "df"
   ],
   "id": "7152b9341859846c",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                               Open      High       Low     Close   returns  \\\n",
       "Datetime                                                                      \n",
       "2022-08-28 23:00:00+00:00  0.995520  0.995520  0.993838  0.993838       NaN   \n",
       "2022-08-29 00:00:00+00:00  0.994233  0.995124  0.993246  0.993345 -0.000497   \n",
       "2022-08-29 01:00:00+00:00  0.993345  0.993937  0.992162  0.993443  0.000099   \n",
       "2022-08-29 02:00:00+00:00  0.993147  0.993937  0.992753  0.993049 -0.000397   \n",
       "2022-08-29 03:00:00+00:00  0.992950  0.993345  0.992457  0.992654 -0.000397   \n",
       "...                             ...       ...       ...       ...       ...   \n",
       "2024-08-19 18:00:00+00:00  1.108156  1.108770  1.108033  1.108525  0.000554   \n",
       "2024-08-19 19:00:00+00:00  1.108402  1.108893  1.108279  1.108647  0.000111   \n",
       "2024-08-19 20:00:00+00:00  1.108647  1.108893  1.108525  1.108893  0.000222   \n",
       "2024-08-19 21:00:00+00:00  1.108525  1.109139  1.108279  1.108647 -0.000222   \n",
       "2024-08-19 22:00:00+00:00  1.109016  1.109016  1.108647  1.108770  0.000111   \n",
       "\n",
       "                           day_of_week  \n",
       "Datetime                                \n",
       "2022-08-28 23:00:00+00:00            6  \n",
       "2022-08-29 00:00:00+00:00            0  \n",
       "2022-08-29 01:00:00+00:00            0  \n",
       "2022-08-29 02:00:00+00:00            0  \n",
       "2022-08-29 03:00:00+00:00            0  \n",
       "...                                ...  \n",
       "2024-08-19 18:00:00+00:00            0  \n",
       "2024-08-19 19:00:00+00:00            0  \n",
       "2024-08-19 20:00:00+00:00            0  \n",
       "2024-08-19 21:00:00+00:00            0  \n",
       "2024-08-19 22:00:00+00:00            0  \n",
       "\n",
       "[12247 rows x 6 columns]"
      ],
      "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>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>returns</th>\n",
       "      <th>day_of_week</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-28 23:00:00+00:00</th>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 00:00:00+00:00</th>\n",
       "      <td>0.994233</td>\n",
       "      <td>0.995124</td>\n",
       "      <td>0.993246</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>-0.000497</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 01:00:00+00:00</th>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992162</td>\n",
       "      <td>0.993443</td>\n",
       "      <td>0.000099</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 02:00:00+00:00</th>\n",
       "      <td>0.993147</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992753</td>\n",
       "      <td>0.993049</td>\n",
       "      <td>-0.000397</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 03:00:00+00:00</th>\n",
       "      <td>0.992950</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.992457</td>\n",
       "      <td>0.992654</td>\n",
       "      <td>-0.000397</td>\n",
       "      <td>0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>1.108156</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>1.108033</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>0.000554</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>1.108402</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>0.000111</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>0.000222</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.109139</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>-0.000222</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>0.000111</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12247 rows × 6 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 5
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:41:55.817354Z",
     "start_time": "2024-08-24T06:41:55.543776Z"
    }
   },
   "cell_type": "code",
   "source": [
    "snp = yf.download('^SPX', start, end, interval=interval)\n",
    "snp.index = snp.index.tz_convert('utc')\n",
    "snp.index = snp.index.ceil('h')\n",
    "snp['snp_volatility'] = (snp['High'] -snp['Low']) / (snp['Open']+snp['Close']+snp['Low']+snp['High'])  *  4 ## not ideal because previous day might close lower or higher\n",
    "snp['snp_returns'] = np.log(snp['Close'].div(snp['Close'].shift(1)))\n",
    "snp.rename(columns={\"Close\": \"snp_price\", 'Volume': 'snp_volume' }, inplace=True)\n",
    "snp.drop(columns=['Adj Close', 'Open', 'High', 'Low'], inplace=True)\n",
    "snp"
   ],
   "id": "2c0faafed42bce0c",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[*********************100%%**********************]  1 of 1 completed\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "                             snp_price  snp_volume  snp_volatility  \\\n",
       "Datetime                                                             \n",
       "2022-08-29 14:00:00+00:00  4040.330078           0        0.006902   \n",
       "2022-08-29 15:00:00+00:00  4027.120117   238991365        0.006495   \n",
       "2022-08-29 16:00:00+00:00  4043.340088   183154616        0.006174   \n",
       "2022-08-29 17:00:00+00:00  4049.459961   169493518        0.005165   \n",
       "2022-08-29 18:00:00+00:00  4045.560059   163613948        0.004866   \n",
       "...                                ...         ...             ...   \n",
       "2024-08-19 16:00:00+00:00  5579.950195   191317840        0.001935   \n",
       "2024-08-19 17:00:00+00:00  5583.229980   160907370        0.001672   \n",
       "2024-08-19 18:00:00+00:00  5588.919922   173170955        0.001423   \n",
       "2024-08-19 19:00:00+00:00  5594.100098   215252315        0.001372   \n",
       "2024-08-19 20:00:00+00:00  5607.520020   242772673        0.002894   \n",
       "\n",
       "                           snp_returns  \n",
       "Datetime                                \n",
       "2022-08-29 14:00:00+00:00          NaN  \n",
       "2022-08-29 15:00:00+00:00    -0.003275  \n",
       "2022-08-29 16:00:00+00:00     0.004020  \n",
       "2022-08-29 17:00:00+00:00     0.001512  \n",
       "2022-08-29 18:00:00+00:00    -0.000964  \n",
       "...                                ...  \n",
       "2024-08-19 16:00:00+00:00     0.000610  \n",
       "2024-08-19 17:00:00+00:00     0.000588  \n",
       "2024-08-19 18:00:00+00:00     0.001019  \n",
       "2024-08-19 19:00:00+00:00     0.000926  \n",
       "2024-08-19 20:00:00+00:00     0.002396  \n",
       "\n",
       "[3456 rows x 4 columns]"
      ],
      "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>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-08-29 14:00:00+00:00</th>\n",
       "      <td>4040.330078</td>\n",
       "      <td>0</td>\n",
       "      <td>0.006902</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 15:00:00+00:00</th>\n",
       "      <td>4027.120117</td>\n",
       "      <td>238991365</td>\n",
       "      <td>0.006495</td>\n",
       "      <td>-0.003275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 16:00:00+00:00</th>\n",
       "      <td>4043.340088</td>\n",
       "      <td>183154616</td>\n",
       "      <td>0.006174</td>\n",
       "      <td>0.004020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 17:00:00+00:00</th>\n",
       "      <td>4049.459961</td>\n",
       "      <td>169493518</td>\n",
       "      <td>0.005165</td>\n",
       "      <td>0.001512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 18:00:00+00:00</th>\n",
       "      <td>4045.560059</td>\n",
       "      <td>163613948</td>\n",
       "      <td>0.004866</td>\n",
       "      <td>-0.000964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 16:00:00+00:00</th>\n",
       "      <td>5579.950195</td>\n",
       "      <td>191317840</td>\n",
       "      <td>0.001935</td>\n",
       "      <td>0.000610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 17:00:00+00:00</th>\n",
       "      <td>5583.229980</td>\n",
       "      <td>160907370</td>\n",
       "      <td>0.001672</td>\n",
       "      <td>0.000588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>5588.919922</td>\n",
       "      <td>173170955</td>\n",
       "      <td>0.001423</td>\n",
       "      <td>0.001019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>5594.100098</td>\n",
       "      <td>215252315</td>\n",
       "      <td>0.001372</td>\n",
       "      <td>0.000926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3456 rows × 4 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 6
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:41:57.973338Z",
     "start_time": "2024-08-24T06:41:57.955204Z"
    }
   },
   "cell_type": "code",
   "source": [
    "forex_combined = snp.join(df, how='outer')\n",
    "forex_combined"
   ],
   "id": "6f35d89cdf0e7331",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                             snp_price   snp_volume  snp_volatility  \\\n",
       "Datetime                                                              \n",
       "2022-08-28 23:00:00+00:00          NaN          NaN             NaN   \n",
       "2022-08-29 00:00:00+00:00          NaN          NaN             NaN   \n",
       "2022-08-29 01:00:00+00:00          NaN          NaN             NaN   \n",
       "2022-08-29 02:00:00+00:00          NaN          NaN             NaN   \n",
       "2022-08-29 03:00:00+00:00          NaN          NaN             NaN   \n",
       "...                                ...          ...             ...   \n",
       "2024-08-19 18:00:00+00:00  5588.919922  173170955.0        0.001423   \n",
       "2024-08-19 19:00:00+00:00  5594.100098  215252315.0        0.001372   \n",
       "2024-08-19 20:00:00+00:00  5607.520020  242772673.0        0.002894   \n",
       "2024-08-19 21:00:00+00:00          NaN          NaN             NaN   \n",
       "2024-08-19 22:00:00+00:00          NaN          NaN             NaN   \n",
       "\n",
       "                           snp_returns      Open      High       Low  \\\n",
       "Datetime                                                               \n",
       "2022-08-28 23:00:00+00:00          NaN  0.995520  0.995520  0.993838   \n",
       "2022-08-29 00:00:00+00:00          NaN  0.994233  0.995124  0.993246   \n",
       "2022-08-29 01:00:00+00:00          NaN  0.993345  0.993937  0.992162   \n",
       "2022-08-29 02:00:00+00:00          NaN  0.993147  0.993937  0.992753   \n",
       "2022-08-29 03:00:00+00:00          NaN  0.992950  0.993345  0.992457   \n",
       "...                                ...       ...       ...       ...   \n",
       "2024-08-19 18:00:00+00:00     0.001019  1.108156  1.108770  1.108033   \n",
       "2024-08-19 19:00:00+00:00     0.000926  1.108402  1.108893  1.108279   \n",
       "2024-08-19 20:00:00+00:00     0.002396  1.108647  1.108893  1.108525   \n",
       "2024-08-19 21:00:00+00:00          NaN  1.108525  1.109139  1.108279   \n",
       "2024-08-19 22:00:00+00:00          NaN  1.109016  1.109016  1.108647   \n",
       "\n",
       "                              Close   returns  day_of_week  \n",
       "Datetime                                                    \n",
       "2022-08-28 23:00:00+00:00  0.993838       NaN            6  \n",
       "2022-08-29 00:00:00+00:00  0.993345 -0.000497            0  \n",
       "2022-08-29 01:00:00+00:00  0.993443  0.000099            0  \n",
       "2022-08-29 02:00:00+00:00  0.993049 -0.000397            0  \n",
       "2022-08-29 03:00:00+00:00  0.992654 -0.000397            0  \n",
       "...                             ...       ...          ...  \n",
       "2024-08-19 18:00:00+00:00  1.108525  0.000554            0  \n",
       "2024-08-19 19:00:00+00:00  1.108647  0.000111            0  \n",
       "2024-08-19 20:00:00+00:00  1.108893  0.000222            0  \n",
       "2024-08-19 21:00:00+00:00  1.108647 -0.000222            0  \n",
       "2024-08-19 22:00:00+00:00  1.108770  0.000111            0  \n",
       "\n",
       "[12247 rows x 10 columns]"
      ],
      "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>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>returns</th>\n",
       "      <th>day_of_week</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-28 23:00:00+00:00</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 00:00:00+00:00</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.994233</td>\n",
       "      <td>0.995124</td>\n",
       "      <td>0.993246</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>-0.000497</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 01:00:00+00:00</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992162</td>\n",
       "      <td>0.993443</td>\n",
       "      <td>0.000099</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 02:00:00+00:00</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.993147</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992753</td>\n",
       "      <td>0.993049</td>\n",
       "      <td>-0.000397</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 03:00:00+00:00</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.992950</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.992457</td>\n",
       "      <td>0.992654</td>\n",
       "      <td>-0.000397</td>\n",
       "      <td>0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>5588.919922</td>\n",
       "      <td>173170955.0</td>\n",
       "      <td>0.001423</td>\n",
       "      <td>0.001019</td>\n",
       "      <td>1.108156</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>1.108033</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>0.000554</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>5594.100098</td>\n",
       "      <td>215252315.0</td>\n",
       "      <td>0.001372</td>\n",
       "      <td>0.000926</td>\n",
       "      <td>1.108402</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>0.000111</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>0.000222</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.109139</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>-0.000222</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>0.000111</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12247 rows × 10 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 7
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:01.795065Z",
     "start_time": "2024-08-24T06:42:01.431473Z"
    }
   },
   "cell_type": "code",
   "source": [
    "gold_ticker = 'GC=F'\n",
    "# Fetch historical data\n",
    "gold_data = yf.download(gold_ticker, start=start, end=end,  interval=interval)\n",
    "gold_data.index = gold_data.index.tz_convert('utc')\n",
    "gold_data.rename(columns={\"Close\": \"gold_price\", 'Volume': 'gold_volume' }, inplace=True)\n",
    "gold_data.drop(columns=['Adj Close', 'Open', 'High', 'Low'], inplace=True)\n",
    "gold_data"
   ],
   "id": "7dcfa16dc002077a",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[*********************100%%**********************]  1 of 1 completed\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume\n",
       "Datetime                                           \n",
       "2022-08-28 22:00:00+00:00  1747.099976         2303\n",
       "2022-08-28 23:00:00+00:00  1746.199951         3175\n",
       "2022-08-29 00:00:00+00:00  1745.800049         3193\n",
       "2022-08-29 01:00:00+00:00  1735.500000         9299\n",
       "2022-08-29 02:00:00+00:00  1737.900024         4304\n",
       "...                                ...          ...\n",
       "2024-08-19 23:00:00+00:00  2541.800049         1335\n",
       "2024-08-20 00:00:00+00:00  2543.000000         2746\n",
       "2024-08-20 01:00:00+00:00  2540.000000         5307\n",
       "2024-08-20 02:00:00+00:00  2537.399902         4353\n",
       "2024-08-20 03:00:00+00:00  2539.399902         2002\n",
       "\n",
       "[11348 rows x 2 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-08-28 22:00:00+00:00</th>\n",
       "      <td>1747.099976</td>\n",
       "      <td>2303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-28 23:00:00+00:00</th>\n",
       "      <td>1746.199951</td>\n",
       "      <td>3175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 00:00:00+00:00</th>\n",
       "      <td>1745.800049</td>\n",
       "      <td>3193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 01:00:00+00:00</th>\n",
       "      <td>1735.500000</td>\n",
       "      <td>9299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 02:00:00+00:00</th>\n",
       "      <td>1737.900024</td>\n",
       "      <td>4304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 23:00:00+00:00</th>\n",
       "      <td>2541.800049</td>\n",
       "      <td>1335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 00:00:00+00:00</th>\n",
       "      <td>2543.000000</td>\n",
       "      <td>2746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 01:00:00+00:00</th>\n",
       "      <td>2540.000000</td>\n",
       "      <td>5307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 02:00:00+00:00</th>\n",
       "      <td>2537.399902</td>\n",
       "      <td>4353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 03:00:00+00:00</th>\n",
       "      <td>2539.399902</td>\n",
       "      <td>2002</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11348 rows × 2 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 8
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:04.355522Z",
     "start_time": "2024-08-24T06:42:04.341501Z"
    }
   },
   "cell_type": "code",
   "source": "forex_combined = gold_data.join(forex_combined, how='outer')",
   "id": "5711a33f86fbb710",
   "outputs": [],
   "execution_count": 9
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:06.400572Z",
     "start_time": "2024-08-24T06:42:06.381657Z"
    }
   },
   "cell_type": "code",
   "source": "forex_combined",
   "id": "7baea7bcbed038d8",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume  snp_price  snp_volume  \\\n",
       "Datetime                                                                     \n",
       "2022-08-28 22:00:00+00:00  1747.099976       2303.0        NaN         NaN   \n",
       "2022-08-28 23:00:00+00:00  1746.199951       3175.0        NaN         NaN   \n",
       "2022-08-29 00:00:00+00:00  1745.800049       3193.0        NaN         NaN   \n",
       "2022-08-29 01:00:00+00:00  1735.500000       9299.0        NaN         NaN   \n",
       "2022-08-29 02:00:00+00:00  1737.900024       4304.0        NaN         NaN   \n",
       "...                                ...          ...        ...         ...   \n",
       "2024-08-19 23:00:00+00:00  2541.800049       1335.0        NaN         NaN   \n",
       "2024-08-20 00:00:00+00:00  2543.000000       2746.0        NaN         NaN   \n",
       "2024-08-20 01:00:00+00:00  2540.000000       5307.0        NaN         NaN   \n",
       "2024-08-20 02:00:00+00:00  2537.399902       4353.0        NaN         NaN   \n",
       "2024-08-20 03:00:00+00:00  2539.399902       2002.0        NaN         NaN   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2022-08-28 22:00:00+00:00             NaN          NaN       NaN       NaN   \n",
       "2022-08-28 23:00:00+00:00             NaN          NaN  0.995520  0.995520   \n",
       "2022-08-29 00:00:00+00:00             NaN          NaN  0.994233  0.995124   \n",
       "2022-08-29 01:00:00+00:00             NaN          NaN  0.993345  0.993937   \n",
       "2022-08-29 02:00:00+00:00             NaN          NaN  0.993147  0.993937   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-08-19 23:00:00+00:00             NaN          NaN       NaN       NaN   \n",
       "2024-08-20 00:00:00+00:00             NaN          NaN       NaN       NaN   \n",
       "2024-08-20 01:00:00+00:00             NaN          NaN       NaN       NaN   \n",
       "2024-08-20 02:00:00+00:00             NaN          NaN       NaN       NaN   \n",
       "2024-08-20 03:00:00+00:00             NaN          NaN       NaN       NaN   \n",
       "\n",
       "                                Low     Close   returns  day_of_week  \n",
       "Datetime                                                              \n",
       "2022-08-28 22:00:00+00:00       NaN       NaN       NaN          NaN  \n",
       "2022-08-28 23:00:00+00:00  0.993838  0.993838       NaN          6.0  \n",
       "2022-08-29 00:00:00+00:00  0.993246  0.993345 -0.000497          0.0  \n",
       "2022-08-29 01:00:00+00:00  0.992162  0.993443  0.000099          0.0  \n",
       "2022-08-29 02:00:00+00:00  0.992753  0.993049 -0.000397          0.0  \n",
       "...                             ...       ...       ...          ...  \n",
       "2024-08-19 23:00:00+00:00       NaN       NaN       NaN          NaN  \n",
       "2024-08-20 00:00:00+00:00       NaN       NaN       NaN          NaN  \n",
       "2024-08-20 01:00:00+00:00       NaN       NaN       NaN          NaN  \n",
       "2024-08-20 02:00:00+00:00       NaN       NaN       NaN          NaN  \n",
       "2024-08-20 03:00:00+00:00       NaN       NaN       NaN          NaN  \n",
       "\n",
       "[12375 rows x 12 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>returns</th>\n",
       "      <th>day_of_week</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-28 22:00:00+00:00</th>\n",
       "      <td>1747.099976</td>\n",
       "      <td>2303.0</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",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-28 23:00:00+00:00</th>\n",
       "      <td>1746.199951</td>\n",
       "      <td>3175.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 00:00:00+00:00</th>\n",
       "      <td>1745.800049</td>\n",
       "      <td>3193.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.994233</td>\n",
       "      <td>0.995124</td>\n",
       "      <td>0.993246</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>-0.000497</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 01:00:00+00:00</th>\n",
       "      <td>1735.500000</td>\n",
       "      <td>9299.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992162</td>\n",
       "      <td>0.993443</td>\n",
       "      <td>0.000099</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 02:00:00+00:00</th>\n",
       "      <td>1737.900024</td>\n",
       "      <td>4304.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.993147</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992753</td>\n",
       "      <td>0.993049</td>\n",
       "      <td>-0.000397</td>\n",
       "      <td>0.0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 23:00:00+00:00</th>\n",
       "      <td>2541.800049</td>\n",
       "      <td>1335.0</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",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 00:00:00+00:00</th>\n",
       "      <td>2543.000000</td>\n",
       "      <td>2746.0</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",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 01:00:00+00:00</th>\n",
       "      <td>2540.000000</td>\n",
       "      <td>5307.0</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",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 02:00:00+00:00</th>\n",
       "      <td>2537.399902</td>\n",
       "      <td>4353.0</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",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 03:00:00+00:00</th>\n",
       "      <td>2539.399902</td>\n",
       "      <td>2002.0</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",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12375 rows × 12 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 10
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:09.803419Z",
     "start_time": "2024-08-24T06:42:09.786172Z"
    }
   },
   "cell_type": "code",
   "source": [
    "ffill_columns = ['gold_price', 'gold_volume', 'snp_returns', 'snp_volume', 'snp_price', 'snp_volatility']\n",
    "forex_combined[ffill_columns] = forex_combined[ffill_columns].ffill()\n",
    "forex_combined['dir'] = np.where(forex_combined['returns'] > 0, 1,0)\n",
    "forex_combined"
   ],
   "id": "f01a1f1eda1c065e",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume   snp_price   snp_volume  \\\n",
       "Datetime                                                                       \n",
       "2022-08-28 22:00:00+00:00  1747.099976       2303.0         NaN          NaN   \n",
       "2022-08-28 23:00:00+00:00  1746.199951       3175.0         NaN          NaN   \n",
       "2022-08-29 00:00:00+00:00  1745.800049       3193.0         NaN          NaN   \n",
       "2022-08-29 01:00:00+00:00  1735.500000       9299.0         NaN          NaN   \n",
       "2022-08-29 02:00:00+00:00  1737.900024       4304.0         NaN          NaN   \n",
       "...                                ...          ...         ...          ...   \n",
       "2024-08-19 23:00:00+00:00  2541.800049       1335.0  5607.52002  242772673.0   \n",
       "2024-08-20 00:00:00+00:00  2543.000000       2746.0  5607.52002  242772673.0   \n",
       "2024-08-20 01:00:00+00:00  2540.000000       5307.0  5607.52002  242772673.0   \n",
       "2024-08-20 02:00:00+00:00  2537.399902       4353.0  5607.52002  242772673.0   \n",
       "2024-08-20 03:00:00+00:00  2539.399902       2002.0  5607.52002  242772673.0   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2022-08-28 22:00:00+00:00             NaN          NaN       NaN       NaN   \n",
       "2022-08-28 23:00:00+00:00             NaN          NaN  0.995520  0.995520   \n",
       "2022-08-29 00:00:00+00:00             NaN          NaN  0.994233  0.995124   \n",
       "2022-08-29 01:00:00+00:00             NaN          NaN  0.993345  0.993937   \n",
       "2022-08-29 02:00:00+00:00             NaN          NaN  0.993147  0.993937   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-08-19 23:00:00+00:00        0.002894     0.002396       NaN       NaN   \n",
       "2024-08-20 00:00:00+00:00        0.002894     0.002396       NaN       NaN   \n",
       "2024-08-20 01:00:00+00:00        0.002894     0.002396       NaN       NaN   \n",
       "2024-08-20 02:00:00+00:00        0.002894     0.002396       NaN       NaN   \n",
       "2024-08-20 03:00:00+00:00        0.002894     0.002396       NaN       NaN   \n",
       "\n",
       "                                Low     Close   returns  day_of_week  dir  \n",
       "Datetime                                                                   \n",
       "2022-08-28 22:00:00+00:00       NaN       NaN       NaN          NaN    0  \n",
       "2022-08-28 23:00:00+00:00  0.993838  0.993838       NaN          6.0    0  \n",
       "2022-08-29 00:00:00+00:00  0.993246  0.993345 -0.000497          0.0    0  \n",
       "2022-08-29 01:00:00+00:00  0.992162  0.993443  0.000099          0.0    1  \n",
       "2022-08-29 02:00:00+00:00  0.992753  0.993049 -0.000397          0.0    0  \n",
       "...                             ...       ...       ...          ...  ...  \n",
       "2024-08-19 23:00:00+00:00       NaN       NaN       NaN          NaN    0  \n",
       "2024-08-20 00:00:00+00:00       NaN       NaN       NaN          NaN    0  \n",
       "2024-08-20 01:00:00+00:00       NaN       NaN       NaN          NaN    0  \n",
       "2024-08-20 02:00:00+00:00       NaN       NaN       NaN          NaN    0  \n",
       "2024-08-20 03:00:00+00:00       NaN       NaN       NaN          NaN    0  \n",
       "\n",
       "[12375 rows x 13 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>returns</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>dir</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-28 22:00:00+00:00</th>\n",
       "      <td>1747.099976</td>\n",
       "      <td>2303.0</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",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-28 23:00:00+00:00</th>\n",
       "      <td>1746.199951</td>\n",
       "      <td>3175.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.995520</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>0.993838</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 00:00:00+00:00</th>\n",
       "      <td>1745.800049</td>\n",
       "      <td>3193.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.994233</td>\n",
       "      <td>0.995124</td>\n",
       "      <td>0.993246</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>-0.000497</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 01:00:00+00:00</th>\n",
       "      <td>1735.500000</td>\n",
       "      <td>9299.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.993345</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992162</td>\n",
       "      <td>0.993443</td>\n",
       "      <td>0.000099</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 02:00:00+00:00</th>\n",
       "      <td>1737.900024</td>\n",
       "      <td>4304.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.993147</td>\n",
       "      <td>0.993937</td>\n",
       "      <td>0.992753</td>\n",
       "      <td>0.993049</td>\n",
       "      <td>-0.000397</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 23:00:00+00:00</th>\n",
       "      <td>2541.800049</td>\n",
       "      <td>1335.0</td>\n",
       "      <td>5607.52002</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</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>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 00:00:00+00:00</th>\n",
       "      <td>2543.000000</td>\n",
       "      <td>2746.0</td>\n",
       "      <td>5607.52002</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</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>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 01:00:00+00:00</th>\n",
       "      <td>2540.000000</td>\n",
       "      <td>5307.0</td>\n",
       "      <td>5607.52002</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</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>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 02:00:00+00:00</th>\n",
       "      <td>2537.399902</td>\n",
       "      <td>4353.0</td>\n",
       "      <td>5607.52002</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</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>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-20 03:00:00+00:00</th>\n",
       "      <td>2539.399902</td>\n",
       "      <td>2002.0</td>\n",
       "      <td>5607.52002</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</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>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12375 rows × 13 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 11
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:12.644741Z",
     "start_time": "2024-08-24T06:42:12.621753Z"
    }
   },
   "cell_type": "code",
   "source": [
    "forex_combined.dropna(inplace=True)\n",
    "forex_combined"
   ],
   "id": "737dafbb595735dd",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume    snp_price   snp_volume  \\\n",
       "Datetime                                                                        \n",
       "2022-08-29 15:00:00+00:00  1753.099976       6503.0  4027.120117  238991365.0   \n",
       "2022-08-29 16:00:00+00:00  1750.599976       5476.0  4043.340088  183154616.0   \n",
       "2022-08-29 17:00:00+00:00  1750.699951       6480.0  4049.459961  169493518.0   \n",
       "2022-08-29 18:00:00+00:00  1749.500000       2913.0  4045.560059  163613948.0   \n",
       "2022-08-29 19:00:00+00:00  1750.500000       3471.0  4046.800049  190743137.0   \n",
       "...                                ...          ...          ...          ...   \n",
       "2024-08-19 18:00:00+00:00  2542.800049       5353.0  5588.919922  173170955.0   \n",
       "2024-08-19 19:00:00+00:00  2543.899902       3701.0  5594.100098  215252315.0   \n",
       "2024-08-19 20:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 21:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 22:00:00+00:00  2542.500000          0.0  5607.520020  242772673.0   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2022-08-29 15:00:00+00:00        0.006495    -0.003275  0.998702  1.000200   \n",
       "2022-08-29 16:00:00+00:00        0.006174     0.004020  0.999900  1.001803   \n",
       "2022-08-29 17:00:00+00:00        0.005165     0.001512  1.002205  1.002205   \n",
       "2022-08-29 18:00:00+00:00        0.004866    -0.000964  1.001001  1.001302   \n",
       "2022-08-29 19:00:00+00:00        0.004446     0.000306  0.999500  1.000300   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-08-19 18:00:00+00:00        0.001423     0.001019  1.108156  1.108770   \n",
       "2024-08-19 19:00:00+00:00        0.001372     0.000926  1.108402  1.108893   \n",
       "2024-08-19 20:00:00+00:00        0.002894     0.002396  1.108647  1.108893   \n",
       "2024-08-19 21:00:00+00:00        0.002894     0.002396  1.108525  1.109139   \n",
       "2024-08-19 22:00:00+00:00        0.002894     0.002396  1.109016  1.109016   \n",
       "\n",
       "                                Low     Close   returns  day_of_week  dir  \n",
       "Datetime                                                                   \n",
       "2022-08-29 15:00:00+00:00  0.998403  0.999900  0.001299          0.0    1  \n",
       "2022-08-29 16:00:00+00:00  0.999900  1.001703  0.001801          0.0    1  \n",
       "2022-08-29 17:00:00+00:00  1.000901  1.001001 -0.000701          0.0    0  \n",
       "2022-08-29 18:00:00+00:00  0.999400  0.999400 -0.001600          0.0    0  \n",
       "2022-08-29 19:00:00+00:00  0.999400  1.000000  0.000600          0.0    1  \n",
       "...                             ...       ...       ...          ...  ...  \n",
       "2024-08-19 18:00:00+00:00  1.108033  1.108525  0.000554          0.0    1  \n",
       "2024-08-19 19:00:00+00:00  1.108279  1.108647  0.000111          0.0    1  \n",
       "2024-08-19 20:00:00+00:00  1.108525  1.108893  0.000222          0.0    1  \n",
       "2024-08-19 21:00:00+00:00  1.108279  1.108647 -0.000222          0.0    0  \n",
       "2024-08-19 22:00:00+00:00  1.108647  1.108770  0.000111          0.0    1  \n",
       "\n",
       "[12231 rows x 13 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>returns</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>dir</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-29 15:00:00+00:00</th>\n",
       "      <td>1753.099976</td>\n",
       "      <td>6503.0</td>\n",
       "      <td>4027.120117</td>\n",
       "      <td>238991365.0</td>\n",
       "      <td>0.006495</td>\n",
       "      <td>-0.003275</td>\n",
       "      <td>0.998702</td>\n",
       "      <td>1.000200</td>\n",
       "      <td>0.998403</td>\n",
       "      <td>0.999900</td>\n",
       "      <td>0.001299</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 16:00:00+00:00</th>\n",
       "      <td>1750.599976</td>\n",
       "      <td>5476.0</td>\n",
       "      <td>4043.340088</td>\n",
       "      <td>183154616.0</td>\n",
       "      <td>0.006174</td>\n",
       "      <td>0.004020</td>\n",
       "      <td>0.999900</td>\n",
       "      <td>1.001803</td>\n",
       "      <td>0.999900</td>\n",
       "      <td>1.001703</td>\n",
       "      <td>0.001801</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 17:00:00+00:00</th>\n",
       "      <td>1750.699951</td>\n",
       "      <td>6480.0</td>\n",
       "      <td>4049.459961</td>\n",
       "      <td>169493518.0</td>\n",
       "      <td>0.005165</td>\n",
       "      <td>0.001512</td>\n",
       "      <td>1.002205</td>\n",
       "      <td>1.002205</td>\n",
       "      <td>1.000901</td>\n",
       "      <td>1.001001</td>\n",
       "      <td>-0.000701</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 18:00:00+00:00</th>\n",
       "      <td>1749.500000</td>\n",
       "      <td>2913.0</td>\n",
       "      <td>4045.560059</td>\n",
       "      <td>163613948.0</td>\n",
       "      <td>0.004866</td>\n",
       "      <td>-0.000964</td>\n",
       "      <td>1.001001</td>\n",
       "      <td>1.001302</td>\n",
       "      <td>0.999400</td>\n",
       "      <td>0.999400</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-29 19:00:00+00:00</th>\n",
       "      <td>1750.500000</td>\n",
       "      <td>3471.0</td>\n",
       "      <td>4046.800049</td>\n",
       "      <td>190743137.0</td>\n",
       "      <td>0.004446</td>\n",
       "      <td>0.000306</td>\n",
       "      <td>0.999500</td>\n",
       "      <td>1.000300</td>\n",
       "      <td>0.999400</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>2542.800049</td>\n",
       "      <td>5353.0</td>\n",
       "      <td>5588.919922</td>\n",
       "      <td>173170955.0</td>\n",
       "      <td>0.001423</td>\n",
       "      <td>0.001019</td>\n",
       "      <td>1.108156</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>1.108033</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>0.000554</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>2543.899902</td>\n",
       "      <td>3701.0</td>\n",
       "      <td>5594.100098</td>\n",
       "      <td>215252315.0</td>\n",
       "      <td>0.001372</td>\n",
       "      <td>0.000926</td>\n",
       "      <td>1.108402</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>0.000111</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>0.000222</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.109139</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>-0.000222</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>2542.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>0.000111</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12231 rows × 13 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 12
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:14.830333Z",
     "start_time": "2024-08-24T06:42:14.821696Z"
    }
   },
   "cell_type": "code",
   "source": [
    "extra_features = ['Open', 'High', 'Low', 'Close', 'returns']\n",
    "lags = 24\n",
    "cols = []\n",
    "features = ffill_columns + extra_features\n",
    "features"
   ],
   "id": "6ddb2bf37f4ef15a",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['gold_price',\n",
       " 'gold_volume',\n",
       " 'snp_returns',\n",
       " 'snp_volume',\n",
       " 'snp_price',\n",
       " 'snp_volatility',\n",
       " 'Open',\n",
       " 'High',\n",
       " 'Low',\n",
       " 'Close',\n",
       " 'returns']"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 13
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:17.467714Z",
     "start_time": "2024-08-24T06:42:17.341695Z"
    }
   },
   "cell_type": "code",
   "source": [
    "for f in features:\n",
    "    for lag in range(1, lags + 1):\n",
    "        col = \"{}_lag_{}\".format(f, lag)\n",
    "        forex_combined[col] = forex_combined[f].shift(lag)\n",
    "        cols.append(col)\n",
    "forex_combined.dropna(inplace=True)\n",
    "forex_combined"
   ],
   "id": "855002f24a5f4b86",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n",
      "C:\\Users\\grigo\\AppData\\Local\\Temp\\ipykernel_6176\\1825698716.py:4: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  forex_combined[col] = forex_combined[f].shift(lag)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume    snp_price   snp_volume  \\\n",
       "Datetime                                                                        \n",
       "2022-08-30 15:00:00+00:00  1737.800049      10693.0  3991.939941  303972709.0   \n",
       "2022-08-30 16:00:00+00:00  1735.800049       5181.0  3991.959961  197123531.0   \n",
       "2022-08-30 17:00:00+00:00  1735.800049       6423.0  3975.770020  186525504.0   \n",
       "2022-08-30 18:00:00+00:00  1735.699951       2296.0  3975.189941  193156989.0   \n",
       "2022-08-30 19:00:00+00:00  1734.900024       3164.0  3989.540039  224540573.0   \n",
       "...                                ...          ...          ...          ...   \n",
       "2024-08-19 18:00:00+00:00  2542.800049       5353.0  5588.919922  173170955.0   \n",
       "2024-08-19 19:00:00+00:00  2543.899902       3701.0  5594.100098  215252315.0   \n",
       "2024-08-19 20:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 21:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 22:00:00+00:00  2542.500000          0.0  5607.520020  242772673.0   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2022-08-30 15:00:00+00:00        0.009543    -0.003466  0.999400  1.003512   \n",
       "2022-08-30 16:00:00+00:00        0.004345     0.000005  1.002305  1.003512   \n",
       "2022-08-30 17:00:00+00:00        0.005411    -0.004064  1.002908  1.002908   \n",
       "2022-08-30 18:00:00+00:00        0.004888    -0.000146  1.002004  1.003210   \n",
       "2022-08-30 19:00:00+00:00        0.006125     0.003603  1.002406  1.003512   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-08-19 18:00:00+00:00        0.001423     0.001019  1.108156  1.108770   \n",
       "2024-08-19 19:00:00+00:00        0.001372     0.000926  1.108402  1.108893   \n",
       "2024-08-19 20:00:00+00:00        0.002894     0.002396  1.108647  1.108893   \n",
       "2024-08-19 21:00:00+00:00        0.002894     0.002396  1.108525  1.109139   \n",
       "2024-08-19 22:00:00+00:00        0.002894     0.002396  1.109016  1.109016   \n",
       "\n",
       "                                Low     Close  ...  returns_lag_15  \\\n",
       "Datetime                                       ...                   \n",
       "2022-08-30 15:00:00+00:00  0.999400  1.002104  ...       -0.000400   \n",
       "2022-08-30 16:00:00+00:00  1.002004  1.002707  ...       -0.001600   \n",
       "2022-08-30 17:00:00+00:00  1.001402  1.002004  ...        0.000000   \n",
       "2022-08-30 18:00:00+00:00  1.001904  1.002406  ...        0.000800   \n",
       "2022-08-30 19:00:00+00:00  1.001803  1.002004  ...        0.000100   \n",
       "...                             ...       ...  ...             ...   \n",
       "2024-08-19 18:00:00+00:00  1.108033  1.108525  ...        0.000773   \n",
       "2024-08-19 19:00:00+00:00  1.108279  1.108647  ...        0.000000   \n",
       "2024-08-19 20:00:00+00:00  1.108525  1.108893  ...        0.000773   \n",
       "2024-08-19 21:00:00+00:00  1.108279  1.108647  ...       -0.000552   \n",
       "2024-08-19 22:00:00+00:00  1.108647  1.108770  ...        0.000110   \n",
       "\n",
       "                           returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "Datetime                                                                    \n",
       "2022-08-30 15:00:00+00:00        0.001001        0.000000        0.000100   \n",
       "2022-08-30 16:00:00+00:00       -0.000400        0.001001        0.000000   \n",
       "2022-08-30 17:00:00+00:00       -0.001600       -0.000400        0.001001   \n",
       "2022-08-30 18:00:00+00:00        0.000000       -0.001600       -0.000400   \n",
       "2022-08-30 19:00:00+00:00        0.000800        0.000000       -0.001600   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.000110        0.000000        0.000662   \n",
       "2024-08-19 19:00:00+00:00        0.000773       -0.000110        0.000000   \n",
       "2024-08-19 20:00:00+00:00        0.000000        0.000773       -0.000110   \n",
       "2024-08-19 21:00:00+00:00        0.000773        0.000000        0.000773   \n",
       "2024-08-19 22:00:00+00:00       -0.000552        0.000773        0.000000   \n",
       "\n",
       "                           returns_lag_19  returns_lag_20  returns_lag_21  \\\n",
       "Datetime                                                                    \n",
       "2022-08-30 15:00:00+00:00        0.000200        0.000600       -0.001600   \n",
       "2022-08-30 16:00:00+00:00        0.000100        0.000200        0.000600   \n",
       "2022-08-30 17:00:00+00:00        0.000000        0.000100        0.000200   \n",
       "2022-08-30 18:00:00+00:00        0.001001        0.000000        0.000100   \n",
       "2022-08-30 19:00:00+00:00       -0.000400        0.001001        0.000000   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.000331        0.000110        0.000221   \n",
       "2024-08-19 19:00:00+00:00        0.000662       -0.000331        0.000110   \n",
       "2024-08-19 20:00:00+00:00        0.000000        0.000662       -0.000331   \n",
       "2024-08-19 21:00:00+00:00       -0.000110        0.000000        0.000662   \n",
       "2024-08-19 22:00:00+00:00        0.000773       -0.000110        0.000000   \n",
       "\n",
       "                           returns_lag_22  returns_lag_23  returns_lag_24  \n",
       "Datetime                                                                   \n",
       "2022-08-30 15:00:00+00:00       -0.000701        0.001801        0.001299  \n",
       "2022-08-30 16:00:00+00:00       -0.001600       -0.000701        0.001801  \n",
       "2022-08-30 17:00:00+00:00        0.000600       -0.001600       -0.000701  \n",
       "2022-08-30 18:00:00+00:00        0.000200        0.000600       -0.001600  \n",
       "2022-08-30 19:00:00+00:00        0.000100        0.000200        0.000600  \n",
       "...                                   ...             ...             ...  \n",
       "2024-08-19 18:00:00+00:00        0.000331        0.000771        0.000771  \n",
       "2024-08-19 19:00:00+00:00        0.000221        0.000331        0.000771  \n",
       "2024-08-19 20:00:00+00:00        0.000110        0.000221        0.000331  \n",
       "2024-08-19 21:00:00+00:00       -0.000331        0.000110        0.000221  \n",
       "2024-08-19 22:00:00+00:00        0.000662       -0.000331        0.000110  \n",
       "\n",
       "[12207 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-30 15:00:00+00:00</th>\n",
       "      <td>1737.800049</td>\n",
       "      <td>10693.0</td>\n",
       "      <td>3991.939941</td>\n",
       "      <td>303972709.0</td>\n",
       "      <td>0.009543</td>\n",
       "      <td>-0.003466</td>\n",
       "      <td>0.999400</td>\n",
       "      <td>1.003512</td>\n",
       "      <td>0.999400</td>\n",
       "      <td>1.002104</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000701</td>\n",
       "      <td>0.001801</td>\n",
       "      <td>0.001299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 16:00:00+00:00</th>\n",
       "      <td>1735.800049</td>\n",
       "      <td>5181.0</td>\n",
       "      <td>3991.959961</td>\n",
       "      <td>197123531.0</td>\n",
       "      <td>0.004345</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>1.002305</td>\n",
       "      <td>1.003512</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>1.002707</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000701</td>\n",
       "      <td>0.001801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 17:00:00+00:00</th>\n",
       "      <td>1735.800049</td>\n",
       "      <td>6423.0</td>\n",
       "      <td>3975.770020</td>\n",
       "      <td>186525504.0</td>\n",
       "      <td>0.005411</td>\n",
       "      <td>-0.004064</td>\n",
       "      <td>1.002908</td>\n",
       "      <td>1.002908</td>\n",
       "      <td>1.001402</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 18:00:00+00:00</th>\n",
       "      <td>1735.699951</td>\n",
       "      <td>2296.0</td>\n",
       "      <td>3975.189941</td>\n",
       "      <td>193156989.0</td>\n",
       "      <td>0.004888</td>\n",
       "      <td>-0.000146</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>1.003210</td>\n",
       "      <td>1.001904</td>\n",
       "      <td>1.002406</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000800</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 19:00:00+00:00</th>\n",
       "      <td>1734.900024</td>\n",
       "      <td>3164.0</td>\n",
       "      <td>3989.540039</td>\n",
       "      <td>224540573.0</td>\n",
       "      <td>0.006125</td>\n",
       "      <td>0.003603</td>\n",
       "      <td>1.002406</td>\n",
       "      <td>1.003512</td>\n",
       "      <td>1.001803</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000800</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</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>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>2542.800049</td>\n",
       "      <td>5353.0</td>\n",
       "      <td>5588.919922</td>\n",
       "      <td>173170955.0</td>\n",
       "      <td>0.001423</td>\n",
       "      <td>0.001019</td>\n",
       "      <td>1.108156</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>1.108033</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "      <td>0.000771</td>\n",
       "      <td>0.000771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>2543.899902</td>\n",
       "      <td>3701.0</td>\n",
       "      <td>5594.100098</td>\n",
       "      <td>215252315.0</td>\n",
       "      <td>0.001372</td>\n",
       "      <td>0.000926</td>\n",
       "      <td>1.108402</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "      <td>0.000771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.109139</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000552</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>2542.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>-0.000552</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12207 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 14
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:21.370025Z",
     "start_time": "2024-08-24T06:42:21.323519Z"
    }
   },
   "cell_type": "code",
   "source": [
    "split = int(len(forex_combined)* 0.7)\n",
    "train = forex_combined.iloc[:split].copy()\n",
    "train"
   ],
   "id": "9d8124962b36dd5",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume    snp_price   snp_volume  \\\n",
       "Datetime                                                                        \n",
       "2022-08-30 15:00:00+00:00  1737.800049      10693.0  3991.939941  303972709.0   \n",
       "2022-08-30 16:00:00+00:00  1735.800049       5181.0  3991.959961  197123531.0   \n",
       "2022-08-30 17:00:00+00:00  1735.800049       6423.0  3975.770020  186525504.0   \n",
       "2022-08-30 18:00:00+00:00  1735.699951       2296.0  3975.189941  193156989.0   \n",
       "2022-08-30 19:00:00+00:00  1734.900024       3164.0  3989.540039  224540573.0   \n",
       "...                                ...          ...          ...          ...   \n",
       "2024-01-16 19:00:00+00:00  2032.800049       4989.0  4752.200195  215168361.0   \n",
       "2024-01-16 20:00:00+00:00  2031.800049       2850.0  4756.799805  242464900.0   \n",
       "2024-01-16 21:00:00+00:00  2031.699951          0.0  4765.180176  336315998.0   \n",
       "2024-01-16 22:00:00+00:00  2031.699951          0.0  4765.180176  336315998.0   \n",
       "2024-01-16 23:00:00+00:00  2031.300049       3405.0  4765.180176  336315998.0   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2022-08-30 15:00:00+00:00        0.009543    -0.003466  0.999400  1.003512   \n",
       "2022-08-30 16:00:00+00:00        0.004345     0.000005  1.002305  1.003512   \n",
       "2022-08-30 17:00:00+00:00        0.005411    -0.004064  1.002908  1.002908   \n",
       "2022-08-30 18:00:00+00:00        0.004888    -0.000146  1.002004  1.003210   \n",
       "2022-08-30 19:00:00+00:00        0.006125     0.003603  1.002406  1.003512   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-01-16 19:00:00+00:00        0.004323    -0.002912  1.087311  1.087784   \n",
       "2024-01-16 20:00:00+00:00        0.002288     0.000967  1.087666  1.087903   \n",
       "2024-01-16 21:00:00+00:00        0.002521     0.001760  1.087548  1.088021   \n",
       "2024-01-16 22:00:00+00:00        0.002521     0.001760  1.088021  1.088139   \n",
       "2024-01-16 23:00:00+00:00        0.002521     0.001760  1.087903  1.088139   \n",
       "\n",
       "                                Low     Close  ...  returns_lag_15  \\\n",
       "Datetime                                       ...                   \n",
       "2022-08-30 15:00:00+00:00  0.999400  1.002104  ...       -0.000400   \n",
       "2022-08-30 16:00:00+00:00  1.002004  1.002707  ...       -0.001600   \n",
       "2022-08-30 17:00:00+00:00  1.001402  1.002004  ...        0.000000   \n",
       "2022-08-30 18:00:00+00:00  1.001904  1.002406  ...        0.000800   \n",
       "2022-08-30 19:00:00+00:00  1.001803  1.002004  ...        0.000100   \n",
       "...                             ...       ...  ...             ...   \n",
       "2024-01-16 19:00:00+00:00  1.087075  1.087666  ...       -0.000109   \n",
       "2024-01-16 20:00:00+00:00  1.087311  1.087548  ...       -0.000218   \n",
       "2024-01-16 21:00:00+00:00  1.087429  1.087784  ...        0.000328   \n",
       "2024-01-16 22:00:00+00:00  1.087666  1.088021  ...       -0.000109   \n",
       "2024-01-16 23:00:00+00:00  1.087903  1.088139  ...       -0.001200   \n",
       "\n",
       "                           returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "Datetime                                                                    \n",
       "2022-08-30 15:00:00+00:00        0.001001        0.000000        0.000100   \n",
       "2022-08-30 16:00:00+00:00       -0.000400        0.001001        0.000000   \n",
       "2022-08-30 17:00:00+00:00       -0.001600       -0.000400        0.001001   \n",
       "2022-08-30 18:00:00+00:00        0.000000       -0.001600       -0.000400   \n",
       "2022-08-30 19:00:00+00:00        0.000800        0.000000       -0.001600   \n",
       "...                                   ...             ...             ...   \n",
       "2024-01-16 19:00:00+00:00       -0.000437       -0.000328        0.000656   \n",
       "2024-01-16 20:00:00+00:00       -0.000109       -0.000437       -0.000328   \n",
       "2024-01-16 21:00:00+00:00       -0.000218       -0.000109       -0.000437   \n",
       "2024-01-16 22:00:00+00:00        0.000328       -0.000218       -0.000109   \n",
       "2024-01-16 23:00:00+00:00       -0.000109        0.000328       -0.000218   \n",
       "\n",
       "                           returns_lag_19  returns_lag_20  returns_lag_21  \\\n",
       "Datetime                                                                    \n",
       "2022-08-30 15:00:00+00:00        0.000200        0.000600       -0.001600   \n",
       "2022-08-30 16:00:00+00:00        0.000100        0.000200        0.000600   \n",
       "2022-08-30 17:00:00+00:00        0.000000        0.000100        0.000200   \n",
       "2022-08-30 18:00:00+00:00        0.001001        0.000000        0.000100   \n",
       "2022-08-30 19:00:00+00:00       -0.000400        0.001001        0.000000   \n",
       "...                                   ...             ...             ...   \n",
       "2024-01-16 19:00:00+00:00       -0.002296       -0.000438       -0.000109   \n",
       "2024-01-16 20:00:00+00:00        0.000656       -0.002296       -0.000438   \n",
       "2024-01-16 21:00:00+00:00       -0.000328        0.000656       -0.002296   \n",
       "2024-01-16 22:00:00+00:00       -0.000437       -0.000328        0.000656   \n",
       "2024-01-16 23:00:00+00:00       -0.000109       -0.000437       -0.000328   \n",
       "\n",
       "                           returns_lag_22  returns_lag_23  returns_lag_24  \n",
       "Datetime                                                                   \n",
       "2022-08-30 15:00:00+00:00       -0.000701        0.001801        0.001299  \n",
       "2022-08-30 16:00:00+00:00       -0.001600       -0.000701        0.001801  \n",
       "2022-08-30 17:00:00+00:00        0.000600       -0.001600       -0.000701  \n",
       "2022-08-30 18:00:00+00:00        0.000200        0.000600       -0.001600  \n",
       "2022-08-30 19:00:00+00:00        0.000100        0.000200        0.000600  \n",
       "...                                   ...             ...             ...  \n",
       "2024-01-16 19:00:00+00:00       -0.000110        0.000110       -0.000110  \n",
       "2024-01-16 20:00:00+00:00       -0.000109       -0.000110        0.000110  \n",
       "2024-01-16 21:00:00+00:00       -0.000438       -0.000109       -0.000110  \n",
       "2024-01-16 22:00:00+00:00       -0.002296       -0.000438       -0.000109  \n",
       "2024-01-16 23:00:00+00:00        0.000656       -0.002296       -0.000438  \n",
       "\n",
       "[8544 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-30 15:00:00+00:00</th>\n",
       "      <td>1737.800049</td>\n",
       "      <td>10693.0</td>\n",
       "      <td>3991.939941</td>\n",
       "      <td>303972709.0</td>\n",
       "      <td>0.009543</td>\n",
       "      <td>-0.003466</td>\n",
       "      <td>0.999400</td>\n",
       "      <td>1.003512</td>\n",
       "      <td>0.999400</td>\n",
       "      <td>1.002104</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000701</td>\n",
       "      <td>0.001801</td>\n",
       "      <td>0.001299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 16:00:00+00:00</th>\n",
       "      <td>1735.800049</td>\n",
       "      <td>5181.0</td>\n",
       "      <td>3991.959961</td>\n",
       "      <td>197123531.0</td>\n",
       "      <td>0.004345</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>1.002305</td>\n",
       "      <td>1.003512</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>1.002707</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000701</td>\n",
       "      <td>0.001801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 17:00:00+00:00</th>\n",
       "      <td>1735.800049</td>\n",
       "      <td>6423.0</td>\n",
       "      <td>3975.770020</td>\n",
       "      <td>186525504.0</td>\n",
       "      <td>0.005411</td>\n",
       "      <td>-0.004064</td>\n",
       "      <td>1.002908</td>\n",
       "      <td>1.002908</td>\n",
       "      <td>1.001402</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 18:00:00+00:00</th>\n",
       "      <td>1735.699951</td>\n",
       "      <td>2296.0</td>\n",
       "      <td>3975.189941</td>\n",
       "      <td>193156989.0</td>\n",
       "      <td>0.004888</td>\n",
       "      <td>-0.000146</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>1.003210</td>\n",
       "      <td>1.001904</td>\n",
       "      <td>1.002406</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000800</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</td>\n",
       "      <td>-0.001600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 19:00:00+00:00</th>\n",
       "      <td>1734.900024</td>\n",
       "      <td>3164.0</td>\n",
       "      <td>3989.540039</td>\n",
       "      <td>224540573.0</td>\n",
       "      <td>0.006125</td>\n",
       "      <td>0.003603</td>\n",
       "      <td>1.002406</td>\n",
       "      <td>1.003512</td>\n",
       "      <td>1.001803</td>\n",
       "      <td>1.002004</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000800</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001600</td>\n",
       "      <td>-0.000400</td>\n",
       "      <td>0.001001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000600</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>2024-01-16 19:00:00+00:00</th>\n",
       "      <td>2032.800049</td>\n",
       "      <td>4989.0</td>\n",
       "      <td>4752.200195</td>\n",
       "      <td>215168361.0</td>\n",
       "      <td>0.004323</td>\n",
       "      <td>-0.002912</td>\n",
       "      <td>1.087311</td>\n",
       "      <td>1.087784</td>\n",
       "      <td>1.087075</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "      <td>-0.002296</td>\n",
       "      <td>-0.000438</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>-0.000110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 20:00:00+00:00</th>\n",
       "      <td>2031.800049</td>\n",
       "      <td>2850.0</td>\n",
       "      <td>4756.799805</td>\n",
       "      <td>242464900.0</td>\n",
       "      <td>0.002288</td>\n",
       "      <td>0.000967</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.087903</td>\n",
       "      <td>1.087311</td>\n",
       "      <td>1.087548</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "      <td>-0.002296</td>\n",
       "      <td>-0.000438</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 21:00:00+00:00</th>\n",
       "      <td>2031.699951</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.087548</td>\n",
       "      <td>1.088021</td>\n",
       "      <td>1.087429</td>\n",
       "      <td>1.087784</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "      <td>-0.002296</td>\n",
       "      <td>-0.000438</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 22:00:00+00:00</th>\n",
       "      <td>2031.699951</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.088021</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.088021</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "      <td>-0.002296</td>\n",
       "      <td>-0.000438</td>\n",
       "      <td>-0.000109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 23:00:00+00:00</th>\n",
       "      <td>2031.300049</td>\n",
       "      <td>3405.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.087903</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.087903</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "      <td>-0.002296</td>\n",
       "      <td>-0.000438</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8544 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 15
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:23.974825Z",
     "start_time": "2024-08-24T06:42:23.947479Z"
    }
   },
   "cell_type": "code",
   "source": [
    "test = forex_combined.iloc[split:].copy()\n",
    "test"
   ],
   "id": "16ab63bc350ed2e8",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume    snp_price   snp_volume  \\\n",
       "Datetime                                                                        \n",
       "2024-01-17 00:00:00+00:00  2031.599976       2532.0  4765.180176  336315998.0   \n",
       "2024-01-17 01:00:00+00:00  2033.400024       4852.0  4765.180176  336315998.0   \n",
       "2024-01-17 02:00:00+00:00  2032.000000       5865.0  4765.180176  336315998.0   \n",
       "2024-01-17 03:00:00+00:00  2027.199951       4409.0  4765.180176  336315998.0   \n",
       "2024-01-17 04:00:00+00:00  2028.500000       2729.0  4765.180176  336315998.0   \n",
       "...                                ...          ...          ...          ...   \n",
       "2024-08-19 18:00:00+00:00  2542.800049       5353.0  5588.919922  173170955.0   \n",
       "2024-08-19 19:00:00+00:00  2543.899902       3701.0  5594.100098  215252315.0   \n",
       "2024-08-19 20:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 21:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 22:00:00+00:00  2542.500000          0.0  5607.520020  242772673.0   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2024-01-17 00:00:00+00:00        0.002521     0.001760  1.088139  1.088139   \n",
       "2024-01-17 01:00:00+00:00        0.002521     0.001760  1.088139  1.088376   \n",
       "2024-01-17 02:00:00+00:00        0.002521     0.001760  1.088021  1.088732   \n",
       "2024-01-17 03:00:00+00:00        0.002521     0.001760  1.087666  1.087784   \n",
       "2024-01-17 04:00:00+00:00        0.002521     0.001760  1.086838  1.087429   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-08-19 18:00:00+00:00        0.001423     0.001019  1.108156  1.108770   \n",
       "2024-08-19 19:00:00+00:00        0.001372     0.000926  1.108402  1.108893   \n",
       "2024-08-19 20:00:00+00:00        0.002894     0.002396  1.108647  1.108893   \n",
       "2024-08-19 21:00:00+00:00        0.002894     0.002396  1.108525  1.109139   \n",
       "2024-08-19 22:00:00+00:00        0.002894     0.002396  1.109016  1.109016   \n",
       "\n",
       "                                Low     Close  ...  returns_lag_15  \\\n",
       "Datetime                                       ...                   \n",
       "2024-01-17 00:00:00+00:00  1.087548  1.088139  ...       -0.001635   \n",
       "2024-01-17 01:00:00+00:00  1.087666  1.088021  ...       -0.000544   \n",
       "2024-01-17 02:00:00+00:00  1.087666  1.087784  ...        0.000980   \n",
       "2024-01-17 03:00:00+00:00  1.086838  1.086957  ...       -0.000545   \n",
       "2024-01-17 04:00:00+00:00  1.086838  1.087311  ...       -0.001197   \n",
       "...                             ...       ...  ...             ...   \n",
       "2024-08-19 18:00:00+00:00  1.108033  1.108525  ...        0.000773   \n",
       "2024-08-19 19:00:00+00:00  1.108279  1.108647  ...        0.000000   \n",
       "2024-08-19 20:00:00+00:00  1.108525  1.108893  ...        0.000773   \n",
       "2024-08-19 21:00:00+00:00  1.108279  1.108647  ...       -0.000552   \n",
       "2024-08-19 22:00:00+00:00  1.108647  1.108770  ...        0.000110   \n",
       "\n",
       "                           returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "Datetime                                                                    \n",
       "2024-01-17 00:00:00+00:00       -0.001200       -0.000109        0.000328   \n",
       "2024-01-17 01:00:00+00:00       -0.001635       -0.001200       -0.000109   \n",
       "2024-01-17 02:00:00+00:00       -0.000544       -0.001635       -0.001200   \n",
       "2024-01-17 03:00:00+00:00        0.000980       -0.000544       -0.001635   \n",
       "2024-01-17 04:00:00+00:00       -0.000545        0.000980       -0.000544   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.000110        0.000000        0.000662   \n",
       "2024-08-19 19:00:00+00:00        0.000773       -0.000110        0.000000   \n",
       "2024-08-19 20:00:00+00:00        0.000000        0.000773       -0.000110   \n",
       "2024-08-19 21:00:00+00:00        0.000773        0.000000        0.000773   \n",
       "2024-08-19 22:00:00+00:00       -0.000552        0.000773        0.000000   \n",
       "\n",
       "                           returns_lag_19  returns_lag_20  returns_lag_21  \\\n",
       "Datetime                                                                    \n",
       "2024-01-17 00:00:00+00:00       -0.000218       -0.000109       -0.000437   \n",
       "2024-01-17 01:00:00+00:00        0.000328       -0.000218       -0.000109   \n",
       "2024-01-17 02:00:00+00:00       -0.000109        0.000328       -0.000218   \n",
       "2024-01-17 03:00:00+00:00       -0.001200       -0.000109        0.000328   \n",
       "2024-01-17 04:00:00+00:00       -0.001635       -0.001200       -0.000109   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.000331        0.000110        0.000221   \n",
       "2024-08-19 19:00:00+00:00        0.000662       -0.000331        0.000110   \n",
       "2024-08-19 20:00:00+00:00        0.000000        0.000662       -0.000331   \n",
       "2024-08-19 21:00:00+00:00       -0.000110        0.000000        0.000662   \n",
       "2024-08-19 22:00:00+00:00        0.000773       -0.000110        0.000000   \n",
       "\n",
       "                           returns_lag_22  returns_lag_23  returns_lag_24  \n",
       "Datetime                                                                   \n",
       "2024-01-17 00:00:00+00:00       -0.000328        0.000656       -0.002296  \n",
       "2024-01-17 01:00:00+00:00       -0.000437       -0.000328        0.000656  \n",
       "2024-01-17 02:00:00+00:00       -0.000109       -0.000437       -0.000328  \n",
       "2024-01-17 03:00:00+00:00       -0.000218       -0.000109       -0.000437  \n",
       "2024-01-17 04:00:00+00:00        0.000328       -0.000218       -0.000109  \n",
       "...                                   ...             ...             ...  \n",
       "2024-08-19 18:00:00+00:00        0.000331        0.000771        0.000771  \n",
       "2024-08-19 19:00:00+00:00        0.000221        0.000331        0.000771  \n",
       "2024-08-19 20:00:00+00:00        0.000110        0.000221        0.000331  \n",
       "2024-08-19 21:00:00+00:00       -0.000331        0.000110        0.000221  \n",
       "2024-08-19 22:00:00+00:00        0.000662       -0.000331        0.000110  \n",
       "\n",
       "[3663 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2024-01-17 00:00:00+00:00</th>\n",
       "      <td>2031.599976</td>\n",
       "      <td>2532.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.087548</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "      <td>-0.002296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 01:00:00+00:00</th>\n",
       "      <td>2033.400024</td>\n",
       "      <td>4852.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.088376</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.088021</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 02:00:00+00:00</th>\n",
       "      <td>2032.000000</td>\n",
       "      <td>5865.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.088021</td>\n",
       "      <td>1.088732</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.087784</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000980</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 03:00:00+00:00</th>\n",
       "      <td>2027.199951</td>\n",
       "      <td>4409.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.087784</td>\n",
       "      <td>1.086838</td>\n",
       "      <td>1.086957</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000545</td>\n",
       "      <td>0.000980</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 04:00:00+00:00</th>\n",
       "      <td>2028.500000</td>\n",
       "      <td>2729.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.086838</td>\n",
       "      <td>1.087429</td>\n",
       "      <td>1.086838</td>\n",
       "      <td>1.087311</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001197</td>\n",
       "      <td>-0.000545</td>\n",
       "      <td>0.000980</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</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>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>2542.800049</td>\n",
       "      <td>5353.0</td>\n",
       "      <td>5588.919922</td>\n",
       "      <td>173170955.0</td>\n",
       "      <td>0.001423</td>\n",
       "      <td>0.001019</td>\n",
       "      <td>1.108156</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>1.108033</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "      <td>0.000771</td>\n",
       "      <td>0.000771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>2543.899902</td>\n",
       "      <td>3701.0</td>\n",
       "      <td>5594.100098</td>\n",
       "      <td>215252315.0</td>\n",
       "      <td>0.001372</td>\n",
       "      <td>0.000926</td>\n",
       "      <td>1.108402</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "      <td>0.000771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.109139</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000552</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>2542.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>-0.000552</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3663 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 16
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:41.228233Z",
     "start_time": "2024-08-24T06:42:41.191940Z"
    }
   },
   "cell_type": "code",
   "source": "mu, std = train.mean(), train.std()",
   "id": "8f48ac99ae1032ad",
   "outputs": [],
   "execution_count": 17
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:43.122374Z",
     "start_time": "2024-08-24T06:42:43.086645Z"
    }
   },
   "cell_type": "code",
   "source": "train_standardized = (train-mu) / std",
   "id": "1b08cdd81e4ea00e",
   "outputs": [],
   "execution_count": 18
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:44.707853Z",
     "start_time": "2024-08-24T06:42:44.685508Z"
    }
   },
   "cell_type": "code",
   "source": "train_standardized",
   "id": "98f12f4f990ffb87",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                           gold_price  gold_volume  snp_price  snp_volume  \\\n",
       "Datetime                                                                    \n",
       "2022-08-30 15:00:00+00:00   -1.390359     0.377020  -0.680478    0.005995   \n",
       "2022-08-30 16:00:00+00:00   -1.407350    -0.266903  -0.680409   -0.894752   \n",
       "2022-08-30 17:00:00+00:00   -1.407350    -0.121810  -0.735621   -0.984094   \n",
       "2022-08-30 18:00:00+00:00   -1.408200    -0.603934  -0.737599   -0.928190   \n",
       "2022-08-30 19:00:00+00:00   -1.414996    -0.502532  -0.688662   -0.663624   \n",
       "...                               ...          ...        ...         ...   \n",
       "2024-01-16 19:00:00+00:00    1.115798    -0.289333   1.912170   -0.742633   \n",
       "2024-01-16 20:00:00+00:00    1.107303    -0.539215   1.927855   -0.512521   \n",
       "2024-01-16 21:00:00+00:00    1.106452    -0.872157   1.956434    0.278652   \n",
       "2024-01-16 22:00:00+00:00    1.106452    -0.872157   1.956434    0.278652   \n",
       "2024-01-16 23:00:00+00:00    1.103055    -0.474378   1.956434    0.278652   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2022-08-30 15:00:00+00:00        2.723099    -1.203663 -1.805296 -1.720132   \n",
       "2022-08-30 16:00:00+00:00        0.391589    -0.009288 -1.725986 -1.720132   \n",
       "2022-08-30 17:00:00+00:00        0.869719    -1.409390 -1.709517 -1.736682   \n",
       "2022-08-30 18:00:00+00:00        0.635038    -0.061222 -1.734210 -1.728408   \n",
       "2022-08-30 19:00:00+00:00        1.189899     1.228917 -1.723242 -1.720132   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-01-16 19:00:00+00:00        0.381743    -1.013122  0.594868  0.589711   \n",
       "2024-01-16 20:00:00+00:00       -0.530956     0.321875  0.604557  0.592955   \n",
       "2024-01-16 21:00:00+00:00       -0.426632     0.594675  0.601325  0.596200   \n",
       "2024-01-16 22:00:00+00:00       -0.426632     0.594675  0.614249  0.599444   \n",
       "2024-01-16 23:00:00+00:00       -0.426632     0.594675  0.611017  0.599444   \n",
       "\n",
       "                                Low     Close  ...  returns_lag_15  \\\n",
       "Datetime                                       ...                   \n",
       "2022-08-30 15:00:00+00:00 -1.779931 -1.731903  ...       -0.352815   \n",
       "2022-08-30 16:00:00+00:00 -1.709102 -1.715443  ...       -1.383947   \n",
       "2022-08-30 17:00:00+00:00 -1.725482 -1.734644  ...       -0.008608   \n",
       "2022-08-30 18:00:00+00:00 -1.711833 -1.723674  ...        0.678824   \n",
       "2022-08-30 19:00:00+00:00 -1.714563 -1.734644  ...        0.077347   \n",
       "...                             ...       ...  ...             ...   \n",
       "2024-01-16 19:00:00+00:00  0.605115  0.604304  ...       -0.102523   \n",
       "2024-01-16 20:00:00+00:00  0.611545  0.601072  ...       -0.196281   \n",
       "2024-01-16 21:00:00+00:00  0.614762  0.607533  ...        0.272980   \n",
       "2024-01-16 22:00:00+00:00  0.621199  0.613997  ...       -0.102523   \n",
       "2024-01-16 23:00:00+00:00  0.627637  0.617230  ...       -1.040472   \n",
       "\n",
       "                           returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "Datetime                                                                    \n",
       "2022-08-30 15:00:00+00:00        0.851402       -0.008841        0.077130   \n",
       "2022-08-30 16:00:00+00:00       -0.353044        0.851392       -0.008818   \n",
       "2022-08-30 17:00:00+00:00       -1.384197       -0.353055        0.851418   \n",
       "2022-08-30 18:00:00+00:00       -0.008830       -1.384209       -0.353034   \n",
       "2022-08-30 19:00:00+00:00        0.678617       -0.008841       -1.384192   \n",
       "...                                   ...             ...             ...   \n",
       "2024-01-16 19:00:00+00:00       -0.384395       -0.290688        0.554691   \n",
       "2024-01-16 20:00:00+00:00       -0.102746       -0.384406       -0.290666   \n",
       "2024-01-16 21:00:00+00:00       -0.196507       -0.102758       -0.384385   \n",
       "2024-01-16 22:00:00+00:00        0.272764       -0.196518       -0.102735   \n",
       "2024-01-16 23:00:00+00:00       -0.102746        0.272754       -0.196496   \n",
       "\n",
       "                           returns_lag_19  returns_lag_20  returns_lag_21  \\\n",
       "Datetime                                                                    \n",
       "2022-08-30 15:00:00+00:00        0.163062        0.506624       -1.384195   \n",
       "2022-08-30 16:00:00+00:00        0.077088        0.162988        0.506688   \n",
       "2022-08-30 17:00:00+00:00       -0.008860        0.077016        0.163089   \n",
       "2022-08-30 18:00:00+00:00        0.851377       -0.008931        0.077125   \n",
       "2022-08-30 19:00:00+00:00       -0.353076        0.851293       -0.008813   \n",
       "...                                   ...             ...             ...   \n",
       "2024-01-16 19:00:00+00:00       -1.982756       -0.385462       -0.102910   \n",
       "2024-01-16 20:00:00+00:00        0.554649       -1.982799       -0.385305   \n",
       "2024-01-16 21:00:00+00:00       -0.290709        0.554570       -1.982477   \n",
       "2024-01-16 22:00:00+00:00       -0.384427       -0.290776        0.554630   \n",
       "2024-01-16 23:00:00+00:00       -0.102777       -0.384493       -0.290629   \n",
       "\n",
       "                           returns_lag_22  returns_lag_23  returns_lag_24  \n",
       "Datetime                                                                   \n",
       "2022-08-30 15:00:00+00:00       -0.611203        1.539181        1.107184  \n",
       "2022-08-30 16:00:00+00:00       -1.384134       -0.611245        1.539063  \n",
       "2022-08-30 17:00:00+00:00        0.506717       -1.384083       -0.611703  \n",
       "2022-08-30 18:00:00+00:00        0.163123        0.506541       -1.384664  \n",
       "2022-08-30 19:00:00+00:00        0.077161        0.162988        0.506259  \n",
       "...                                   ...             ...             ...  \n",
       "2024-01-16 19:00:00+00:00       -0.102954        0.085278       -0.103435  \n",
       "2024-01-16 20:00:00+00:00       -0.102871       -0.103057        0.084930  \n",
       "2024-01-16 21:00:00+00:00       -0.385261       -0.102974       -0.103435  \n",
       "2024-01-16 22:00:00+00:00       -1.982406       -0.385329       -0.103352  \n",
       "2024-01-16 23:00:00+00:00        0.554658       -1.982282       -0.385752  \n",
       "\n",
       "[8544 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2022-08-30 15:00:00+00:00</th>\n",
       "      <td>-1.390359</td>\n",
       "      <td>0.377020</td>\n",
       "      <td>-0.680478</td>\n",
       "      <td>0.005995</td>\n",
       "      <td>2.723099</td>\n",
       "      <td>-1.203663</td>\n",
       "      <td>-1.805296</td>\n",
       "      <td>-1.720132</td>\n",
       "      <td>-1.779931</td>\n",
       "      <td>-1.731903</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.352815</td>\n",
       "      <td>0.851402</td>\n",
       "      <td>-0.008841</td>\n",
       "      <td>0.077130</td>\n",
       "      <td>0.163062</td>\n",
       "      <td>0.506624</td>\n",
       "      <td>-1.384195</td>\n",
       "      <td>-0.611203</td>\n",
       "      <td>1.539181</td>\n",
       "      <td>1.107184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 16:00:00+00:00</th>\n",
       "      <td>-1.407350</td>\n",
       "      <td>-0.266903</td>\n",
       "      <td>-0.680409</td>\n",
       "      <td>-0.894752</td>\n",
       "      <td>0.391589</td>\n",
       "      <td>-0.009288</td>\n",
       "      <td>-1.725986</td>\n",
       "      <td>-1.720132</td>\n",
       "      <td>-1.709102</td>\n",
       "      <td>-1.715443</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.383947</td>\n",
       "      <td>-0.353044</td>\n",
       "      <td>0.851392</td>\n",
       "      <td>-0.008818</td>\n",
       "      <td>0.077088</td>\n",
       "      <td>0.162988</td>\n",
       "      <td>0.506688</td>\n",
       "      <td>-1.384134</td>\n",
       "      <td>-0.611245</td>\n",
       "      <td>1.539063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 17:00:00+00:00</th>\n",
       "      <td>-1.407350</td>\n",
       "      <td>-0.121810</td>\n",
       "      <td>-0.735621</td>\n",
       "      <td>-0.984094</td>\n",
       "      <td>0.869719</td>\n",
       "      <td>-1.409390</td>\n",
       "      <td>-1.709517</td>\n",
       "      <td>-1.736682</td>\n",
       "      <td>-1.725482</td>\n",
       "      <td>-1.734644</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008608</td>\n",
       "      <td>-1.384197</td>\n",
       "      <td>-0.353055</td>\n",
       "      <td>0.851418</td>\n",
       "      <td>-0.008860</td>\n",
       "      <td>0.077016</td>\n",
       "      <td>0.163089</td>\n",
       "      <td>0.506717</td>\n",
       "      <td>-1.384083</td>\n",
       "      <td>-0.611703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 18:00:00+00:00</th>\n",
       "      <td>-1.408200</td>\n",
       "      <td>-0.603934</td>\n",
       "      <td>-0.737599</td>\n",
       "      <td>-0.928190</td>\n",
       "      <td>0.635038</td>\n",
       "      <td>-0.061222</td>\n",
       "      <td>-1.734210</td>\n",
       "      <td>-1.728408</td>\n",
       "      <td>-1.711833</td>\n",
       "      <td>-1.723674</td>\n",
       "      <td>...</td>\n",
       "      <td>0.678824</td>\n",
       "      <td>-0.008830</td>\n",
       "      <td>-1.384209</td>\n",
       "      <td>-0.353034</td>\n",
       "      <td>0.851377</td>\n",
       "      <td>-0.008931</td>\n",
       "      <td>0.077125</td>\n",
       "      <td>0.163123</td>\n",
       "      <td>0.506541</td>\n",
       "      <td>-1.384664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-08-30 19:00:00+00:00</th>\n",
       "      <td>-1.414996</td>\n",
       "      <td>-0.502532</td>\n",
       "      <td>-0.688662</td>\n",
       "      <td>-0.663624</td>\n",
       "      <td>1.189899</td>\n",
       "      <td>1.228917</td>\n",
       "      <td>-1.723242</td>\n",
       "      <td>-1.720132</td>\n",
       "      <td>-1.714563</td>\n",
       "      <td>-1.734644</td>\n",
       "      <td>...</td>\n",
       "      <td>0.077347</td>\n",
       "      <td>0.678617</td>\n",
       "      <td>-0.008841</td>\n",
       "      <td>-1.384192</td>\n",
       "      <td>-0.353076</td>\n",
       "      <td>0.851293</td>\n",
       "      <td>-0.008813</td>\n",
       "      <td>0.077161</td>\n",
       "      <td>0.162988</td>\n",
       "      <td>0.506259</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>2024-01-16 19:00:00+00:00</th>\n",
       "      <td>1.115798</td>\n",
       "      <td>-0.289333</td>\n",
       "      <td>1.912170</td>\n",
       "      <td>-0.742633</td>\n",
       "      <td>0.381743</td>\n",
       "      <td>-1.013122</td>\n",
       "      <td>0.594868</td>\n",
       "      <td>0.589711</td>\n",
       "      <td>0.605115</td>\n",
       "      <td>0.604304</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.102523</td>\n",
       "      <td>-0.384395</td>\n",
       "      <td>-0.290688</td>\n",
       "      <td>0.554691</td>\n",
       "      <td>-1.982756</td>\n",
       "      <td>-0.385462</td>\n",
       "      <td>-0.102910</td>\n",
       "      <td>-0.102954</td>\n",
       "      <td>0.085278</td>\n",
       "      <td>-0.103435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 20:00:00+00:00</th>\n",
       "      <td>1.107303</td>\n",
       "      <td>-0.539215</td>\n",
       "      <td>1.927855</td>\n",
       "      <td>-0.512521</td>\n",
       "      <td>-0.530956</td>\n",
       "      <td>0.321875</td>\n",
       "      <td>0.604557</td>\n",
       "      <td>0.592955</td>\n",
       "      <td>0.611545</td>\n",
       "      <td>0.601072</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.196281</td>\n",
       "      <td>-0.102746</td>\n",
       "      <td>-0.384406</td>\n",
       "      <td>-0.290666</td>\n",
       "      <td>0.554649</td>\n",
       "      <td>-1.982799</td>\n",
       "      <td>-0.385305</td>\n",
       "      <td>-0.102871</td>\n",
       "      <td>-0.103057</td>\n",
       "      <td>0.084930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 21:00:00+00:00</th>\n",
       "      <td>1.106452</td>\n",
       "      <td>-0.872157</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.601325</td>\n",
       "      <td>0.596200</td>\n",
       "      <td>0.614762</td>\n",
       "      <td>0.607533</td>\n",
       "      <td>...</td>\n",
       "      <td>0.272980</td>\n",
       "      <td>-0.196507</td>\n",
       "      <td>-0.102758</td>\n",
       "      <td>-0.384385</td>\n",
       "      <td>-0.290709</td>\n",
       "      <td>0.554570</td>\n",
       "      <td>-1.982477</td>\n",
       "      <td>-0.385261</td>\n",
       "      <td>-0.102974</td>\n",
       "      <td>-0.103435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 22:00:00+00:00</th>\n",
       "      <td>1.106452</td>\n",
       "      <td>-0.872157</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.614249</td>\n",
       "      <td>0.599444</td>\n",
       "      <td>0.621199</td>\n",
       "      <td>0.613997</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.102523</td>\n",
       "      <td>0.272764</td>\n",
       "      <td>-0.196518</td>\n",
       "      <td>-0.102735</td>\n",
       "      <td>-0.384427</td>\n",
       "      <td>-0.290776</td>\n",
       "      <td>0.554630</td>\n",
       "      <td>-1.982406</td>\n",
       "      <td>-0.385329</td>\n",
       "      <td>-0.103352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-16 23:00:00+00:00</th>\n",
       "      <td>1.103055</td>\n",
       "      <td>-0.474378</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.611017</td>\n",
       "      <td>0.599444</td>\n",
       "      <td>0.627637</td>\n",
       "      <td>0.617230</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.040472</td>\n",
       "      <td>-0.102746</td>\n",
       "      <td>0.272754</td>\n",
       "      <td>-0.196496</td>\n",
       "      <td>-0.102777</td>\n",
       "      <td>-0.384493</td>\n",
       "      <td>-0.290629</td>\n",
       "      <td>0.554658</td>\n",
       "      <td>-1.982282</td>\n",
       "      <td>-0.385752</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8544 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 19
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:53.687764Z",
     "start_time": "2024-08-24T06:42:53.411944Z"
    }
   },
   "cell_type": "code",
   "source": "train_standardized.describe()",
   "id": "e6c67bda1c7076a9",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "         gold_price   gold_volume     snp_price    snp_volume  snp_volatility  \\\n",
       "count  8.544000e+03  8.544000e+03  8.544000e+03  8.544000e+03    8.544000e+03   \n",
       "mean   3.326511e-16 -4.989766e-17 -8.515868e-16  1.097749e-16    2.328558e-17   \n",
       "std    1.000000e+00  1.000000e+00  1.000000e+00  1.000000e+00    1.000000e+00   \n",
       "min   -2.388574e+00 -8.721570e-01 -2.252687e+00 -2.556520e+00   -1.296278e+00   \n",
       "25%   -6.538043e-01 -5.919013e-01 -7.651186e-01 -4.081163e-01   -6.475696e-01   \n",
       "50%    3.274200e-01 -3.490285e-01 -1.827236e-01  7.146359e-03   -2.587220e-01   \n",
       "75%    7.190599e-01  1.521083e-01  8.410904e-01  3.241067e-01    3.892928e-01   \n",
       "max    1.819222e+00  9.515998e+00  2.047486e+00  1.609350e+01    1.263305e+01   \n",
       "\n",
       "        snp_returns          Open          High           Low         Close  \\\n",
       "count  8.544000e+03  8.544000e+03  8.544000e+03  8.544000e+03  8.544000e+03   \n",
       "mean  -2.328558e-17 -1.543501e-15 -2.355170e-15  2.428353e-15  2.860799e-16   \n",
       "std    1.000000e+00  1.000000e+00  1.000000e+00  1.000000e+00  1.000000e+00   \n",
       "min   -9.669546e+00 -3.024487e+00 -3.011762e+00 -3.012056e+00 -3.037459e+00   \n",
       "25%   -4.139266e-01 -2.548273e-01 -2.579295e-01 -2.509985e-01 -2.551464e-01   \n",
       "50%    1.767115e-02  2.282563e-01  2.276699e-01  2.305846e-01  2.279746e-01   \n",
       "75%    4.528527e-01  7.115255e-01  7.101979e-01  7.115824e-01  7.112811e-01   \n",
       "max    1.401245e+01  1.651148e+00  1.685896e+00  1.646766e+00  1.654439e+00   \n",
       "\n",
       "       ...  returns_lag_15  returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "count  ...    8.544000e+03    8.544000e+03    8.544000e+03    8.544000e+03   \n",
       "mean   ...   -1.122697e-17   -8.316277e-19    2.079069e-18    2.494883e-18   \n",
       "std    ...    1.000000e+00    1.000000e+00    1.000000e+00    1.000000e+00   \n",
       "min    ...   -1.200014e+01   -1.200062e+01   -1.200063e+01   -1.200066e+01   \n",
       "25%    ...   -3.854174e-01   -3.856387e-01   -3.856499e-01   -3.856286e-01   \n",
       "50%    ...   -8.607894e-03   -8.829530e-03   -8.840528e-03   -8.817664e-03   \n",
       "75%    ...    3.723860e-01    3.723056e-01    3.722948e-01    3.723192e-01   \n",
       "max    ...    1.500463e+01    1.500473e+01    1.500473e+01    1.500481e+01   \n",
       "\n",
       "       returns_lag_19  returns_lag_20  returns_lag_21  returns_lag_22  \\\n",
       "count    8.544000e+03    8.544000e+03    8.544000e+03    8.544000e+03   \n",
       "mean     1.663255e-18   -6.237208e-19    4.989766e-18   -5.613487e-18   \n",
       "std      1.000000e+00    1.000000e+00    1.000000e+00    1.000000e+00   \n",
       "min     -1.200071e+01   -1.200061e+01   -1.199925e+01   -1.199901e+01   \n",
       "25%     -3.856709e-01   -3.857367e-01   -3.855880e-01   -3.855542e-01   \n",
       "50%     -8.859759e-03   -8.930964e-03   -8.813015e-03   -8.775341e-03   \n",
       "75%      3.722774e-01    3.726822e-01    3.727607e-01    3.727919e-01   \n",
       "max      1.500478e+01    1.500449e+01    1.500305e+01    1.500284e+01   \n",
       "\n",
       "       returns_lag_23  returns_lag_24  \n",
       "count    8.544000e+03    8.544000e+03  \n",
       "mean     4.158139e-19   -9.355812e-18  \n",
       "std      1.000000e+00    1.000000e+00  \n",
       "min     -1.199768e+01   -1.199994e+01  \n",
       "25%     -3.856230e-01   -3.860354e-01  \n",
       "50%     -8.889533e-03   -9.252614e-03  \n",
       "75%      3.726317e-01    3.727810e-01  \n",
       "max      1.500091e+01    1.500292e+01  \n",
       "\n",
       "[8 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>...</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "      <td>8.544000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>3.326511e-16</td>\n",
       "      <td>-4.989766e-17</td>\n",
       "      <td>-8.515868e-16</td>\n",
       "      <td>1.097749e-16</td>\n",
       "      <td>2.328558e-17</td>\n",
       "      <td>-2.328558e-17</td>\n",
       "      <td>-1.543501e-15</td>\n",
       "      <td>-2.355170e-15</td>\n",
       "      <td>2.428353e-15</td>\n",
       "      <td>2.860799e-16</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.122697e-17</td>\n",
       "      <td>-8.316277e-19</td>\n",
       "      <td>2.079069e-18</td>\n",
       "      <td>2.494883e-18</td>\n",
       "      <td>1.663255e-18</td>\n",
       "      <td>-6.237208e-19</td>\n",
       "      <td>4.989766e-18</td>\n",
       "      <td>-5.613487e-18</td>\n",
       "      <td>4.158139e-19</td>\n",
       "      <td>-9.355812e-18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-2.388574e+00</td>\n",
       "      <td>-8.721570e-01</td>\n",
       "      <td>-2.252687e+00</td>\n",
       "      <td>-2.556520e+00</td>\n",
       "      <td>-1.296278e+00</td>\n",
       "      <td>-9.669546e+00</td>\n",
       "      <td>-3.024487e+00</td>\n",
       "      <td>-3.011762e+00</td>\n",
       "      <td>-3.012056e+00</td>\n",
       "      <td>-3.037459e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.200014e+01</td>\n",
       "      <td>-1.200062e+01</td>\n",
       "      <td>-1.200063e+01</td>\n",
       "      <td>-1.200066e+01</td>\n",
       "      <td>-1.200071e+01</td>\n",
       "      <td>-1.200061e+01</td>\n",
       "      <td>-1.199925e+01</td>\n",
       "      <td>-1.199901e+01</td>\n",
       "      <td>-1.199768e+01</td>\n",
       "      <td>-1.199994e+01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-6.538043e-01</td>\n",
       "      <td>-5.919013e-01</td>\n",
       "      <td>-7.651186e-01</td>\n",
       "      <td>-4.081163e-01</td>\n",
       "      <td>-6.475696e-01</td>\n",
       "      <td>-4.139266e-01</td>\n",
       "      <td>-2.548273e-01</td>\n",
       "      <td>-2.579295e-01</td>\n",
       "      <td>-2.509985e-01</td>\n",
       "      <td>-2.551464e-01</td>\n",
       "      <td>...</td>\n",
       "      <td>-3.854174e-01</td>\n",
       "      <td>-3.856387e-01</td>\n",
       "      <td>-3.856499e-01</td>\n",
       "      <td>-3.856286e-01</td>\n",
       "      <td>-3.856709e-01</td>\n",
       "      <td>-3.857367e-01</td>\n",
       "      <td>-3.855880e-01</td>\n",
       "      <td>-3.855542e-01</td>\n",
       "      <td>-3.856230e-01</td>\n",
       "      <td>-3.860354e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.274200e-01</td>\n",
       "      <td>-3.490285e-01</td>\n",
       "      <td>-1.827236e-01</td>\n",
       "      <td>7.146359e-03</td>\n",
       "      <td>-2.587220e-01</td>\n",
       "      <td>1.767115e-02</td>\n",
       "      <td>2.282563e-01</td>\n",
       "      <td>2.276699e-01</td>\n",
       "      <td>2.305846e-01</td>\n",
       "      <td>2.279746e-01</td>\n",
       "      <td>...</td>\n",
       "      <td>-8.607894e-03</td>\n",
       "      <td>-8.829530e-03</td>\n",
       "      <td>-8.840528e-03</td>\n",
       "      <td>-8.817664e-03</td>\n",
       "      <td>-8.859759e-03</td>\n",
       "      <td>-8.930964e-03</td>\n",
       "      <td>-8.813015e-03</td>\n",
       "      <td>-8.775341e-03</td>\n",
       "      <td>-8.889533e-03</td>\n",
       "      <td>-9.252614e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.190599e-01</td>\n",
       "      <td>1.521083e-01</td>\n",
       "      <td>8.410904e-01</td>\n",
       "      <td>3.241067e-01</td>\n",
       "      <td>3.892928e-01</td>\n",
       "      <td>4.528527e-01</td>\n",
       "      <td>7.115255e-01</td>\n",
       "      <td>7.101979e-01</td>\n",
       "      <td>7.115824e-01</td>\n",
       "      <td>7.112811e-01</td>\n",
       "      <td>...</td>\n",
       "      <td>3.723860e-01</td>\n",
       "      <td>3.723056e-01</td>\n",
       "      <td>3.722948e-01</td>\n",
       "      <td>3.723192e-01</td>\n",
       "      <td>3.722774e-01</td>\n",
       "      <td>3.726822e-01</td>\n",
       "      <td>3.727607e-01</td>\n",
       "      <td>3.727919e-01</td>\n",
       "      <td>3.726317e-01</td>\n",
       "      <td>3.727810e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.819222e+00</td>\n",
       "      <td>9.515998e+00</td>\n",
       "      <td>2.047486e+00</td>\n",
       "      <td>1.609350e+01</td>\n",
       "      <td>1.263305e+01</td>\n",
       "      <td>1.401245e+01</td>\n",
       "      <td>1.651148e+00</td>\n",
       "      <td>1.685896e+00</td>\n",
       "      <td>1.646766e+00</td>\n",
       "      <td>1.654439e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>1.500463e+01</td>\n",
       "      <td>1.500473e+01</td>\n",
       "      <td>1.500473e+01</td>\n",
       "      <td>1.500481e+01</td>\n",
       "      <td>1.500478e+01</td>\n",
       "      <td>1.500449e+01</td>\n",
       "      <td>1.500305e+01</td>\n",
       "      <td>1.500284e+01</td>\n",
       "      <td>1.500091e+01</td>\n",
       "      <td>1.500292e+01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 20
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:42:58.749843Z",
     "start_time": "2024-08-24T06:42:57.013485Z"
    }
   },
   "cell_type": "code",
   "source": "from DNNModel import *",
   "id": "3e39c64e99d5d3c8",
   "outputs": [],
   "execution_count": 21
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:44:00.385359Z",
     "start_time": "2024-08-24T06:43:40.401546Z"
    }
   },
   "cell_type": "code",
   "source": [
    "set_seeds(100)\n",
    "\n",
    "model = create_model(hl=3, hu=50, dropout= True, input_dim=len(cols))\n",
    "model.fit(x=train_standardized[cols], y=train['dir'], epochs=50, verbose=True,\n",
    "          validation_split=0.2, shuffle=False, class_weight=cw(train))"
   ],
   "id": "68eedf41d74ae073",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\code\\trading\\v1\\pythonProject\\.venv\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n",
      "C:\\code\\trading\\v1\\pythonProject\\.venv\\Lib\\site-packages\\keras\\src\\trainers\\data_adapters\\data_adapter_utils.py:126: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`\n",
      "  sample_weight[i] = class_weight.get(int(y[i]), 1.0)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m2s\u001B[0m 2ms/step - accuracy: 0.5150 - loss: 0.7380 - val_accuracy: 0.5354 - val_loss: 0.6911\n",
      "Epoch 2/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.4972 - loss: 0.7178 - val_accuracy: 0.5354 - val_loss: 0.6908\n",
      "Epoch 3/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5046 - loss: 0.7099 - val_accuracy: 0.5372 - val_loss: 0.6914\n",
      "Epoch 4/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5003 - loss: 0.7067 - val_accuracy: 0.5366 - val_loss: 0.6913\n",
      "Epoch 5/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5119 - loss: 0.7031 - val_accuracy: 0.5366 - val_loss: 0.6917\n",
      "Epoch 6/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5089 - loss: 0.7015 - val_accuracy: 0.5377 - val_loss: 0.6917\n",
      "Epoch 7/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5069 - loss: 0.7010 - val_accuracy: 0.5266 - val_loss: 0.6915\n",
      "Epoch 8/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5059 - loss: 0.7009 - val_accuracy: 0.5319 - val_loss: 0.6915\n",
      "Epoch 9/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.4914 - loss: 0.7026 - val_accuracy: 0.5231 - val_loss: 0.6915\n",
      "Epoch 10/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5072 - loss: 0.6971 - val_accuracy: 0.5255 - val_loss: 0.6915\n",
      "Epoch 11/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5004 - loss: 0.6988 - val_accuracy: 0.5260 - val_loss: 0.6919\n",
      "Epoch 12/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5190 - loss: 0.6955 - val_accuracy: 0.5272 - val_loss: 0.6913\n",
      "Epoch 13/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5134 - loss: 0.6949 - val_accuracy: 0.5155 - val_loss: 0.6919\n",
      "Epoch 14/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5084 - loss: 0.6960 - val_accuracy: 0.5184 - val_loss: 0.6921\n",
      "Epoch 15/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5139 - loss: 0.6950 - val_accuracy: 0.5167 - val_loss: 0.6921\n",
      "Epoch 16/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 1ms/step - accuracy: 0.5204 - loss: 0.6962 - val_accuracy: 0.5120 - val_loss: 0.6923\n",
      "Epoch 17/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5131 - loss: 0.6927 - val_accuracy: 0.5214 - val_loss: 0.6922\n",
      "Epoch 18/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.4960 - loss: 0.6962 - val_accuracy: 0.5196 - val_loss: 0.6923\n",
      "Epoch 19/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5205 - loss: 0.6949 - val_accuracy: 0.5196 - val_loss: 0.6922\n",
      "Epoch 20/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5291 - loss: 0.6927 - val_accuracy: 0.5284 - val_loss: 0.6920\n",
      "Epoch 21/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5209 - loss: 0.6925 - val_accuracy: 0.5190 - val_loss: 0.6917\n",
      "Epoch 22/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5018 - loss: 0.6954 - val_accuracy: 0.5243 - val_loss: 0.6922\n",
      "Epoch 23/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5250 - loss: 0.6921 - val_accuracy: 0.5342 - val_loss: 0.6917\n",
      "Epoch 24/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5195 - loss: 0.6935 - val_accuracy: 0.5190 - val_loss: 0.6920\n",
      "Epoch 25/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5078 - loss: 0.6917 - val_accuracy: 0.5196 - val_loss: 0.6919\n",
      "Epoch 26/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5189 - loss: 0.6944 - val_accuracy: 0.5196 - val_loss: 0.6917\n",
      "Epoch 27/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5270 - loss: 0.6927 - val_accuracy: 0.5237 - val_loss: 0.6919\n",
      "Epoch 28/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5184 - loss: 0.6961 - val_accuracy: 0.5255 - val_loss: 0.6917\n",
      "Epoch 29/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m1s\u001B[0m 2ms/step - accuracy: 0.5214 - loss: 0.6921 - val_accuracy: 0.5067 - val_loss: 0.6923\n",
      "Epoch 30/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5159 - loss: 0.6927 - val_accuracy: 0.5079 - val_loss: 0.6919\n",
      "Epoch 31/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5226 - loss: 0.6898 - val_accuracy: 0.5149 - val_loss: 0.6918\n",
      "Epoch 32/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5145 - loss: 0.6929 - val_accuracy: 0.5143 - val_loss: 0.6918\n",
      "Epoch 33/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5321 - loss: 0.6868 - val_accuracy: 0.5178 - val_loss: 0.6923\n",
      "Epoch 34/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5338 - loss: 0.6891 - val_accuracy: 0.5091 - val_loss: 0.6920\n",
      "Epoch 35/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5298 - loss: 0.6906 - val_accuracy: 0.5132 - val_loss: 0.6920\n",
      "Epoch 36/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5351 - loss: 0.6872 - val_accuracy: 0.5184 - val_loss: 0.6920\n",
      "Epoch 37/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5171 - loss: 0.6914 - val_accuracy: 0.5214 - val_loss: 0.6918\n",
      "Epoch 38/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5452 - loss: 0.6879 - val_accuracy: 0.5138 - val_loss: 0.6919\n",
      "Epoch 39/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5360 - loss: 0.6880 - val_accuracy: 0.5196 - val_loss: 0.6919\n",
      "Epoch 40/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5393 - loss: 0.6875 - val_accuracy: 0.5202 - val_loss: 0.6919\n",
      "Epoch 41/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5439 - loss: 0.6880 - val_accuracy: 0.5272 - val_loss: 0.6918\n",
      "Epoch 42/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5374 - loss: 0.6857 - val_accuracy: 0.5167 - val_loss: 0.6921\n",
      "Epoch 43/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5358 - loss: 0.6895 - val_accuracy: 0.5243 - val_loss: 0.6920\n",
      "Epoch 44/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5283 - loss: 0.6879 - val_accuracy: 0.5272 - val_loss: 0.6921\n",
      "Epoch 45/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5494 - loss: 0.6861 - val_accuracy: 0.5278 - val_loss: 0.6929\n",
      "Epoch 46/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5702 - loss: 0.6839 - val_accuracy: 0.5255 - val_loss: 0.6927\n",
      "Epoch 47/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5466 - loss: 0.6855 - val_accuracy: 0.5278 - val_loss: 0.6930\n",
      "Epoch 48/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5476 - loss: 0.6856 - val_accuracy: 0.5266 - val_loss: 0.6927\n",
      "Epoch 49/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5456 - loss: 0.6848 - val_accuracy: 0.5231 - val_loss: 0.6930\n",
      "Epoch 50/50\n",
      "\u001B[1m214/214\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 2ms/step - accuracy: 0.5515 - loss: 0.6866 - val_accuracy: 0.5225 - val_loss: 0.6935\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.history.History at 0x2b3168b4590>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 22
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:44:25.270334Z",
     "start_time": "2024-08-24T06:44:25.006315Z"
    }
   },
   "cell_type": "code",
   "source": "model.evaluate(train_standardized[cols], train['dir'])",
   "id": "97f69a036b05acda",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m267/267\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 676us/step - accuracy: 0.5905 - loss: 0.6785\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.682673454284668, 0.5775983333587646]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 23
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:44:33.292072Z",
     "start_time": "2024-08-24T06:44:32.971515Z"
    }
   },
   "cell_type": "code",
   "source": [
    "predict = model.predict(train_standardized[cols])\n",
    "predict"
   ],
   "id": "7194899747383ceb",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m267/267\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 552us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0.5363864 ],\n",
       "       [0.50734246],\n",
       "       [0.49099603],\n",
       "       ...,\n",
       "       [0.5501518 ],\n",
       "       [0.5044123 ],\n",
       "       [0.48607168]], dtype=float32)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 24
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:44:42.127577Z",
     "start_time": "2024-08-24T06:44:42.031718Z"
    }
   },
   "cell_type": "code",
   "source": [
    "plt.hist(predict, bins=50)\n",
    "plt.show()"
   ],
   "id": "512239abff645d63",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 25
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:44:50.956930Z",
     "start_time": "2024-08-24T06:44:50.932842Z"
    }
   },
   "cell_type": "code",
   "source": "test",
   "id": "ea7601d723907d7d",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                            gold_price  gold_volume    snp_price   snp_volume  \\\n",
       "Datetime                                                                        \n",
       "2024-01-17 00:00:00+00:00  2031.599976       2532.0  4765.180176  336315998.0   \n",
       "2024-01-17 01:00:00+00:00  2033.400024       4852.0  4765.180176  336315998.0   \n",
       "2024-01-17 02:00:00+00:00  2032.000000       5865.0  4765.180176  336315998.0   \n",
       "2024-01-17 03:00:00+00:00  2027.199951       4409.0  4765.180176  336315998.0   \n",
       "2024-01-17 04:00:00+00:00  2028.500000       2729.0  4765.180176  336315998.0   \n",
       "...                                ...          ...          ...          ...   \n",
       "2024-08-19 18:00:00+00:00  2542.800049       5353.0  5588.919922  173170955.0   \n",
       "2024-08-19 19:00:00+00:00  2543.899902       3701.0  5594.100098  215252315.0   \n",
       "2024-08-19 20:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 21:00:00+00:00  2542.600098       1749.0  5607.520020  242772673.0   \n",
       "2024-08-19 22:00:00+00:00  2542.500000          0.0  5607.520020  242772673.0   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2024-01-17 00:00:00+00:00        0.002521     0.001760  1.088139  1.088139   \n",
       "2024-01-17 01:00:00+00:00        0.002521     0.001760  1.088139  1.088376   \n",
       "2024-01-17 02:00:00+00:00        0.002521     0.001760  1.088021  1.088732   \n",
       "2024-01-17 03:00:00+00:00        0.002521     0.001760  1.087666  1.087784   \n",
       "2024-01-17 04:00:00+00:00        0.002521     0.001760  1.086838  1.087429   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-08-19 18:00:00+00:00        0.001423     0.001019  1.108156  1.108770   \n",
       "2024-08-19 19:00:00+00:00        0.001372     0.000926  1.108402  1.108893   \n",
       "2024-08-19 20:00:00+00:00        0.002894     0.002396  1.108647  1.108893   \n",
       "2024-08-19 21:00:00+00:00        0.002894     0.002396  1.108525  1.109139   \n",
       "2024-08-19 22:00:00+00:00        0.002894     0.002396  1.109016  1.109016   \n",
       "\n",
       "                                Low     Close  ...  returns_lag_15  \\\n",
       "Datetime                                       ...                   \n",
       "2024-01-17 00:00:00+00:00  1.087548  1.088139  ...       -0.001635   \n",
       "2024-01-17 01:00:00+00:00  1.087666  1.088021  ...       -0.000544   \n",
       "2024-01-17 02:00:00+00:00  1.087666  1.087784  ...        0.000980   \n",
       "2024-01-17 03:00:00+00:00  1.086838  1.086957  ...       -0.000545   \n",
       "2024-01-17 04:00:00+00:00  1.086838  1.087311  ...       -0.001197   \n",
       "...                             ...       ...  ...             ...   \n",
       "2024-08-19 18:00:00+00:00  1.108033  1.108525  ...        0.000773   \n",
       "2024-08-19 19:00:00+00:00  1.108279  1.108647  ...        0.000000   \n",
       "2024-08-19 20:00:00+00:00  1.108525  1.108893  ...        0.000773   \n",
       "2024-08-19 21:00:00+00:00  1.108279  1.108647  ...       -0.000552   \n",
       "2024-08-19 22:00:00+00:00  1.108647  1.108770  ...        0.000110   \n",
       "\n",
       "                           returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "Datetime                                                                    \n",
       "2024-01-17 00:00:00+00:00       -0.001200       -0.000109        0.000328   \n",
       "2024-01-17 01:00:00+00:00       -0.001635       -0.001200       -0.000109   \n",
       "2024-01-17 02:00:00+00:00       -0.000544       -0.001635       -0.001200   \n",
       "2024-01-17 03:00:00+00:00        0.000980       -0.000544       -0.001635   \n",
       "2024-01-17 04:00:00+00:00       -0.000545        0.000980       -0.000544   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.000110        0.000000        0.000662   \n",
       "2024-08-19 19:00:00+00:00        0.000773       -0.000110        0.000000   \n",
       "2024-08-19 20:00:00+00:00        0.000000        0.000773       -0.000110   \n",
       "2024-08-19 21:00:00+00:00        0.000773        0.000000        0.000773   \n",
       "2024-08-19 22:00:00+00:00       -0.000552        0.000773        0.000000   \n",
       "\n",
       "                           returns_lag_19  returns_lag_20  returns_lag_21  \\\n",
       "Datetime                                                                    \n",
       "2024-01-17 00:00:00+00:00       -0.000218       -0.000109       -0.000437   \n",
       "2024-01-17 01:00:00+00:00        0.000328       -0.000218       -0.000109   \n",
       "2024-01-17 02:00:00+00:00       -0.000109        0.000328       -0.000218   \n",
       "2024-01-17 03:00:00+00:00       -0.001200       -0.000109        0.000328   \n",
       "2024-01-17 04:00:00+00:00       -0.001635       -0.001200       -0.000109   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.000331        0.000110        0.000221   \n",
       "2024-08-19 19:00:00+00:00        0.000662       -0.000331        0.000110   \n",
       "2024-08-19 20:00:00+00:00        0.000000        0.000662       -0.000331   \n",
       "2024-08-19 21:00:00+00:00       -0.000110        0.000000        0.000662   \n",
       "2024-08-19 22:00:00+00:00        0.000773       -0.000110        0.000000   \n",
       "\n",
       "                           returns_lag_22  returns_lag_23  returns_lag_24  \n",
       "Datetime                                                                   \n",
       "2024-01-17 00:00:00+00:00       -0.000328        0.000656       -0.002296  \n",
       "2024-01-17 01:00:00+00:00       -0.000437       -0.000328        0.000656  \n",
       "2024-01-17 02:00:00+00:00       -0.000109       -0.000437       -0.000328  \n",
       "2024-01-17 03:00:00+00:00       -0.000218       -0.000109       -0.000437  \n",
       "2024-01-17 04:00:00+00:00        0.000328       -0.000218       -0.000109  \n",
       "...                                   ...             ...             ...  \n",
       "2024-08-19 18:00:00+00:00        0.000331        0.000771        0.000771  \n",
       "2024-08-19 19:00:00+00:00        0.000221        0.000331        0.000771  \n",
       "2024-08-19 20:00:00+00:00        0.000110        0.000221        0.000331  \n",
       "2024-08-19 21:00:00+00:00       -0.000331        0.000110        0.000221  \n",
       "2024-08-19 22:00:00+00:00        0.000662       -0.000331        0.000110  \n",
       "\n",
       "[3663 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2024-01-17 00:00:00+00:00</th>\n",
       "      <td>2031.599976</td>\n",
       "      <td>2532.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.087548</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "      <td>-0.002296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 01:00:00+00:00</th>\n",
       "      <td>2033.400024</td>\n",
       "      <td>4852.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.088139</td>\n",
       "      <td>1.088376</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.088021</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "      <td>0.000656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 02:00:00+00:00</th>\n",
       "      <td>2032.000000</td>\n",
       "      <td>5865.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.088021</td>\n",
       "      <td>1.088732</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.087784</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000980</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "      <td>-0.000328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 03:00:00+00:00</th>\n",
       "      <td>2027.199951</td>\n",
       "      <td>4409.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.087666</td>\n",
       "      <td>1.087784</td>\n",
       "      <td>1.086838</td>\n",
       "      <td>1.086957</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000545</td>\n",
       "      <td>0.000980</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>-0.000437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 04:00:00+00:00</th>\n",
       "      <td>2028.500000</td>\n",
       "      <td>2729.0</td>\n",
       "      <td>4765.180176</td>\n",
       "      <td>336315998.0</td>\n",
       "      <td>0.002521</td>\n",
       "      <td>0.001760</td>\n",
       "      <td>1.086838</td>\n",
       "      <td>1.087429</td>\n",
       "      <td>1.086838</td>\n",
       "      <td>1.087311</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001197</td>\n",
       "      <td>-0.000545</td>\n",
       "      <td>0.000980</td>\n",
       "      <td>-0.000544</td>\n",
       "      <td>-0.001635</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>-0.000109</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>-0.000218</td>\n",
       "      <td>-0.000109</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>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>2542.800049</td>\n",
       "      <td>5353.0</td>\n",
       "      <td>5588.919922</td>\n",
       "      <td>173170955.0</td>\n",
       "      <td>0.001423</td>\n",
       "      <td>0.001019</td>\n",
       "      <td>1.108156</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>1.108033</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "      <td>0.000771</td>\n",
       "      <td>0.000771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>2543.899902</td>\n",
       "      <td>3701.0</td>\n",
       "      <td>5594.100098</td>\n",
       "      <td>215252315.0</td>\n",
       "      <td>0.001372</td>\n",
       "      <td>0.000926</td>\n",
       "      <td>1.108402</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "      <td>0.000771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.108893</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>2542.600098</td>\n",
       "      <td>1749.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.108525</td>\n",
       "      <td>1.109139</td>\n",
       "      <td>1.108279</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000552</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>0.000221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>2542.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5607.520020</td>\n",
       "      <td>242772673.0</td>\n",
       "      <td>0.002894</td>\n",
       "      <td>0.002396</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.109016</td>\n",
       "      <td>1.108647</td>\n",
       "      <td>1.108770</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000110</td>\n",
       "      <td>-0.000552</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000773</td>\n",
       "      <td>-0.000110</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000662</td>\n",
       "      <td>-0.000331</td>\n",
       "      <td>0.000110</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3663 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 26
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:44:58.821809Z",
     "start_time": "2024-08-24T06:44:58.781482Z"
    }
   },
   "cell_type": "code",
   "source": [
    "test_standardized = (test-mu) / std\n",
    "test_standardized"
   ],
   "id": "a01273953021c394",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "                           gold_price  gold_volume  snp_price  snp_volume  \\\n",
       "Datetime                                                                    \n",
       "2024-01-17 00:00:00+00:00    1.105603    -0.576364   1.956434    0.278652   \n",
       "2024-01-17 01:00:00+00:00    1.120895    -0.305337   1.956434    0.278652   \n",
       "2024-01-17 02:00:00+00:00    1.109001    -0.186997   1.956434    0.278652   \n",
       "2024-01-17 03:00:00+00:00    1.068223    -0.357089   1.956434    0.278652   \n",
       "2024-01-17 04:00:00+00:00    1.079267    -0.553350   1.956434    0.278652   \n",
       "...                               ...          ...        ...         ...   \n",
       "2024-08-19 18:00:00+00:00    5.448477    -0.246809   4.765560   -1.096674   \n",
       "2024-08-19 19:00:00+00:00    5.457820    -0.439799   4.783225   -0.741925   \n",
       "2024-08-19 20:00:00+00:00    5.446778    -0.667835   4.828990   -0.509926   \n",
       "2024-08-19 21:00:00+00:00    5.446778    -0.667835   4.828990   -0.509926   \n",
       "2024-08-19 22:00:00+00:00    5.445927    -0.872157   4.828990   -0.509926   \n",
       "\n",
       "                           snp_volatility  snp_returns      Open      High  \\\n",
       "Datetime                                                                     \n",
       "2024-01-17 00:00:00+00:00       -0.426632     0.594675  0.617481  0.599444   \n",
       "2024-01-17 01:00:00+00:00       -0.426632     0.594675  0.617481  0.605937   \n",
       "2024-01-17 02:00:00+00:00       -0.426632     0.594675  0.614249  0.615680   \n",
       "2024-01-17 03:00:00+00:00       -0.426632     0.594675  0.604557  0.589711   \n",
       "2024-01-17 04:00:00+00:00       -0.426632     0.594675  0.581963  0.579983   \n",
       "...                                   ...          ...       ...       ...   \n",
       "2024-08-19 18:00:00+00:00       -0.918821     0.339483  1.163983  1.164930   \n",
       "2024-08-19 19:00:00+00:00       -0.941987     0.307772  1.170690  1.168302   \n",
       "2024-08-19 20:00:00+00:00       -0.258942     0.813471  1.177402  1.168302   \n",
       "2024-08-19 21:00:00+00:00       -0.258942     0.813471  1.174046  1.175042   \n",
       "2024-08-19 22:00:00+00:00       -0.258942     0.813471  1.187472  1.171670   \n",
       "\n",
       "                                Low     Close  ...  returns_lag_15  \\\n",
       "Datetime                                       ...                   \n",
       "2024-01-17 00:00:00+00:00  0.617979  0.617230  ...       -1.413768   \n",
       "2024-01-17 01:00:00+00:00  0.621199  0.613997  ...       -0.476470   \n",
       "2024-01-17 02:00:00+00:00  0.621199  0.607533  ...        0.833662   \n",
       "2024-01-17 03:00:00+00:00  0.598687  0.584934  ...       -0.476643   \n",
       "2024-01-17 04:00:00+00:00  0.598687  0.594614  ...       -1.037486   \n",
       "...                             ...       ...  ...             ...   \n",
       "2024-08-19 18:00:00+00:00  1.175263  1.173837  ...        0.655571   \n",
       "2024-08-19 19:00:00+00:00  1.181944  1.177193  ...       -0.008608   \n",
       "2024-08-19 20:00:00+00:00  1.188627  1.183908  ...        0.655986   \n",
       "2024-08-19 21:00:00+00:00  1.181944  1.177193  ...       -0.483371   \n",
       "2024-08-19 22:00:00+00:00  1.191971  1.180549  ...        0.086292   \n",
       "\n",
       "                           returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "Datetime                                                                    \n",
       "2024-01-17 00:00:00+00:00       -1.040715       -0.102758        0.272778   \n",
       "2024-01-17 01:00:00+00:00       -1.414019       -1.040727       -0.102735   \n",
       "2024-01-17 02:00:00+00:00       -0.476701       -1.414031       -1.040708   \n",
       "2024-01-17 03:00:00+00:00        0.833458       -0.476713       -1.414014   \n",
       "2024-01-17 04:00:00+00:00       -0.476874        0.833447       -0.476692   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.103732       -0.008841        0.560109   \n",
       "2024-08-19 19:00:00+00:00        0.655363       -0.103743       -0.008818   \n",
       "2024-08-19 20:00:00+00:00       -0.008830        0.655353       -0.103721   \n",
       "2024-08-19 21:00:00+00:00        0.655778       -0.008841        0.655378   \n",
       "2024-08-19 22:00:00+00:00       -0.483603        0.655767       -0.008818   \n",
       "\n",
       "                           returns_lag_19  returns_lag_20  returns_lag_21  \\\n",
       "Datetime                                                                    \n",
       "2024-01-17 00:00:00+00:00       -0.196538       -0.102847       -0.384336   \n",
       "2024-01-17 01:00:00+00:00        0.272736       -0.196606       -0.102719   \n",
       "2024-01-17 02:00:00+00:00       -0.102777        0.272660       -0.196469   \n",
       "2024-01-17 03:00:00+00:00       -1.040751       -0.102847        0.272749   \n",
       "2024-01-17 04:00:00+00:00       -1.414057       -1.040807       -0.102719   \n",
       "...                                   ...             ...             ...   \n",
       "2024-08-19 18:00:00+00:00       -0.293231        0.085817        0.180787   \n",
       "2024-08-19 19:00:00+00:00        0.560067       -0.293298        0.085925   \n",
       "2024-08-19 20:00:00+00:00       -0.008860        0.559988       -0.293151   \n",
       "2024-08-19 21:00:00+00:00       -0.103763       -0.008931        0.560047   \n",
       "2024-08-19 22:00:00+00:00        0.655337       -0.103833       -0.008813   \n",
       "\n",
       "                           returns_lag_22  returns_lag_23  returns_lag_24  \n",
       "Datetime                                                                   \n",
       "2024-01-17 00:00:00+00:00       -0.290586        0.554476       -1.982957  \n",
       "2024-01-17 01:00:00+00:00       -0.384292       -0.290667        0.554202  \n",
       "2024-01-17 02:00:00+00:00       -0.102680       -0.384361       -0.291074  \n",
       "2024-01-17 03:00:00+00:00       -0.196428       -0.102783       -0.384783  \n",
       "2024-01-17 04:00:00+00:00        0.272782       -0.196520       -0.103161  \n",
       "...                                   ...             ...             ...  \n",
       "2024-08-19 18:00:00+00:00        0.275559        0.654006        0.653235  \n",
       "2024-08-19 19:00:00+00:00        0.180821        0.275410        0.653748  \n",
       "2024-08-19 20:00:00+00:00        0.085961        0.180684        0.275092  \n",
       "2024-08-19 21:00:00+00:00       -0.293108        0.085835        0.180351  \n",
       "2024-08-19 22:00:00+00:00        0.560075       -0.293188        0.085487  \n",
       "\n",
       "[3663 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Datetime</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>2024-01-17 00:00:00+00:00</th>\n",
       "      <td>1.105603</td>\n",
       "      <td>-0.576364</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.617481</td>\n",
       "      <td>0.599444</td>\n",
       "      <td>0.617979</td>\n",
       "      <td>0.617230</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.413768</td>\n",
       "      <td>-1.040715</td>\n",
       "      <td>-0.102758</td>\n",
       "      <td>0.272778</td>\n",
       "      <td>-0.196538</td>\n",
       "      <td>-0.102847</td>\n",
       "      <td>-0.384336</td>\n",
       "      <td>-0.290586</td>\n",
       "      <td>0.554476</td>\n",
       "      <td>-1.982957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 01:00:00+00:00</th>\n",
       "      <td>1.120895</td>\n",
       "      <td>-0.305337</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.617481</td>\n",
       "      <td>0.605937</td>\n",
       "      <td>0.621199</td>\n",
       "      <td>0.613997</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.476470</td>\n",
       "      <td>-1.414019</td>\n",
       "      <td>-1.040727</td>\n",
       "      <td>-0.102735</td>\n",
       "      <td>0.272736</td>\n",
       "      <td>-0.196606</td>\n",
       "      <td>-0.102719</td>\n",
       "      <td>-0.384292</td>\n",
       "      <td>-0.290667</td>\n",
       "      <td>0.554202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 02:00:00+00:00</th>\n",
       "      <td>1.109001</td>\n",
       "      <td>-0.186997</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.614249</td>\n",
       "      <td>0.615680</td>\n",
       "      <td>0.621199</td>\n",
       "      <td>0.607533</td>\n",
       "      <td>...</td>\n",
       "      <td>0.833662</td>\n",
       "      <td>-0.476701</td>\n",
       "      <td>-1.414031</td>\n",
       "      <td>-1.040708</td>\n",
       "      <td>-0.102777</td>\n",
       "      <td>0.272660</td>\n",
       "      <td>-0.196469</td>\n",
       "      <td>-0.102680</td>\n",
       "      <td>-0.384361</td>\n",
       "      <td>-0.291074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 03:00:00+00:00</th>\n",
       "      <td>1.068223</td>\n",
       "      <td>-0.357089</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.604557</td>\n",
       "      <td>0.589711</td>\n",
       "      <td>0.598687</td>\n",
       "      <td>0.584934</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.476643</td>\n",
       "      <td>0.833458</td>\n",
       "      <td>-0.476713</td>\n",
       "      <td>-1.414014</td>\n",
       "      <td>-1.040751</td>\n",
       "      <td>-0.102847</td>\n",
       "      <td>0.272749</td>\n",
       "      <td>-0.196428</td>\n",
       "      <td>-0.102783</td>\n",
       "      <td>-0.384783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-01-17 04:00:00+00:00</th>\n",
       "      <td>1.079267</td>\n",
       "      <td>-0.553350</td>\n",
       "      <td>1.956434</td>\n",
       "      <td>0.278652</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.581963</td>\n",
       "      <td>0.579983</td>\n",
       "      <td>0.598687</td>\n",
       "      <td>0.594614</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.037486</td>\n",
       "      <td>-0.476874</td>\n",
       "      <td>0.833447</td>\n",
       "      <td>-0.476692</td>\n",
       "      <td>-1.414057</td>\n",
       "      <td>-1.040807</td>\n",
       "      <td>-0.102719</td>\n",
       "      <td>0.272782</td>\n",
       "      <td>-0.196520</td>\n",
       "      <td>-0.103161</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>2024-08-19 18:00:00+00:00</th>\n",
       "      <td>5.448477</td>\n",
       "      <td>-0.246809</td>\n",
       "      <td>4.765560</td>\n",
       "      <td>-1.096674</td>\n",
       "      <td>-0.918821</td>\n",
       "      <td>0.339483</td>\n",
       "      <td>1.163983</td>\n",
       "      <td>1.164930</td>\n",
       "      <td>1.175263</td>\n",
       "      <td>1.173837</td>\n",
       "      <td>...</td>\n",
       "      <td>0.655571</td>\n",
       "      <td>-0.103732</td>\n",
       "      <td>-0.008841</td>\n",
       "      <td>0.560109</td>\n",
       "      <td>-0.293231</td>\n",
       "      <td>0.085817</td>\n",
       "      <td>0.180787</td>\n",
       "      <td>0.275559</td>\n",
       "      <td>0.654006</td>\n",
       "      <td>0.653235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 19:00:00+00:00</th>\n",
       "      <td>5.457820</td>\n",
       "      <td>-0.439799</td>\n",
       "      <td>4.783225</td>\n",
       "      <td>-0.741925</td>\n",
       "      <td>-0.941987</td>\n",
       "      <td>0.307772</td>\n",
       "      <td>1.170690</td>\n",
       "      <td>1.168302</td>\n",
       "      <td>1.181944</td>\n",
       "      <td>1.177193</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008608</td>\n",
       "      <td>0.655363</td>\n",
       "      <td>-0.103743</td>\n",
       "      <td>-0.008818</td>\n",
       "      <td>0.560067</td>\n",
       "      <td>-0.293298</td>\n",
       "      <td>0.085925</td>\n",
       "      <td>0.180821</td>\n",
       "      <td>0.275410</td>\n",
       "      <td>0.653748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 20:00:00+00:00</th>\n",
       "      <td>5.446778</td>\n",
       "      <td>-0.667835</td>\n",
       "      <td>4.828990</td>\n",
       "      <td>-0.509926</td>\n",
       "      <td>-0.258942</td>\n",
       "      <td>0.813471</td>\n",
       "      <td>1.177402</td>\n",
       "      <td>1.168302</td>\n",
       "      <td>1.188627</td>\n",
       "      <td>1.183908</td>\n",
       "      <td>...</td>\n",
       "      <td>0.655986</td>\n",
       "      <td>-0.008830</td>\n",
       "      <td>0.655353</td>\n",
       "      <td>-0.103721</td>\n",
       "      <td>-0.008860</td>\n",
       "      <td>0.559988</td>\n",
       "      <td>-0.293151</td>\n",
       "      <td>0.085961</td>\n",
       "      <td>0.180684</td>\n",
       "      <td>0.275092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 21:00:00+00:00</th>\n",
       "      <td>5.446778</td>\n",
       "      <td>-0.667835</td>\n",
       "      <td>4.828990</td>\n",
       "      <td>-0.509926</td>\n",
       "      <td>-0.258942</td>\n",
       "      <td>0.813471</td>\n",
       "      <td>1.174046</td>\n",
       "      <td>1.175042</td>\n",
       "      <td>1.181944</td>\n",
       "      <td>1.177193</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.483371</td>\n",
       "      <td>0.655778</td>\n",
       "      <td>-0.008841</td>\n",
       "      <td>0.655378</td>\n",
       "      <td>-0.103763</td>\n",
       "      <td>-0.008931</td>\n",
       "      <td>0.560047</td>\n",
       "      <td>-0.293108</td>\n",
       "      <td>0.085835</td>\n",
       "      <td>0.180351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2024-08-19 22:00:00+00:00</th>\n",
       "      <td>5.445927</td>\n",
       "      <td>-0.872157</td>\n",
       "      <td>4.828990</td>\n",
       "      <td>-0.509926</td>\n",
       "      <td>-0.258942</td>\n",
       "      <td>0.813471</td>\n",
       "      <td>1.187472</td>\n",
       "      <td>1.171670</td>\n",
       "      <td>1.191971</td>\n",
       "      <td>1.180549</td>\n",
       "      <td>...</td>\n",
       "      <td>0.086292</td>\n",
       "      <td>-0.483603</td>\n",
       "      <td>0.655767</td>\n",
       "      <td>-0.008818</td>\n",
       "      <td>0.655337</td>\n",
       "      <td>-0.103833</td>\n",
       "      <td>-0.008813</td>\n",
       "      <td>0.560075</td>\n",
       "      <td>-0.293188</td>\n",
       "      <td>0.085487</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3663 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 27
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:45:07.570762Z",
     "start_time": "2024-08-24T06:45:07.351758Z"
    }
   },
   "cell_type": "code",
   "source": "test_standardized.describe()",
   "id": "aa05d42551e21b43",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "        gold_price  gold_volume    snp_price   snp_volume  snp_volatility  \\\n",
       "count  3663.000000  3663.000000  3663.000000  3663.000000     3663.000000   \n",
       "mean      3.200951     0.120640     3.543728     0.000627       -0.257474   \n",
       "std       1.284181     1.093883     0.728563     0.664022        0.797679   \n",
       "min       0.836298    -0.872157     1.812557    -2.556520       -1.214839   \n",
       "25%       2.210436    -0.537462     2.993922    -0.252510       -0.768557   \n",
       "50%       3.698414    -0.249964     3.504260     0.002684       -0.426632   \n",
       "75%       4.078160     0.320595     4.218084     0.221289       -0.063816   \n",
       "max       5.479909    12.299163     5.033739     5.022257        7.468390   \n",
       "\n",
       "       snp_returns         Open         High          Low        Close  ...  \\\n",
       "count  3663.000000  3663.000000  3663.000000  3663.000000  3663.000000  ...   \n",
       "mean      0.009429     0.450739     0.444085     0.456196     0.450728  ...   \n",
       "std       0.909800     0.225644     0.226104     0.225226     0.226020  ...   \n",
       "min     -11.958683    -0.114042    -0.128773    -0.119296    -0.117427  ...   \n",
       "25%      -0.349484     0.281881     0.281573     0.287102     0.284760  ...   \n",
       "50%       0.047264     0.475919     0.466975     0.483460     0.475656  ...   \n",
       "75%       0.449283     0.602941     0.596200     0.605115     0.604304  ...   \n",
       "max       6.029409     1.187472     1.175042     1.191971     1.183908  ...   \n",
       "\n",
       "       returns_lag_15  returns_lag_16  returns_lag_17  returns_lag_18  \\\n",
       "count     3663.000000     3663.000000     3663.000000     3663.000000   \n",
       "mean        -0.005646       -0.006175       -0.006082       -0.006164   \n",
       "std          0.608214        0.608466        0.608417        0.608338   \n",
       "min         -6.198638       -6.198990       -6.199004       -6.199006   \n",
       "25%         -0.285327       -0.285590       -0.285566       -0.285544   \n",
       "50%         -0.008608       -0.008830       -0.008841       -0.008818   \n",
       "75%          0.269055        0.268839        0.268828        0.268852   \n",
       "max          5.096859        5.096744        5.096736        5.096780   \n",
       "\n",
       "       returns_lag_19  returns_lag_20  returns_lag_21  returns_lag_22  \\\n",
       "count     3663.000000     3663.000000     3663.000000     3663.000000   \n",
       "mean        -0.006257       -0.006535       -0.006494       -0.006534   \n",
       "std          0.608347        0.608242        0.608209        0.608217   \n",
       "min         -6.199053       -6.199035       -6.198276       -6.198134   \n",
       "25%         -0.285586       -0.285653       -0.285543       -0.285584   \n",
       "50%         -0.008860       -0.008931       -0.008813       -0.008775   \n",
       "75%          0.268810        0.268735        0.268824        0.268857   \n",
       "max          5.096741        5.096596        5.096186        5.096137   \n",
       "\n",
       "       returns_lag_23  returns_lag_24  \n",
       "count     3663.000000     3663.000000  \n",
       "mean        -0.006649       -0.007473  \n",
       "std          0.608142        0.609096  \n",
       "min         -6.197501       -6.198844  \n",
       "25%         -0.285665       -0.286072  \n",
       "50%         -0.008890       -0.009253  \n",
       "75%          0.268709        0.268390  \n",
       "max          5.095407        5.095852  \n",
       "\n",
       "[8 rows x 277 columns]"
      ],
      "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>gold_price</th>\n",
       "      <th>gold_volume</th>\n",
       "      <th>snp_price</th>\n",
       "      <th>snp_volume</th>\n",
       "      <th>snp_volatility</th>\n",
       "      <th>snp_returns</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>...</th>\n",
       "      <th>returns_lag_15</th>\n",
       "      <th>returns_lag_16</th>\n",
       "      <th>returns_lag_17</th>\n",
       "      <th>returns_lag_18</th>\n",
       "      <th>returns_lag_19</th>\n",
       "      <th>returns_lag_20</th>\n",
       "      <th>returns_lag_21</th>\n",
       "      <th>returns_lag_22</th>\n",
       "      <th>returns_lag_23</th>\n",
       "      <th>returns_lag_24</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "      <td>3663.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>3.200951</td>\n",
       "      <td>0.120640</td>\n",
       "      <td>3.543728</td>\n",
       "      <td>0.000627</td>\n",
       "      <td>-0.257474</td>\n",
       "      <td>0.009429</td>\n",
       "      <td>0.450739</td>\n",
       "      <td>0.444085</td>\n",
       "      <td>0.456196</td>\n",
       "      <td>0.450728</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.005646</td>\n",
       "      <td>-0.006175</td>\n",
       "      <td>-0.006082</td>\n",
       "      <td>-0.006164</td>\n",
       "      <td>-0.006257</td>\n",
       "      <td>-0.006535</td>\n",
       "      <td>-0.006494</td>\n",
       "      <td>-0.006534</td>\n",
       "      <td>-0.006649</td>\n",
       "      <td>-0.007473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.284181</td>\n",
       "      <td>1.093883</td>\n",
       "      <td>0.728563</td>\n",
       "      <td>0.664022</td>\n",
       "      <td>0.797679</td>\n",
       "      <td>0.909800</td>\n",
       "      <td>0.225644</td>\n",
       "      <td>0.226104</td>\n",
       "      <td>0.225226</td>\n",
       "      <td>0.226020</td>\n",
       "      <td>...</td>\n",
       "      <td>0.608214</td>\n",
       "      <td>0.608466</td>\n",
       "      <td>0.608417</td>\n",
       "      <td>0.608338</td>\n",
       "      <td>0.608347</td>\n",
       "      <td>0.608242</td>\n",
       "      <td>0.608209</td>\n",
       "      <td>0.608217</td>\n",
       "      <td>0.608142</td>\n",
       "      <td>0.609096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.836298</td>\n",
       "      <td>-0.872157</td>\n",
       "      <td>1.812557</td>\n",
       "      <td>-2.556520</td>\n",
       "      <td>-1.214839</td>\n",
       "      <td>-11.958683</td>\n",
       "      <td>-0.114042</td>\n",
       "      <td>-0.128773</td>\n",
       "      <td>-0.119296</td>\n",
       "      <td>-0.117427</td>\n",
       "      <td>...</td>\n",
       "      <td>-6.198638</td>\n",
       "      <td>-6.198990</td>\n",
       "      <td>-6.199004</td>\n",
       "      <td>-6.199006</td>\n",
       "      <td>-6.199053</td>\n",
       "      <td>-6.199035</td>\n",
       "      <td>-6.198276</td>\n",
       "      <td>-6.198134</td>\n",
       "      <td>-6.197501</td>\n",
       "      <td>-6.198844</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.210436</td>\n",
       "      <td>-0.537462</td>\n",
       "      <td>2.993922</td>\n",
       "      <td>-0.252510</td>\n",
       "      <td>-0.768557</td>\n",
       "      <td>-0.349484</td>\n",
       "      <td>0.281881</td>\n",
       "      <td>0.281573</td>\n",
       "      <td>0.287102</td>\n",
       "      <td>0.284760</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.285327</td>\n",
       "      <td>-0.285590</td>\n",
       "      <td>-0.285566</td>\n",
       "      <td>-0.285544</td>\n",
       "      <td>-0.285586</td>\n",
       "      <td>-0.285653</td>\n",
       "      <td>-0.285543</td>\n",
       "      <td>-0.285584</td>\n",
       "      <td>-0.285665</td>\n",
       "      <td>-0.286072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.698414</td>\n",
       "      <td>-0.249964</td>\n",
       "      <td>3.504260</td>\n",
       "      <td>0.002684</td>\n",
       "      <td>-0.426632</td>\n",
       "      <td>0.047264</td>\n",
       "      <td>0.475919</td>\n",
       "      <td>0.466975</td>\n",
       "      <td>0.483460</td>\n",
       "      <td>0.475656</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008608</td>\n",
       "      <td>-0.008830</td>\n",
       "      <td>-0.008841</td>\n",
       "      <td>-0.008818</td>\n",
       "      <td>-0.008860</td>\n",
       "      <td>-0.008931</td>\n",
       "      <td>-0.008813</td>\n",
       "      <td>-0.008775</td>\n",
       "      <td>-0.008890</td>\n",
       "      <td>-0.009253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>4.078160</td>\n",
       "      <td>0.320595</td>\n",
       "      <td>4.218084</td>\n",
       "      <td>0.221289</td>\n",
       "      <td>-0.063816</td>\n",
       "      <td>0.449283</td>\n",
       "      <td>0.602941</td>\n",
       "      <td>0.596200</td>\n",
       "      <td>0.605115</td>\n",
       "      <td>0.604304</td>\n",
       "      <td>...</td>\n",
       "      <td>0.269055</td>\n",
       "      <td>0.268839</td>\n",
       "      <td>0.268828</td>\n",
       "      <td>0.268852</td>\n",
       "      <td>0.268810</td>\n",
       "      <td>0.268735</td>\n",
       "      <td>0.268824</td>\n",
       "      <td>0.268857</td>\n",
       "      <td>0.268709</td>\n",
       "      <td>0.268390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.479909</td>\n",
       "      <td>12.299163</td>\n",
       "      <td>5.033739</td>\n",
       "      <td>5.022257</td>\n",
       "      <td>7.468390</td>\n",
       "      <td>6.029409</td>\n",
       "      <td>1.187472</td>\n",
       "      <td>1.175042</td>\n",
       "      <td>1.191971</td>\n",
       "      <td>1.183908</td>\n",
       "      <td>...</td>\n",
       "      <td>5.096859</td>\n",
       "      <td>5.096744</td>\n",
       "      <td>5.096736</td>\n",
       "      <td>5.096780</td>\n",
       "      <td>5.096741</td>\n",
       "      <td>5.096596</td>\n",
       "      <td>5.096186</td>\n",
       "      <td>5.096137</td>\n",
       "      <td>5.095407</td>\n",
       "      <td>5.095852</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 277 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 28
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:45:16.325685Z",
     "start_time": "2024-08-24T06:45:16.171748Z"
    }
   },
   "cell_type": "code",
   "source": "model.evaluate(test_standardized[cols], test['dir'])",
   "id": "51a53db207a8c98b",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m115/115\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 700us/step - accuracy: 0.4692 - loss: 0.7019\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.7091071009635925, 0.4641004502773285]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 29
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:45:26.422085Z",
     "start_time": "2024-08-24T06:45:26.211662Z"
    }
   },
   "cell_type": "code",
   "source": [
    "pred = model.predict(test_standardized[cols])\n",
    "pred"
   ],
   "id": "fa91ef684b900a5e",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m115/115\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 905us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0.5185485 ],\n",
       "       [0.52787507],\n",
       "       [0.5073915 ],\n",
       "       ...,\n",
       "       [0.5426675 ],\n",
       "       [0.54983366],\n",
       "       [0.5511843 ]], dtype=float32)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 30
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:45:34.784260Z",
     "start_time": "2024-08-24T06:45:34.631497Z"
    }
   },
   "cell_type": "code",
   "source": "test['probability'] = model.predict(test_standardized[cols])",
   "id": "4a140173f5e29e02",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[1m115/115\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 576us/step\n"
     ]
    }
   ],
   "execution_count": 31
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:45:43.449016Z",
     "start_time": "2024-08-24T06:45:43.331726Z"
    }
   },
   "cell_type": "code",
   "source": [
    "plt.hist(test['probability'], bins=100)\n",
    "plt.show()"
   ],
   "id": "db3a8deb486f68c8",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 32
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:45:59.124142Z",
     "start_time": "2024-08-24T06:45:59.011948Z"
    }
   },
   "cell_type": "code",
   "source": [
    "plt.hist(pred, bins=100)\n",
    "plt.show()"
   ],
   "id": "37d7c0d794fc30ef",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 33
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:46:06.778035Z",
     "start_time": "2024-08-24T06:46:06.771749Z"
    }
   },
   "cell_type": "code",
   "source": "threshold = 0.55",
   "id": "7e79687ff4cba24b",
   "outputs": [],
   "execution_count": 34
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:46:13.558809Z",
     "start_time": "2024-08-24T06:46:13.554294Z"
    }
   },
   "cell_type": "code",
   "source": [
    "test['position'] = np.where(test.probability < (1-threshold), -1, np.nan)\n",
    "test['position'] = np.where(test.probability > threshold, 1, test.position)"
   ],
   "id": "399f1f0204cefc96",
   "outputs": [],
   "execution_count": 35
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:46:29.014598Z",
     "start_time": "2024-08-24T06:46:29.001806Z"
    }
   },
   "cell_type": "code",
   "source": [
    "#test.index = test.index.tz_localize('utc')\n",
    "test['NYTime'] = test.index.tz_convert(\"America/New_York\")\n",
    "test['hour'] = test.NYTime.dt.hour"
   ],
   "id": "e5244dfa8415afe2",
   "outputs": [],
   "execution_count": 37
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:46:35.533779Z",
     "start_time": "2024-08-24T06:46:35.523315Z"
    }
   },
   "cell_type": "code",
   "source": "test['position'] = np.where(~test.hour.between(2,12), 0, test.position)",
   "id": "2fa290f9d3c8d4f6",
   "outputs": [],
   "execution_count": 38
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:46:41.911519Z",
     "start_time": "2024-08-24T06:46:41.902676Z"
    }
   },
   "cell_type": "code",
   "source": "test.position.value_counts(dropna=False)",
   "id": "b85c0892c0090e88",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "position\n",
       " 0.0    1969\n",
       " NaN    1042\n",
       " 1.0     639\n",
       "-1.0      13\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 39
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:46:55.957371Z",
     "start_time": "2024-08-24T06:46:55.951671Z"
    }
   },
   "cell_type": "code",
   "source": "test['position'] = test.position.ffill()",
   "id": "4f3d924945a479f4",
   "outputs": [],
   "execution_count": 40
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:47:01.144846Z",
     "start_time": "2024-08-24T06:47:01.137658Z"
    }
   },
   "cell_type": "code",
   "source": "test.position.value_counts(dropna=False)",
   "id": "b86cc57bf2e50f83",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "position\n",
       " 0.0    2350\n",
       " 1.0    1279\n",
       "-1.0      34\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 41
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:47:15.622794Z",
     "start_time": "2024-08-24T06:47:15.614926Z"
    }
   },
   "cell_type": "code",
   "source": "test['strategy'] = test['position'] * test['returns']",
   "id": "c4c77ed6a040eb54",
   "outputs": [],
   "execution_count": 42
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:47:21.387053Z",
     "start_time": "2024-08-24T06:47:21.376963Z"
    }
   },
   "cell_type": "code",
   "source": [
    "test['creturns'] = test['returns'].cumsum().apply(np.exp)\n",
    "test['cstrategy'] = test['strategy'].cumsum().apply(np.exp)"
   ],
   "id": "a29c0e7477170d34",
   "outputs": [],
   "execution_count": 43
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:47:28.065022Z",
     "start_time": "2024-08-24T06:47:27.916508Z"
    }
   },
   "cell_type": "code",
   "source": [
    "test[['creturns', 'cstrategy']].plot()\n",
    "plt.show()"
   ],
   "id": "e7af252623256524",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 44
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:48:59.307828Z",
     "start_time": "2024-08-24T06:48:59.302017Z"
    }
   },
   "cell_type": "code",
   "source": "ptc = 0.000059",
   "id": "1407c2168acde302",
   "outputs": [],
   "execution_count": 46
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:49:08.854199Z",
     "start_time": "2024-08-24T06:49:08.841902Z"
    }
   },
   "cell_type": "code",
   "source": [
    "test['trades'] = test.position.diff().abs()\n",
    "test.trades.value_counts()"
   ],
   "id": "7a9d8bed2149b1d1",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "trades\n",
       "0.0    3389\n",
       "1.0     264\n",
       "2.0       9\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 47
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:49:17.293813Z",
     "start_time": "2024-08-24T06:49:17.289374Z"
    }
   },
   "cell_type": "code",
   "source": "test['strategy_net'] = test.strategy - test.trades * ptc",
   "id": "b20a43f6becc7e74",
   "outputs": [],
   "execution_count": 48
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:49:23.073354Z",
     "start_time": "2024-08-24T06:49:23.067591Z"
    }
   },
   "cell_type": "code",
   "source": "test['cstrategy_net'] = test['strategy_net'].cumsum().apply(np.exp)",
   "id": "c2e44dea27459331",
   "outputs": [],
   "execution_count": 49
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:49:29.387161Z",
     "start_time": "2024-08-24T06:49:29.203259Z"
    }
   },
   "cell_type": "code",
   "source": [
    "test[['creturns', 'cstrategy', 'cstrategy_net']].plot(figsize=(12, 8))\n",
    "plt.show()"
   ],
   "id": "8cc6af69422e7728",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 50
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:50:13.738152Z",
     "start_time": "2024-08-24T06:50:13.702243Z"
    }
   },
   "cell_type": "code",
   "source": "model.save('DNN_modelv2.keras')",
   "id": "4c761e7916b51440",
   "outputs": [],
   "execution_count": 54
  },
  {
   "metadata": {},
   "cell_type": "code",
   "outputs": [],
   "execution_count": null,
   "source": "",
   "id": "db2f331ca8b307e1"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:50:38.108564Z",
     "start_time": "2024-08-24T06:50:38.101845Z"
    }
   },
   "cell_type": "code",
   "source": "import pickle",
   "id": "9a80b929abc29300",
   "outputs": [],
   "execution_count": 55
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:50:42.613485Z",
     "start_time": "2024-08-24T06:50:42.602258Z"
    }
   },
   "cell_type": "code",
   "source": "params = {'mu': mu, 'std': std}",
   "id": "c515ce6bdbd7f19d",
   "outputs": [],
   "execution_count": 56
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:50:50.868925Z",
     "start_time": "2024-08-24T06:50:50.862123Z"
    }
   },
   "cell_type": "code",
   "source": "pickle.dump(params, open('paramsv2.pkl', 'wb'))",
   "id": "c9bc97b4c1253fb7",
   "outputs": [],
   "execution_count": 57
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-08-24T06:52:13.008500Z",
     "start_time": "2024-08-24T06:52:12.992037Z"
    }
   },
   "cell_type": "code",
   "source": "test[['strategy_net']].mean() * 252",
   "id": "4fea312a935af3f3",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "strategy_net    0.001146\n",
       "dtype: float64"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 59
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}