Predict Future Price - Binary Option of USDCAD - V6.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting Financial Time Series Data with Machine Learning\n",
    "\n",
    "This is a sample that predicts future price behaviour from past price movements.\n",
    "Here we implement it with USD/CAD rate as an example, but you can convert it to other market prices by getting other csv files.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "from sklearn import ensemble"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Structuring data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "h = pd.read_csv('USDCAD_Candlestick_1_Hour_BID_31.12.2015-31.05.2018.csv')\n",
    "o = pd.read_csv('BRENT.CMDUSD_Candlestick_1_Hour_BID_31.12.2014-31.05.2018.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gmt time</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>01.01.2016 00:00:00.000</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>01.01.2016 01:00:00.000</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>01.01.2016 02:00:00.000</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>01.01.2016 03:00:00.000</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>01.01.2016 04:00:00.000</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Gmt time     Open     High      Low    Close  Volume\n",
       "0  01.01.2016 00:00:00.000  1.38326  1.38326  1.38326  1.38326     0.0\n",
       "1  01.01.2016 01:00:00.000  1.38326  1.38326  1.38326  1.38326     0.0\n",
       "2  01.01.2016 02:00:00.000  1.38326  1.38326  1.38326  1.38326     0.0\n",
       "3  01.01.2016 03:00:00.000  1.38326  1.38326  1.38326  1.38326     0.0\n",
       "4  01.01.2016 04:00:00.000  1.38326  1.38326  1.38326  1.38326     0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gmt time</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15116</th>\n",
       "      <td>31.05.2018 19:00:00.000</td>\n",
       "      <td>1.29566</td>\n",
       "      <td>1.29671</td>\n",
       "      <td>1.29541</td>\n",
       "      <td>1.29666</td>\n",
       "      <td>5.330320e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15117</th>\n",
       "      <td>31.05.2018 20:00:00.000</td>\n",
       "      <td>1.29668</td>\n",
       "      <td>1.29682</td>\n",
       "      <td>1.29502</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>3.116830e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15118</th>\n",
       "      <td>31.05.2018 21:00:00.000</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29577</td>\n",
       "      <td>1.29498</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>1.456650e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15119</th>\n",
       "      <td>31.05.2018 22:00:00.000</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29407</td>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.370330e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15120</th>\n",
       "      <td>31.05.2018 23:00:00.000</td>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.29469</td>\n",
       "      <td>1.29406</td>\n",
       "      <td>1.29436</td>\n",
       "      <td>1.259850e+09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Gmt time     Open     High      Low    Close  \\\n",
       "15116  31.05.2018 19:00:00.000  1.29566  1.29671  1.29541  1.29666   \n",
       "15117  31.05.2018 20:00:00.000  1.29668  1.29682  1.29502  1.29563   \n",
       "15118  31.05.2018 21:00:00.000  1.29563  1.29577  1.29498  1.29564   \n",
       "15119  31.05.2018 22:00:00.000  1.29563  1.29563  1.29407  1.29409   \n",
       "15120  31.05.2018 23:00:00.000  1.29409  1.29469  1.29406  1.29436   \n",
       "\n",
       "             Volume  \n",
       "15116  5.330320e+09  \n",
       "15117  3.116830e+09  \n",
       "15118  1.456650e+09  \n",
       "15119  1.370330e+09  \n",
       "15120  1.259850e+09  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Put USD/CAD hour data into Pandas Dataframe\n",
    "\n",
    "df = pd.DataFrame(h)\n",
    "display(df.head())\n",
    "display(df.tail())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gmt time</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>01.01.2015 00:00:00.000</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>01.01.2015 01:00:00.000</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>01.01.2015 02:00:00.000</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>01.01.2015 03:00:00.000</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>01.01.2015 04:00:00.000</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>56.83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Gmt time   Open   High    Low  Close  Volume\n",
       "0  01.01.2015 00:00:00.000  56.83  56.83  56.83  56.83     0.0\n",
       "1  01.01.2015 01:00:00.000  56.83  56.83  56.83  56.83     0.0\n",
       "2  01.01.2015 02:00:00.000  56.83  56.83  56.83  56.83     0.0\n",
       "3  01.01.2015 03:00:00.000  56.83  56.83  56.83  56.83     0.0\n",
       "4  01.01.2015 04:00:00.000  56.83  56.83  56.83  56.83     0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gmt time</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>21380</th>\n",
       "      <td>31.05.2018 19:00:00.000</td>\n",
       "      <td>77.835</td>\n",
       "      <td>77.865</td>\n",
       "      <td>77.625</td>\n",
       "      <td>77.655</td>\n",
       "      <td>102294.9964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21381</th>\n",
       "      <td>31.05.2018 20:00:00.000</td>\n",
       "      <td>77.655</td>\n",
       "      <td>77.788</td>\n",
       "      <td>77.535</td>\n",
       "      <td>77.725</td>\n",
       "      <td>125650.0036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21382</th>\n",
       "      <td>31.05.2018 21:00:00.000</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21383</th>\n",
       "      <td>31.05.2018 22:00:00.000</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21384</th>\n",
       "      <td>31.05.2018 23:00:00.000</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>77.725</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Gmt time    Open    High     Low   Close       Volume\n",
       "21380  31.05.2018 19:00:00.000  77.835  77.865  77.625  77.655  102294.9964\n",
       "21381  31.05.2018 20:00:00.000  77.655  77.788  77.535  77.725  125650.0036\n",
       "21382  31.05.2018 21:00:00.000  77.725  77.725  77.725  77.725       0.0000\n",
       "21383  31.05.2018 22:00:00.000  77.725  77.725  77.725  77.725       0.0000\n",
       "21384  31.05.2018 23:00:00.000  77.725  77.725  77.725  77.725       0.0000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Oil data\n",
    "\n",
    "df_oil = pd.DataFrame(o)\n",
    "display(df_oil.head())\n",
    "display(df_oil.tail())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the date as datetime\n",
    "df['datetime'] = pd.to_datetime(df['Gmt time'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.set_index(df['datetime'])\n",
    "df.index.names = [None]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set datetime for oil\n",
    "\n",
    "df_oil['datetime'] = pd.to_datetime(df_oil['Gmt time'])\n",
    "df_oil = df_oil.set_index(df_oil['datetime'])\n",
    "df_oil.index.names = [None]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-31 19:00:00</th>\n",
       "      <td>1.29566</td>\n",
       "      <td>1.29671</td>\n",
       "      <td>1.29541</td>\n",
       "      <td>1.29666</td>\n",
       "      <td>5.330320e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 20:00:00</th>\n",
       "      <td>1.29668</td>\n",
       "      <td>1.29682</td>\n",
       "      <td>1.29502</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>3.116830e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 21:00:00</th>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29577</td>\n",
       "      <td>1.29498</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>1.456650e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 22:00:00</th>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29407</td>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.370330e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 23:00:00</th>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.29469</td>\n",
       "      <td>1.29406</td>\n",
       "      <td>1.29436</td>\n",
       "      <td>1.259850e+09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     Open_oil  High_oil  Low_oil  Close_oil    Volume_oil\n",
       "2018-05-31 19:00:00   1.29566   1.29671  1.29541    1.29666  5.330320e+09\n",
       "2018-05-31 20:00:00   1.29668   1.29682  1.29502    1.29563  3.116830e+09\n",
       "2018-05-31 21:00:00   1.29563   1.29577  1.29498    1.29564  1.456650e+09\n",
       "2018-05-31 22:00:00   1.29563   1.29563  1.29407    1.29409  1.370330e+09\n",
       "2018-05-31 23:00:00   1.29409   1.29469  1.29406    1.29436  1.259850e+09"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_oil = df.loc[:,['Open', 'High', 'Low', 'Close', 'Volume']]\n",
    "\n",
    "df_oil = df_oil.rename(columns={'Open': 'Open_oil', 'High': 'High_oil', 'Low':'Low_oil', 'Close':'Close_oil', 'Volume':'Volume_oil'})\n",
    "\n",
    "df_oil.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10e8d3ac8>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Just some exploration \n",
    "\n",
    "df['Close'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10b55fd30>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAD6CAYAAABDPiuvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzsfXecXFXZ//eZ2ZYtySbZTSe9hxRCSIQQktAhKlUg8gIiiCiIrz8RASkqKggWwBcEREBRUESahl5CCAGSEEIaaYSF9N6TzZY5vz/OPXfOvXPOvWfazuzO+X4++9mZe8+c+9z2POfpxBiDhYWFhUVhI5JrAiwsLCwscg8rDCwsLCwsrDCwsLCwsLDCwMLCwsICVhhYWFhYWMAKAwsLCwsLWGFgYWFhYQErDCwsLCwsYIWBhYWFhQWAolwTYIqamhrWt2/fXJNhYWFh0arw4YcfbmOM1YaNazXCoG/fvpg/f36uybCwsLBoVSCiz03GWTORhYWFhYUVBhYWFhYWVhhYWFhYWMAKAwsLCwsLWGFgYWFhYQErDCwsLCwsUMDCYM7qbdh3qCnXZFhYWFjkBQpSGGzdewhff/gDXPPkR7kmxcLCwiIvUJDCoLE5BgBYtmFPjimxsLCwyA8UpDCIEAEAmmKxHFNiYWFhkR8oSGEQFwKUUzosLCws8gUFKQyaYwwAELGywMLCwgJAgQqDJlcYWGlgYWFhARSoMNiw6yAAqxlYWFhYCBSkMLjoz3MBAGQ1AwsLCwsABSoMBKwssLCwsOBIWRgQ0SNEtIWIloSMO4qImojoXGlbMxEtdP5eSJWGVLBy8173c0OTDS21sLCwANLrdPYYgP8D8FfdACKKAvg1gFd9uw4yxsakceyUcdlf5rmfY4zlggQLCwuLvEPKmgFjbBaAHSHDvgfg3wC2pHqcTGPtjoPu5237GnJIiYWFhUX+IGs+AyLqCeAsAH9U7C4jovlE9D4RnZktGlQ498heLXk4CwsLi1aBbDqQ7wbwY8aYyjDfhzE2DsDXAdxNRANUExDRFY7QmL9169aMEDWiR/uMzGNhYWHRlpBNYTAOwD+IqA7AuQDuF1oAY2y9838NgJkAjlBNwBh7iDE2jjE2rra2NiNEiexjCwsLC4s4siYMGGP9GGN9GWN9ATwN4LuMseeIqCMRlQIAEdUAmAhgWbbo8MMKAwsLC4tEpBxNRERPApgCoIaI1gG4FUAxADDGHgj46TAADxJRDFwY3cEYazlhYCOILCwsLBKQsjBgjE1PYuw3pM9zAIxM9bjpormZC4Mjelfb2kQWFhYWDgouA1loBhUlRWBWS7CwsLAAUIDCYH7dTgC8FIUVBRYWFhYcBScMZq/e5n62ioGFhYUFR8EJAwEispqBhYWFhYPCFQaAVQ0sLCwsHBSuMLA+AwsLCwsXBSkMvtS/EwhcMdh/qCnX5FhYWFjkHAUnDDpVlGBQlyoQERav340Rt76CTbvrc02WhYWFRU5RcMKgqTmGaIQgp5stXr87Z/S0dVz1xAK8tSJvKphbWFhoUHDCoDnGUBQhT8vLrXsP5Y6gNo4Zizbi0kfnhQ+0sLDIKQpOGDTGGIqiERySWl52qijOIUVtFzbD28Ki9SCdtpetDnvqG9HQFMPf3v8c+yTHcXV5SQ6parvYc9A65y0sWgsKSjPYsoebg/YdakJFSdTdXlFSUDIxY3hp8UZMf+h97f7Pd+xvQWosLCzSQUEJAxmXT+rvfi4ustVLU8F3/r4A763ZjvrGZuV+WxXWwqL1oKCEQWlR/HSjkTijsqbt1CAuoa5hkJUFFhatBwUlDNo5pqGfnD4MxdH4qVthkAjGGObV7QgcE9Y0Tha4FhYW+Y2CEgYCZcURFEfjjCrWxqTBmq378PHaXWnN8fcPvsDXHngPLy/ZlPIc1hdjYdF6UFBvq8zz27JmcPxv3wYA1N0xLeU5Pt26DwCwbueBlOcocgSuNRdZWOQ/ClIzAJFXGLSiknV/mVOHbz8+P+vHEQKS0uDkbU3IWli0ZRSmMEDc+QkA97y+KneEJIlbX1iKV5ZubrHjWbO/hUVhoKCEgawBbNoTL073xvLWVztHF8GTKQg/SjqywCoGFhatBwUlDAQIwIUT+iT1m7U7DmS9vEJTcwznP/ge3l+zPXRsY3MsdEw6SMZMZJm+hUXrR0EKAwCorSrFeeN6GY1d8MVOTLrzLTw5d21WaVq38yA++GwHrn5iQejYloqASsf5a2sTWVi0HhSWMPDxprLieEmKCb96Xdvo5vPtvKzC3M/CV+zp4OWlPIxz276G0LFNWTYTtSanuoWFRfooCGEQ8zFOsdod0q3K3bZ5zyFMvmum8vdFEX6ZGrPMgJPxAzQ1Z1kYCDNRVo9iYWGRL0hZGBDRI0S0hYiWhIw7ioiaiOhcadslRLTK+bskVRpM8OnWfeh/44t4afHGhH1fH9/b833bPnVfgyInpKYpy3b6Cf06AQDaSRqLDtmmxRU1NrTUwqIgkI5m8BiAU4MGEFEUwK8BvCpt6wTgVgATAIwHcCsRdUyDjkAscbqYvbhkU4LhwzSGXuQkNGZ5NV5axIVAv5qK0LHZ1lIEI7ehpRYWhYGUhQFjbBaA4OI1wPcA/BuAHLt5CoDXGGM7GGM7AbyGEKGSDgTDl52ZpDF+TBzYWbld1NjJdjingIlzOOuagRtaaqWBhUUhIGs+AyLqCeAsAH/07eoJQA7LWedsywrEytbEZDFlcBf1DjFHmrSc98B7eGT2Z9r9wmlrIgyyraXEQ0tNxlp7kIVFa0c2Hch3A/gxYyzlJSwRXUFE84lo/tatW1OaQ9TUN2GwRVE15xNb02V6c+t24Of/XRY6zkQDaYplVzMQMNEL1u86qNxuZYSFRetBNoXBOAD/IKI6AOcCuJ+IzgSwHsBh0rhezrYEMMYeYoyNY4yNq62tTYkIwcxijCmZU48OZe7nIo2B3MS3sHjdbtz83JK0BIb4adAUotqqSTRROrQcauINa95eGS6E/zV/XcrHsbCwyA9kTRgwxvoxxvoyxvoCeBrAdxljzwF4BcDJRNTRcRyf7GzLCgQjn/vZDmlbfP+s66aiUwXvgVymieKJawb641zw0Ht43NdbOVUEaTFumKuBzyAoX2Hpht34bJu+LeW6nXy1/5JBCesXPt6g3C7MXlZDsLDIf6QTWvokgPcADCGidUR0GRFdSURXBv2OMbYDwG0A5jl/P3e2ZQXCnLLzQKNyf1E0gscvGw8AqChVV/Qm12eg52r7G/hKWq6GmizE7EFWoqIknNlBY6bdOxtTfzNTfxyNyUyF3QfV19bCwqL1IOV+Boyx6UmM/Ybv+yMAHkn12MnApFS1WG3rICJqTFa4yzftxZjDqhO2+xPfghDExKMOkzZxIO880IBukhksGYRdExkNTS3jv7CwsMge2nwGsui2NbZ3nEHr1rw6Zk9JRCTpGHkytYRMzEQmDuR0tJTlm/Yaj5V7S8uw5iELi9aDNi8MBGMtD2jB6PcPH2xoxvkPvufa/12fgUFwaYmGAZvwReHwDRYG5g7k4iRMPX7oGLwK1506NOXjWFhY5AfavDBoFslTFL5SFcz+f//5ET74bAcOv9XxayehGZRomGgymkGQb7jINRNl1zQzpneiqcuPKsfH0rV9qXK/VQwsLFoP2rwwEKvtiLT892sC/vXzdl8UjuszMDiezvFqIgvEkKCQUMpQAlwYIgbhtOWlPPrKmoMsLFo/2rwwEKZ1s0xa/n/+5zulbUwqzaCHCE/NBGNsqV4FQTCpSRQmJG1msoVF60HK0UStBW+u4GWRIkQprab73fCi+zlIoMTLXqiPkgxfbKkaSEHoWF4SOiadxjcWFhb5hTavGQh0qYrbtf3F10yZWnDRNlH2Qr3XxPlskoHcUovtYd15r4fTR3YLHasVgBmlyMLCIpto88LgV2eNRDRC6FwZvtINY14mmoHOxJPMYr85D8wrgt6yIn1vBasYWFi0HbR5YQBwpsVYkA3bjK0FCQMKEQZm9nPzqqXZhktC4Dm3PXHw7upt2iZHFhZtGW3eZwA4YaWeDepxYQw7yEwUcfsmaOYOnNmLIC2ipeREMj2Qteece5mWNC58+AMAQN0d03JMiYVFy6JANIPgFayxzyDQTMR3HnBqFPmRDGNMpnRFthDvgRx+cZIRHLnEkvW7cd4D7+Gg5h5ZWBQyCkIYAMJMlL35RU3/W57XtIQ2yTNwxuSFmchgTLgQzdx5MMbw9sqtaQnKW55fgrl1O7Bkw+6M0ZVNrNi0V9srwhTrdx3Ei4r+3xYWfhSGMCDv6jVZS7fItDWxka/crK7pk8zqOQ8Ug6QkZ0vIrteWbcYlj8zFI++qO8Xt3N+AvtfPwM79+rLdC77YBaD19HU+5e5ZmHjHm9r9m3bX44G3Pw00b06840189+8LAsds3H0Q9Y1WWyp0FIQwICBwkRrGG/Ym0aNAx8hNGHwuZEBYWKiJ07wlfAaiN8OnW/cp9x9x22sAgIm/1jPPOHRZ4nGC6xub8bP/LMWe+uyU596xvwEbd6e36r/6iQW446XlWLVFfU1kBOWuHH37mzjtnnfSosWi9aMwhIHhSlDwgoqS4CY3QSgr1lXwbHlWb1QCQye8YuFZ1yb+hExB1P/buLteub/cuWc6n40J5Gvxrw/X4dF363D3a6tSni8IY297DUffbiK49BCZ8v+YuzZkJHBQs/Lf6wi7oEZHFoWBghAGgHfV7Tf3+L/r6vKYmBdMookyIRiyLVtMNAP/WNPtKsxauTWwS5wo3V3drlizP5xQcS66aq4eep0LLNp/5jMWfLEzdIzOfLlL0/TJovBQEMKAQEkx4CaNSm3iM9DnGag/68a0FPR1hcJ/GzcTpUf4ys17cfEjc+NVYhV4Y/lmAMAXOw4o908f3xsA8I1j+mrnGN69PQB1Y6D7Z67GvLp4w71oEu1F/dixvwHPftRyfaFNypfUVqqbHLXBVBGLFFEYwoDMoomEk1d+uY7s0zGpY+kEiexAzgf/sECoz6AFTEGfb1czeBkvLua9mBeu3aXc384xE7XXaA4AENW0DP1s237c+fIKXPDQ++42oWg8NX8dPlizPZQ+GT98aiF+8M+P8fn2ljG9RAO0IlFSvdGgGZJFYaMwhEGS30UXsaunDkRULn0dcIw+ncsBACcM7aLcL/PcdLKUM+17yETF0XS0C8DM/HbKiK4AgKMHdA4hRn9QYf7zd4lbvnFP4liJqHc/TU4YCOGmWxhkGqKOlAri+c2H4ocW+Y2CEAaAaacxzgRjDPj+CYNw7SlDILcCDlKph3bjL+SkQbXauVWfc410aHEvR5rn07U9N2F0LNev6if040JgUBc14zPRYKKaLnG/eumTxLHSzY4maUsRwt6kJ0Q6OHtsTwDA2N567TWZzngWhY2CEAZExBm9hmvJ76xYQYmXyKuC619ut7a/1uwS354PSWUCumviZiCnUZvINLdC2OUrSs1bk+r2Bx1RZyZauyMxxFNeBKzWhLPqIAoNJitEkoVbAiVgTDRqNQMLMxSGMIAv6SzgHRWqvXiJgjqkJYNkOp2lOyYZaENLpXahoXOkSZVgVEG27zC4faoDSIm6ZqJwemW/8X8+3pAULck0VEoHJppZVGMas7DwoyCEgakPlLFgzcBkGpMwy3zSDHQwydf2M+AXF2/EUqnUQ9hprt1xAGff/66bG2BiVgnzZQQJprhmEM4YTUqea2nIo/ur04YsLPwoiKqlQHA0kWxvdjUDx04QNdQMwrJx5Zo6eRVamg4tvuvx3b8vAGBe8XPSnW8BAB6ezUtMBF7fMFIMJLVgjKrQUj/SYejiVuumaEohXDUIJgKwpZzZFq0XBaEZJEQP6UpYI1EziGShkE06mkGmBYbWZ5BMCWvf94v+zMtAB9EqC8el67k2kY7DVfgvAs1ESaySZeWhprJUP1AB4TPQ3edDTZkRBmELEMBqBhbmSFkYENEjRLSFiJRlOonoDCJaREQLiWg+ER0r7Wt2ti8kohdSpcEUe+qbsCGg+qPMg4RtVQiB15Ztjo8zKues2c4MxuQgAyFMMzBZrTMG/OTZxe72d1ZtQ3OMBQq9/jfGe0tvd4rLZcLhGnQFI8n4DCTaDzQ04Z1VW41pEIJOd/4NBsLARDNxgxYCxoRpBnlk0bLIMdLRDB4DcGrA/jcAjGaMjQHwTQAPS/sOMsbGOH9fTYMGY7y6bHMoq2WMJWgGMoycqQbRRG3lBZSjif7+wReefQNufBFf/sPspOYLKtcsLtnrn2wJnCN4lcz/m5hpmEcYNOOiP88FY8yoPIUQArrVuKwZ6Epybw+ovipgZBoTAjDDpimLtoeUhQFjbBaAHQH797H4G1WBPEq8DVrhB0W2ZMq8kxkHY2Yu59qdB9D3+hn48HNvfRtBI4EzkqC6QclqNDrGFHQMEX6qExjJ+AzkVfLku7jfYvJgb36Iikf/5tUVGHLTy4F0AsBOp97Pr19ertwvC5RXJc1TxlG/fD3wGADw3ML1oWOsz8DCFFn1GRDRWUS0HMAMcO1AoMwxHb1PRGdmk4Zk4dcMjpcyitN5n2RBopUFGfcHhGP2qm0AgKfmeStfinMlAq56YoGybpBJOKcKuw4mXxwtKAdBBgNDfWMz9isYttBkRGG3hqaYmy1cHPW+CirB//h7nwMAduwLX7UDei1GNhOt1dRaCrume+sbUd8YCx1rfQYWpsiqMGCMPcsYGwrgTAC3Sbv6MMbGAfg6gLuJaIDq90R0hSM05m/dam6zDaAncP+2fQ1SNBF/iR75xlFY9vNT0LV9aeDvw1am+RpaqrOjy20vX1mqXr0mY+KXr10qvoGyIl576Oj+6nIUrrbHgPMffA8jFMKr0WHCzyzgK+pLHpnr7nv9E+85isvx7HePcbftqecCpr6p2ajj2qAulcrtsploSDd1RvVhndoFzi0f3kYTWWQCLRJN5JiU+hNRjfN9vfN/DYCZAI7Q/O4hxtg4xti42lp1mQdTTBwYZyI6XvTrl5dLmkH80pSXFKFLVZnR6sokVDMdxSDTTmZhDVu1xVviWHUcXTcsE4rkS9ecgjAUv9BVERX3dMPueny8Tt3WsoNTxO5/TxwEAHgvoACdyEXooCh8d/LvZ3kc4DrooqNkM1FlmVrjCYtgkqcOupxFIbkVVkhYCGRNGBDRQHL0ciIaC6AUwHYi6khEpc72GgATASzLFh0AX6G1L9PXvdm675D7WdRw8VkNEImQkZlot9YEEv9xPmkG63ZyG/wiHwNVkbh5j7exTHw1nlzcvhCqV031KoSlRfrHUfxedaQDDU2uuWuj5FOQ6/y/snQT/vUhLystKnlecVx/7fEE70wn3FUXlixrBrpLV+J/AH0wfYQanOd5Xp2658EvZyTWZbIoTKScdEZETwKYAqCGiNYBuBVAMQAwxh4AcA6Ai4moEcBBAOczxhgRDQPwIBHFwIXRHYyxrAqDaITQHNOvqWUbbtyB7H0ZI2TGxP/vrdU458he2HWgAUdIBcQ8P80fWYByjS1eRaL/9JPhk7Ig3eYI3+JoBEO7VWH5Jq6V6MpRfP1P72OOUzlUdQ+G3xI3Cckredlv8O3HP3Q/iyk6luuzjOVic18Z3SPpkhT8t+rtXmGgfhhKi6OeMUF1oIIep5JocM2sWSvTN79atA2kLAwYY9ND9v8awK8V2+cAGJnqcVPB8k17sXzTXpch+VsnlkgrUmHC8IeWRpxidzrIJqSpv5kJwJuJ6/UZqOdIp02lCnsMHLWVpeoWn6oD+bcIJp6seeus++YAAB59t86jSel8CXOkEtJh5y9P0bM62O4eFO8v51kkE/0lm7F0Tm/5uLqZZc2AsRDBG0BfqeNrObxnB/VP82llYpFTFEQGsoBwEr60eKNne59O5e7nhY5pwb9K/fDznZi9ept27jBzgieaKEMv4Jqt+7TRKABwxn3vhs6h6ynM3P9xWuWQ0Gn3xhuoJyvEGpx5rj1liGeMSbZ3eK5I/HOMcUY+Y9FG5djG5hiIuLnooYuO9OxzNYMI4c3lPCpo9GHVofR9IvVGOMcpMe2H7HzWOaJlk5lSI80QD88ji6VFjlFQwkBgv69pemfHWXfisK746X+4xUrXU3fGoo3oe/0M3PzcEo82UF7CV4FnaxiAn0kpxyT5hh//27fd+j46rN91MMHpKq90t+w95P+JQyMfs6++CdVOnwFh3qjbth9LNyQ2hAlCjDEwxjDlrji9I3q0x5WT434Do6qlCu4lGgsBXh558u/fRr8bXsRVTyzwTcFHvbRkIxgDbjx9GAb4In+aXTNRPIS2XnpuqjW9F+RMd90CwRsJpEbHivj8YQ53I81M+8xZWHAUpDBQraa7dyjzNFfRMSbBWB5//3N3xSijoiRuGpCZbqaSzpL95cQ73sR1Ty/ybPMIoxCT1XMLN7iRVWJFP8Uxg8XHxif5kbTal8MjGeMF4uqkFpdNzQzfmtTP/S4zz2Ub9mC1E+E0dUg8kkwlSOXfyavoMId/uXSv/Hdb/DZKhF+exa2a4/rGfUCdNP6GfjXqcFIZJtnosokp7HFJZ3WfT8EMFrlFQQiDSYNqcETvuIqvitTgPQ/iKIqGr1K/9df57mfxgsuherIGkmw5ikNNzfjpC0uxw1eWQJhq3l6pN1n58exH3kxVVWSPH/LWYudaNGps7PLYzhUl+NPF43DDaUPxpX7xcN4YS6xVNP9zbwK7LH9Pv/cdnPi7WQnHCmNe1QE9kF16nSk6V5ZgVC9uS/c7aN0MbCKceURPzL3xBG9AgG5uA3EdM1kYeDTJ9Bm2ZfkWYSiYEtYy9mqyU+V3zq/inz/uMLztRF5s2qO2swNeG/Ceg42odFZ4Xs0AeO/T7RhQW4EuTstH/5j/frwRj82pw/5DTbjra6Pd7aLUwZNzvbWAkoHMGPY38Gsx1Jf8JNMiBGNDcyy0Nk+ECCcN7wqgK/bWN2Lmyq3YuvcQYiyRqX1lVI/kaQ9bJZvM4fw/2NCMdk7UToJmEIubiQB47hOnQ30kkx4yHo0xhEZAreF4Fxf6s16m6O/spSVwt0UBoSA0AxP4o0aK/KGlEW67rZCib2RtQ7ybsmawpz4eKeNNOmOY/qf3Awu5HXQSvFJJ0AqDzJRFVq1f+MnMpti5Fo3NMYyQwjjPOoL7R3QkVpUV4zvCJ8C81+bio/vgMMlxz+ni/2VHtV/4hF0Nk1wQxoCbnluMDz7b4Z6/bh6TPINPNu7BdU9/DMaYkWbgXxiox8jmrtSfAbeOkuX6FiGwwsABkZfR+H0GESLEYgyVUvLaKEW4npzpuedgnNF4eyDz/37nrXz8TU6Ujz/cckBtBQBgfL9O+pMJgcwX9jqhnQk+EpVm0BTzMPSvjumRQLd/iS3IjzHm0Zrmfqaqccj3N0jCQBY+nPYQZ6oB09t1sAF/e59rViL6x8/zw5ray0c57Z538NT8dVizbb9RpnnMgNF75gnRNiybt8gErDBwwPMIJM0gqhAGjOnj8h00SgxPtsd7QisN6tlXOWUKtu3zCozuHbhT9ksGwqCjJuJFRlwz8G73+E8czeBQUwynjujmbh8r2dB1kJu2yxnBD/pCOYH4NZKvj79cQoOiHIWJ2cUzh+L6+yuRunkG0htCiv0yHnu3zmgVbyYw5M+Jo5Jd6FuBYRGGghIGPTpwu+/Ma6ck7CN4X8DPt3sjjkQWsxwtJPsewl42eb9JTXzBsN5aoc4QNemW1b9WHdkiM5e9jilL50AFJAdyM1OarVZsitul/evoiKQZiNIXANCnc0XCcQVdQcKyXXGiMJZpUmkGf7tsgue7HGorBKbfUd8cqhkwjLz1FfS9foa7rVNFiZGZKhZCr5hfNV49n3q7/JzpTXkF6Ta0UKCghEFZSRRfHtUdfWsqEvYRkYdhd6rwhg6KDOQyiRl1qSrD715bib7Xz4gnNhnYgEXp4aAxYcz+gBSp5C/XXFbMb+vw7u01x4l/Fsfxawbe8hGcUTY0xVwa5aztp+avcz8nlE1wvj/30Xrc8vxSAHoTlzik/9zlS6piXs3NMnNNnLfYp+XJykWDe/7eMXEzkZJUMJYYiHDPG6uS1ww0w/dJvowwAaMTKPs0/hAZ/tLdFoWLglkWMMbNNrr8AeFALi2K4FBTDEf28ZpAIsRXizLPiEaAe99YBUBjvlD4CQAzzUA3Rsx5UKog6jelFEciqEfMbeGZOEciZGb4+rLN+LPTpB6IN5RpbI5hb30TenVsh9d+MNnoPMTl/oVUEO07U5QVy12fwswV+m5mKsYoawaCGd99/hgc1qkc9Y3NOKpvJ1w5eQAundgXE371hsevc5zT1Mb/VIgpdTkMOrz3qb4Sqju3HAmkuBsNTTG3qB6nRWEmUtDqh5xdrhMYuuRKi8JDQQgDsVoNFAbgL1V1eTGmDO6SsD8aoQRmEBby598joNUMpM8lARU8Aa8phTPlYjDGuGBwTrGpWU2LiqnJTO9yKX/Cf8wPHMdvu5KoRyDpoDKz6GoQ6TQDzxgF7bIclnfLAv3604a6n8X1//4Jg1zB5I9sEv4emdQ/vLlaeRwZd72yQku7wB0vxTugqeZZvWWf57vqOZMTHnXPnEnbUasZWAgUhDAQaI4xLSPiZqJEBiDv9y+0k3HiyWN1fQFkLHPKPfjbMQrIq/4d+xvQpaoMF/15Lmav3oYqJ7dBF5aq2jy3bofH/q2CX/sxKR+hGqG1wzt0DdQ0hQE0moF0LUxs9i8v3QSAl7X+wUmDAQDdOnjzCITAlJ+XLwLqQCUDkSvCj5O4v3Ol10Spuo/X/utj97PJc6gbMrRbVWDvaYvCQUEtC4I0gwjxl0r3YkUjiS+lUn3X+Qykz3L+gW6QcBz7a/yL+Q9J2oWI9hGF9IQtW9uMJwkh1qUq3mRliy/ZTtX4JdGBzLfI/F9vh2dK8uThqnNq8kRtBZ8cUdyX8s1j+2nHqfIM5LpTmeljrZ4nwX8RIuHS6U8jfGPlJcFRchZtH1YYOCDEzUCqhasILZXf3WT6ysq/+8mzS4x/V6kpg1wv2et1tmxdF6tkkphkbeAvTg/gpOBcy76d4057XW1+QVUQo/WX0ti4+6C3CmjAPRQQ17RXx3ibEqpzAAAgAElEQVTtpESfQeI8PTpItZb00ycFk3nCs66Tc1rLyEZSo0XrRGEJAxbmQNa/nCKaiAHoV1OBytIiJbP1vFvS5yAGXN/YrFXVdb/asV82NahHfbx2l0SX7LQ0x8GGcJOW3JfYz4TFKle+7jrNQJzH+2u4X2JQl0oPw+Zj4p8fnPUpjr79TV8NqGCQdBySRID/uYg5wQKy4BJ1jFS454Ixgft1UDuH/b6p4LNKh58LQWplgkVhCYPAaCI5tFTh9HR+J688ddE6KgS9bFc8/iEm3vGmUtPwMwvxVa6b3xxjuOEZb2VSAJ64/kZP+KX5m69y5l58dB/3c21VqcfEkCgM+H85akXXt4Ax7gx/aNYaAPya+0mVGeO7iv4SYcqafJ9lWv3aSowlmmtkh7RM15pfnY4zxvT0hB2bwuRWhAsDA81As10TY2BRgCg8YaBzIIO/VLr3yk2ecguYkTJaR2ea0anyO/Y3uK0HTaJzlHMz4Mm5awPHeM1KKR3GxV8lcxEh2NQgLrcctSLLAn9W75CbXna/b9vXgPW7DnoS72Ta312dGMZpxBgNxsQYS9Bg5IWEuJ+TBtW4wm2rpjdEIC0GG8PzDJI+LACeES7aedqOZxYFJQwONDRrGZcwE4nPfogX/kBDM+obm0EUX20Pk5K7tKs4xeb+tRUYe9tr7nelZqCezQMTH0C9ppx2uiAK9p2I1bVc3kPvM/DO4y/FAYQ7U8MuhVyqPCgWirFEOuXvjPFkNrmd5Gfb9vsmCaYFMLt3YcLLrDhf4qC35PBUKwsKHgUhDJZv3IOFjv380XfrlGMiRHhj+RYlAxL7AeDtlVuxcXc91wwcM5HI+AV89Yik36veNb8ZQpWw5H9JVYzcxJHtyW1I88W/9uTB7mcC4Z1VcXON/5wEA/3oi7j/4sAhtQaUbCkHFdJxpnqOEwvWDATSTdlS0eLfFKoZaM559GHVGNtb36rTpptZyCgIYaBr7Sijbnt8VVeseOn95qUIxZO6QvvVwowB/XN+3NQjDif/7M6Xl7vOVRkmTFQ2Qc1cqa539J+rjw2fCN4G9fJl6VhejNNHdveMVTEcf1E4Af/qVRXuGGMMddv24w9O5nfC/hA3jlyqXKeh8OMkCjb5sWBIvKdfHuU9dx3GHFaNntXt3HnCkGptIjAWeI6emk4GdGQSb6/ciil3vWWUc2PRMigIYWCCvVIdF1WT+EQTRgNmLOb1iGQnq+7FVW0PWpl1dNoqygzy/pmfKsfKY752ZC/lGPml+/1rK5Vjiov0FP3xwrHu581SvoH8i1u+Mjwho1WVYHagQSMMfN/9ORYAd4pP+c1M/FZzDmGaAYGMGXCiMPCaiRi8wlAw+DAw8P4YfJ6QiDSon50qOeQ4yGcTMERRQaVFwBjDJY/MRd32A9juKxBokTtYYaCAShgcCAixlE0gKqcmoF55BcXCCwZgwrjkY+radcqager8AHU7UIFyifnsl8w88spTxfj9mwZ3rcTxQxPLfQDxciACAxRVV8MK+Jk2t1HRJqM5xrRhss4sfA5JHOqipBKPHw9kCFr0nzisi3YM+bQU5XFg9owFTpIFfLRWfl/M27daZBcFJwzul1a4OsiN3AVSKejledeUmkGACi+4msFLKvcS1oXOyprBGWPU7SaD6iHJZjLV9QHUZhc/Oa/+YDKqpWbyCfZxiZsfrmgeFIqw60VmfoWYIifFYyZiifdUPCNhZToYiwuOIBOQXFPLD1kDC5zDecZU5xzmjM8W5Ci8655ODIm2yA0KThjoVqUyH1MV75IZ3TXHD0TfzuUJY4KQ7GsnXtQ3lm/G0g278cQH+p7HsgPX365TQHYgD+5apRxTWqSPk1+zLV48bdKgeL0k+bodUth/g2zWgKKpjnShTPI4Xrh6In721RHu92RKSAdRpvIZ+Jm8f+UtxvtLZiccH3ETlNqBLMKXBS2Jg8YcFncMB55yACneAIeWEwzf/fuCFjuWhTlSFgZE9AgRbSEiZW0FIjqDiBYR0UIimk9Ex0r7LiGiVc7fJanSkAp0VRrld0a1spNXM988tp/H+WZSsE31rokXsLuvSBoAN6u2vjGGaffOxo3PLg4/BvQajKwZ6CKq/JrBaCmjdnj39m6yme5sVY7psCsjawmAl/Fdc/wgbU8GgVG9qj3nZqAYmBV2U+QZeEJLpfkERM0pnUAWiMXimlYQKfFER/X+ET3aoyQa0e5nLNhnoDMpZhtnSprpMQM6B4xMDm8t3+JpXGSRHNLRDB4DcGrA/jcAjGaMjQHwTQAPAwARdQJwK4AJAMYDuJWIwvsnZgg6vi2/6MWKl/n3r6/0jF27I57de8Vx/UOPq2tdSBTvJZwJDNBU/JQZpip89tFLj/I4bOvumOY2vAe41iAay8irYVlIXvylPgnzmjSUP7p/Zzc8N8Z48bRvHNMXXdqX4cXvT9L+TjTJkSPBwjQDIrP6RbGQSBxVgqIQsrpoKfe3CDYTxX0aQntQmHiEgzvE7BV0jhMH1ADglUtbMs+gX228TlWmKsHO/WwHLn1sHgb95KWMzFeISFkYMMZmAVB1NRf797H4U1yB+CLoFACvMcZ2MMZ2AngNwUIlowgzWwDqFZNw5gGJAkVng5ehjOZwGIqqlWMquHrqQEwf3xuzfjQ1YZ9wIG/Q1ECaOqRLQvROVNKiiqKEqCMkZe1KXol1UPRcDlkkAwCevOJL+N7xgwBwJvfV0T3wU8n0o8ONpw8D4C2CJzvzVahvjEnM2nsj5fvAzUT6edzbafA8JfxW1joCmLDQHtR9Mzh9ETmLLuiYim1CULdXVJ/NJhqlIAC5ZEo6eGzOZxmZp5CRVZ8BEZ1FRMsBzADXDgCgJwC5dsI6Z1tOIb/SKlOS3FdAXu3ef+FYrVnA29EqEYKRyuaZP0w/wpDiRFx7yhAAQG+FP0P4DE7+/Szt74WgvHBCbwBep3GP6na4cnJ/XHFcf1xyTF93uxwaWFWmKmltxizdvApmxl8fvfQo124+zTC+X0DHgGb/eCpqnZLdn27dp9Rqpo/vjV4d2yn9DhccdZjR8RmLm4BUq3qxJchnEGP8fsnVdhPnCfZNqI7ZEpAr4Sbje9uw6yCemq8uu/Li4k3u5yXrdwfO8/qyzTa/QYGsCgPG2LOMsaEAzgRwW7K/J6IrHH/D/K1b1YlS2UCYD0BmEscP7aINybz79VV4ci53/KpU/cYmvk02S/VJ0jFtCqEZjOjBbfC/OPNw5bhPf3W6u0/4H8qKI+jQrhjlJUW48fRh2oJsqv7EpgvnoKgXFWTfSK+O5fjt10abHSiAts6VpW59oY++2KUUBrefPRLHDqzB7oONCXMM8jnmF65Taykykw4K6Ak2JfHQV9GHQ4eg6+82c9IPyQpEGZfRvTqgZ0ez3AwAuPTRebju6UXutRfw9wAP6vC2aN0uXP7X+fj5f5cFHuujL3Zi9Za9xrS1BbRINJFjUupPRDUA1gOQl1C9nG2q3z3EGBvHGBtXW6vu+JUpiJemJBrBtycH+wC81S71jrgPP9+JG57hjl/V+yps9+9+qi/nkCmISJ/+tRXoUlWK/1HY9wEuCIWGMMehS9em049yhZCQzXKdK0oS9gvEV8GJ2sTKX5zmqZQq6JRxjibZLh3obsU/5qlXp6eM6Or5rosAi8magcoE5GwMykVgTrQTEYU4kMOFrJyV3RIQrUHLiqPugsgE4n1p8OWa6O6HCkKQfL59f+C4s+6fgxN/p9eiAeD5hev1japaIbImDIhoIDmcgIjGAigFsB3AKwBOJqKOjuP4ZGdb1jC4q76NooB4aR699Ch0qUqM7pEhM+wokVEOguplExVOZ67Ivtaz+2AjznvwPTw5d63xan3WquQSgooUGpJ8aYIS98gVBonXqaQogp+fcbinb4Kq+mzvTuFaVTRCaC8c4SFjTQSzLLh6dfQe/8rJA5S/kX0GQUzaXzZdhqiqSiFzmNxr06zsTKOkKILGJMrAiwg+/+vWx+C+JwNdhryMNVv34fv/WIj//cfCjB47l0gntPRJAO8BGEJE64joMiK6koiudIacA2AJES0EcB+A8xnHDnCT0Tzn7+fOtqzhlBHdjMdq3x1Ppm28W1Y0QkbdokwXXpnSDIZ285os/jFvLeY6zew37zErtRwWLy/jx6cOVW6Xz6d7tV7IyrZt3SU4vGc8zFSljU0dEq49NscY9tSHv+ycpvAxflqfv2oi3rluKkqKIlomzZ2/Ac5hXzSRakyMMRDIU203CEFjGptjaUcT1W3bj/8u2mBAR/xA63cdxEdf7DJOfmtuFlFg3oueTLc2UVk26B04+/45ofOI+7d6y76Qka0H6p6KBmCMTQ/Z/2sAv9bsewTAI6keO1mYRBDFB4dvjhDhuauOwayV20BE6BqiSQDmDjqT6BsTLN+Uvr1z535zFXiQJqRVvm6PXzYhTYriiGbgQqmei3PG9sK/F6xDNELG5SVkjHac2kFRPrIDOYgLC0VLxSyF0NxT34TH5tQpo6/kqe96ZQWumjpQeZz5n+8EwBllv5oK5ZgwfPX/ZmNPfROmjeweWgBQYM1WzpjH/+oNFEcJ791wQuAxhCYtBMrHa3dheI/2oSVKZPzTMSkFMXH53dm5vwEdFebNUicUOkjbbW0oiAxko1eaxD8D0wABA7tUuQ3VTZhG2OLlzDE98JPThxlH3/ix7OenpPS7IJxyuLlGFdRBTqCLE6mjQrIaUSrlQUxw+SR+T5miUJ3A908Y5H7WUSF6ZqvApFIXQY9FXHtQC4OgazZj0UbsOtAQkivh/b7RCTtujrGkS1UIbatBk/S1ZP1urNi01z2X00fGn61t+w5p62XJkOtrfbp1H864712ced+7mCn1ZQhDtSL8OQiLNZFJnzmCLJl+4vmOghAGZrbfkP0ep3EKseUhusHdFxyBbx3XP9Q0seinJ+PmLw9P2F5e4lXypo00D7fU2ba/o9mugk4YmL4s8iWVezfrkKo5TTafqWaQs35198IzR4AmqeOncqmLIKYbZEoSPZpVWLfzAK56YgE27K53u+ipcPEjcz3f+zhawYAbX8QPnorbwl9bthl9r5+BpwwctToXwJf/MBun3D0La50ksxE9EutO/f61lXh+YWIsyRufbMb1//bWMBKO4KUb9uCZj5TxJ0rIZTx0OHZgjftZJwy+/vAHAMx6ibQWFIgwMB+rf7n5junj1bHkYYlnQc/Mr84a6X7etk9d0rdP53IcP7QL2pcVK0M4E2BwzsInICfUyagsNbci6pjzS0s2Gv1e/rUwWwQh1VIKHcv1EU2A9zx052SiCUYosX+zAA8tFZ8V+5n3+ErNIIC+vYY+EQHhUJ94x5voe/0MAMDzC+P2/2/9dT4A4Lp/hxeVm1sX7P6bX8fvrYr2e95Yhe//Y2FCsMVlf5mfEDEUVGE3CMO7x4WQThCv2xnPil6kCQ8WsMKglcHk5RXPZthI3Qt+zwXByWJBoXvnHBnPuduyV60uf779gPsC6fo4yzBhlS7T0Zp4wuc4wumkpTPhy+UqgqZL1j5vVA8qBKrz8/Rn1lwA+frrxhDxezljUaIwlH0GP/vPMpcBJxwnoOeBXzN4ZekmlzHJ/Samj++NkT07oLwkittf+kR5nDG9M1cN5uF31gTuF6GYQc5mnalJgEFfY8wdo3nf5GumK4UxUdYM1gUnsFlh0EoxxSDaJOjlzhbkVVJQ2WbBqGRGeNYRPXHXuaMSxgaZsh6+eByA+KpUJ1xMzlkkzOlWqSZqOZB84pOJQFQeR2b2iqPK11Ynb0wEERHhv4s24qonFmCHr4FLmL0/fvwgMxE/xqRBnHF9+/EP8dicOgDAq8s2u+PeXrEFEeKOzgffXqOMi89kaOauA8FBB7+YwQXS0g17tGPqG+LCYK2GYQeVWweAJ+eGm7QWfKHWQOW5N+yu17bCBbx+DD827j6I99eo+5uY4lBTs/YaZBoFIQzESzVEU7oZMC+bEGQC76GoPqr73dlj49qAzBiqAkwz4iWXmVGfzuX42rhE05UYcdJwbyJU3R3TcKKzzU1u0jA3E4Z1qIm/DKquZABwZB+zVWeyfhhlP2ITjSlkSObMRPHPTb6VbpBzGpAygwPMRHDyDDpJkS5Cq5ST3Tbsrvdcl/2HmvDPeV/gi+0H3DIeI1PpG6HBroOJZk5Vb++e1e08hRBlyAJr0p1vKceEJcm9/snmwP2Avo6Vv5S7zm8QhtPueQcXPPR+Sr8VGHLTy5h051st0nuiQIQB/x/cSMT7P2G/wXHE7Bcpsnv9DuTfnTcGy287FTOvneJhbF3ah4epyuOHaUo8i/OQozb8EM+XPhIolBTc/OXhGNGjvZYOUyS70E/VZyALfaWZyEQzoOA5+G/jO5p8LzKDmXbhhpZqNAO/QBF2dP+qWT7UAzM/xY//vRin3/tO3ISlIWXXgeRbUqo0g+l/ijPEK47rj6lDavHcVRPxszPUxQjlchMnDuuqHBNmSppXtyPQhNOxvNhIMyAKNxVt12gOYVoSwJPX/BnVKrSEMapAhIFe3fYjHWuQkDXCjg7EX1CVHCorjqJvCnHdMiM5ebj6ZREjGAN+f35w3Z50NINxfTthxjWTtPWKjPMrjFb1EoNtTu31CDuMiT9A9o/oNEr5p36mFGMsULvwO5C1PgOoCyzKzOXqqQM91/Yv730OwFtmW0eJn+4JTsnwIISFbv6/kwbj0UvHo7aqFO3LilF3x7SE4oxyZd3Jg2v8U4Ax4B8+M9AZY3rghtPiiY9765uwfFOiKUpcyvH9OmH5pr3KbOMSaaHRv6YCi0KEweyQ1p07NX2e9x1qwvG/fTshUipXKAhhIN4FkyjHMGYRFCIq9i1ZH38I27fjZp9ktLwHLzoS1506RLt/kxSTrfdxCEYCnHVEcN0eHSPOhJvENAw72WOZNp9PFjKjT0czkO+Ln6mygLBVDy0heQb+e+/Xls4b1wvXnjIkUNAO794+YR4RRSbIFqa+oNV4RQlfDIRVl1AtGvwVAuScg5ufXwoAuPUr8XDqDz7bnhDlU1lalFDkUWTcqzC2d0c0xxg+XpvI6GXNYFSvaixeHxxR9E5I2ZZlG9X+kX1O1NczH63PC0d0QQiDoJdKgBSfPPsNBIrYJzucRKioWN1VlxfjayFF1U4Z0Q3fnaLOFgXMUuBdzSB0pFnCWKowrUKazKGGdW+fUnawCTx1pzTHkLfrXmL5p/5yCTEW7AB3Hfsi50HBYEVtIhn+cMvLJ/GCi0HJ2lOG1CY88UJr8DN/UdE1iOatew8lXfTOn0C4YXdiiXG57PjVT3yEBT57/5Nzv0gQekHCQAQ2qExF8jwje3bA5j2HPBFafryzamvgOevMRXJb13dDtIuWQIEIA/4/6IaFMT4TB/MW52URXbsEdh1ocF+W/1x9LO5KodyyjC/1N2gVaHDOAulEE4VBZmRBdvJMCB4TUMiqXr4WOvOALIheW6Z2VMrPi7/kMpBoJjr3j4n1cMT1UtXeUUUkCaZ6vhNQIHpdB2kGL3ysD/H8lRP5I56hbfv0jD7GGIqjhIbmmJb5yZnbMvzXYuMuznjl6yabXtXH92pGQ7tVYe5nO7T0dqwoQf+aCnykEAbiJ2cf0ROjnNavQX6DzXsOYeVm/QLtz7PV4bayqfPpD9dpf99SKAhhEFTwK3Fs8H6TNU//Wm+dnsl3zXR/mAzPO6pvR/ToUIa/fnO8Z/sQJwO2plJf3iFeupjjn1d8Ca//v+OUY3Urx0wUzRNznzS8ayDDz7QoSGTAZseRGZOu5o0sMHRd6mT+dv9bqz37mEIzkBPtBAMTQzYrSjWIPIP9Um2cYse8EYnAbdLDaQn2T2zRrPhnruBlHsQz5O0SlziPqParmy8sHFRgo6MZyKvx6nbeZMEH/udIz/dJg2o853nMgBps39+AT7fqS1Uf0bsjFnyxK0FgCG3252cejuE92iNCwCJNRJHQMN5Zpc/0Xr1lX2DVYoDnieieWaBlSowXhDAwKRdMvv/aAQa4+Og+ePf6493OV7sPNrrHTobB/uvKYzDnhhNw3GBvfoR4MGoqw/sDiFOe0L8zBnZRh9bqVuwmUQ5hGNqtPe48dxR+e16wNpTpPg67NavTZPIMdJDHVGqywWXB9/on3to5YQ5k9zjOHL99bSUW+kp0iAxkWTMploIV5OnfC4h1v2naMPxJkyi2v6EZuw80ekyjOlMRA9C1PRdAIsS1qTnmybEwrYIrfAZi5Xz+uMM89+3tH03BqYd3w0c3n4Q3fjgZvTuV4wcnDfbcl6lD+TsT6DfoU40d+xvw+XZvHH/cgc/LvAzsUqntnlZbVYqBXSqV5d6/4XQE3FPfhLcVZUFkE+OhppgyQbElURDCQLyYIdFonrF+iDA7E3WuJBpBz+p2HskfS0Ez0GFgl0p8e3J/PHTROO0YcRyT2kA6Riyc3+nivHGHob2iJaaMZK6LydAdmrDIUM0gCccuoPcZBJ0PCzmOmFE+zjVPfuQZo6pNJJjtonW7PSWag5yTjTHmyYP4nU9oT/kNj/MXjFbrN2BANyfPRhz75LtnYextr0n0hbObsuIINu+pRyzG3Gf3xOFdPfetj9PzumNFCQbUVmLWdVMxtndHjzA4dmANaqtKMfczvSAc62Re+/0G7rvqHHVkz2osWrfbszp3TXgxhkmDavDBmu3KVprtiqPo3qEM97/1acI++b4M7lqJpz80b9KTDRSEMAgK0XMRwgSSadwtHhRZcIjWfKlWJZVBRLjhtGHKXsfuGJibxsqK1KYOf/G7bCLjmoHOTBRyHBM6ZKajY47y8zKka5WHnliMGWYxxz/7SyeoookEs/VHrwTZ21ds2oMKX6JjL6kV5c4DjWAAah2T5FZNTH2MMclMxFf2a3wmmiBhIC5Hz+p2aGxm2LbvkLuYMi1DJDSp6vJiEBHG9+uEDwL8BoO7VqGytCgh+Sye9Me/j+rVAdv2HcImyWwljtUUYzhuUC0ONcXcuksCj82pw8HGZlxxXH/MrduRoKXIDuRzj+yFBV/swqdbc9cfoUCEAf9vIgt0r6hY9ZSXqBknAPz98gk4b1wv9yWVE7FEGn6qPO/KyQPwp4v1moAf/3Qah89YHN5wpENIbLguuziTyLT/+NqT9aG5An7TC2BmJpKHmGSnrti8F6N/9qr7XZh4/BArS/GcBq3o5W5pArqy3nIHuPN92eqlRdGEXgjPXTXR29eAMXRxTEBBZqKK0igqS4uwRdM4Jqi4nLgePZ1ucRt216PZYZbRSMRIHRQOZHHdJvTrhI276z2CWTYVRyOE0Yd1SNAMmE+LH+k4keWAgl6duMA8ZkBnTOjfCSXRiNZvcMFRvdG5ogT3+XxHsuXgzDE9EY0Q/p1DR3JBCINkTCY6piRWz+cpSj8ITBxYgzvPjavZlaWKnsChFKhx/WlDE0pLBEE0mzlmQGLSTjL45VmHY8Y1k9KaI1lMHpxev+v+tRXooclDkK//CkVSklHNIInpdgro66yFpjbRb5zewAJBHbxUGci6p0uuQHrGET0813dw10pPJddohFBTWYq3rp0ik4vq8hIURUgvDJzOa12qSrVjiov011ZcU5E/snHXQSx0cgAammJGGrXcLQ/giWWA2m8gZhvbu2NC8pnrwHdGDe/eHtEIeSKK+tdUoEO7Ynz7uP4oLynCuL4dtW1i25VE8c1j++HtlVs9vgdZ2HdpX4bJg2vxzILc5RwUiDBIIppI89AJQZKMOUNlkmipEMpjBvDwU7k2ux9PfGsCbpo2LHCeCyf0wUBNF7NMQr6u/kgRAdMrFzRulZSjcdXxibkcJj4DORIoFWGg60UQN0PwZ61Wihbr7DuOyED2llQPfsCry4txzIAa3HfhWPzlm+Px5g8n49TDu3vOWWfKiRCPXgvSDCLEHaq6mHwTM5EwUW3YXY/b/rsMAO9nYIIiJ3RNMNPBXarQoV1x0slnfs2grDiKQV0qPRFFjHFaxfs8aVAtPtm4R1t1+KKj+6CqrMijHfiz6M89shc27anPWc5BQQgDE5+BuKk6Xi1HGJhCNbSFZAFuOH0YHv3GUW4bRhWOGVDjJiblGvJ1SdcsFSRwhe19aLcq18Zt+lsB2ZRkYo7yg0H9bIin88TfzQLgTUQ6dpBXqAufwXGDaj3b1ihszpdO7AsgXk6isrQIkwfXuiHQ8jmrTE2McXprq0q1PgPGABCha/sybWhpsDDgx+1UUYLSogg2SYlnRw/obPTeiDBmoVFFIoSj+nYK7LGgSj4T90E+5KheHbB4nTcMVaZJVI+d7WgH/ki89mXFuOTovnh56Sas3sLbavo1gBOGdUGHdsXKIJWW0BUKRBjw/0EXNOxhE5pBMsz8jDGJVRlbSBagrDiKqUPVTWvyEZl0IBut7g0GXTiht3K7bIM+NYnWoJ84jl0RWtrbVzraH1oo9+JNLGnBfQbyZWMAjv/t2wnHFdqutgqrtFlVAJCBgYi4MFAw+rhZhbc2/WLHAVz39McJ40x8BlEi9Khu57nGXx3dw+i9iWdsx6/V+H4d8dm2/dii0VZUyWf+2lAAMLJXNXYeaMR6p27SnvpGt5wEwE1JnStK3NIU+xX5GJdO7Iuyoijun8kji5p8qeWlRVGcMaZHaM5BtlAQwiAZn4EOMcUDEgZV68mWMhO1NshXJd1LZGJfNumhfPZYddmQAbXhxQVFv2cRaw4Addt4dI1YaftxvE94yxVs/Y8u72egr4Ekl0IXQ3S5DRGPZhBnCf0dJ3JzjNNbU1miEQbx4whH81Pz1yWWywjQ+GQau7Uvw0tLNkn7zB4IcU/l93x8P24uDdIO/MlnolCefNhRPb2ZyPPqdqJOyk+IRAjHDqrBO6u2IRZjyuS8zpWlmD6+N55fuAFrdxxQ+gbOPbJXznIOCkIYmFQtdaOJNM9dXDMw51Qdyovx8a0ne1ZEVhSo4S0TETECxUQAAB0gSURBVHyVdLuFc87kFhUZxCum2loRAM50avXLGsgDb/MVIbeoJBL55vItWL/roHtcuddygmYA3hNBZuQiGql9WRE+vvVkd7sYoSs74hEGkmYwfTzXjA46ztXaqlJs39+QUFtfzouQM4VjDPj2cXEzZJCZSFynCKm7/Zm8d6r3fESP9igviSaVfCYi8eRjDu1eheIoaTORAe432LbvEJZv2ov9imqoAPCt4/ohQsCDsz5Fo6Ly7sieHZQ5B69ryp5kEgUmDAyiiULYdbL10Tq0K8Z9F45NoMXCi0xcFlHSwSRW26RTWqo9E4A4A5afl4+dVSVzHMgizFGutjnxjjddm7fMdFXF7oi8T6soLPf9Ewd7tABxqtpS5RIXkBn2K0v56rxu+wEQcYd2c4xhpy+hLyaZifwO9eryEjdCKCgD2TUTRUgbCRaGIkVdleJoBEf26RjqRAaAj9bqe2+XFkUxpFtVYI0i4Td4Z9VWpZkIALp3aIdzj+yFp+avw8uK/uBEpMw5uP2l5drjZgoFIQzEy5COA1m8mKkw82Hd4yu8sKYchQqT6/rgLF42wZ/MJCDCaTu0C4/wMfEZpGNWFFw6EiF8qX+8D8DSDbvR2MwjgcT0/3uit4Cb0ALkoyesxh2BIvO/nzy7hB/Td2phz7bXTBT/LCLSAF6XqNZxuPudyLKZSK6JBHDG2Og88ybRRESkjNBKxoHsx/i+vHfBrgMN2OmUKZEvp0g+W/B5cKlqnomcWMtIoGv7MgztVoV3Vm3DLU7pbbnHgsC3jxuApuYYnluozgE68wiec/CPufGOdaqGWZlGQQgDf/xxKoj7DJL/ba+O8ZWfaUnnQkMyl1XXd1bUtAnqWStgsupPtYEO4HXaHt0/Hgk07d7Z7nYBnSCUhVGjpieCMnxZ893ETCQzbNmBPXv1NpfR+/0GQngRUULHu6qyIje5KshnIGiIkDrPxOT5UGkGQDzfYF7dTtz8HBeY8z+Pawq65DM/RvbsgD31TQnZ4DImDarB3Lodbo9nVTRf35oKfGV0D8/xZXSpKsOUwbWePs6XSL6nbCFlYUBEjxDRFiJaotl/IREtIqLFRDSHiEZL++qc7QuJaH6qNJjCpO2lKKilC4tLxWeggq7KZaFj7c70m35/yyBM9qvOS2iiGfQ3cBTr4JpmiNwyyKr9AC9S9tntpyeMkR/X3QrTTITURs2EZzTMTKSJJvq6L5pKJwxufp6zgMbmWALDP2FYV1czCHLay9FEZ4/thX9/52jtWB10C63Rh1WjJBrx1Cnq7utXrko+82OUlIncrjiKK45LfN4mDar1hJXq3ne5X8ny205N2H/ukb1cJ3RVWZFxxdd0kM4RHgOQeBZxfAZgMmNsJIDbADzk2z+VMTaGMWZeYyFFHDe4FscOrMH1pwUnWAGJNWAEhGqYrs2/KqRgW6Fih6Y1oIywBDnZHKPD4K7clGQSTaS91waPgOwzUIX4EpGbnLVo3S7lIkNevPTs2M63jx9DFd2TKAv4BlUhNUGLgLy69j+rOmHw7EfrAcQ1qcM6xWk9cVhX7HVCMCtL9bWuRMKdoKVfDb9PIsJKkJjK61dWHMWYw6o9foPenbyCXk4+mzSoRlnPaXDXKpREI1iyfrdW8Izv18nDuHVJiUOk4ACV+ez4YfFnZm+9XkBlEikLA8bYLABarwxjbA5jTOhd7wMIbu+VRZSXFOFvl0/w1lvxQdz8U0aoSz5ksuqoRSL8TeNVUOVtyDCJEBLMJtlS1alCOHL9z1VTc8wtA3HsQHX5DcaAWT+aCsCbjdzYHMPug43YsveQcsU4Z7W3UufSDdzpqbNRezOQ4186tPMKg4qSKNoVR7F17yFs33fINQ8JoSUY5NodPDSzprLEwww7GmRrC1o6VZTgqW8fLfVHDjf1Bpn1xvfrhCUb4uVH/JqBnHwmtC4/SooiGNa9yqlgql4TlBVHPb2iu3VITGx05wt4Xks1xSOziZbyGVwG4CXpOwPwKhF9SERXtBANgXj2uxNRd8c0ZVYqIFUybEmiCggm9VgqFLWeZCQTCqqzL8tIRwuUwy0B4JaveIvB3T/zU3eBIYTO2z+a4hkTYwy9O5ejU0WJJ5ro1he4c/KlJZuUq8qXl27yfG8MCVrwhpbG5xvYpRI3nh53gIrEs5Vb9uHIX7yOafe+w8/VIe2LHd7KvsJsd+PpQ9G+rCjQgayKeBrfr5NbUfVV3zmp0LW9nvGO79fJ84z5C07KyWexmN43OLJXB64Z6NLIEY8qgn4IAGDmj6bgv987Vrv/9+dzy3rQIjaTyLowIKKp4MLgx9LmYxljYwGcBuAqIlK24CKiK4hoPhHN37pV30moJeCvV5Is5lx/PJ741oTMEdTGYOKsDfO3BBVCExD379Vl4cwlHcVArJYFc+uhWCHGx/Dvok6/gMhCjRB5enE88UE8ysSExrBVpseB7JvwtMO9iZM1lSVumYzlm/Zi1ea9rhAu82kpgvlfcdwALPrpKYE0qLJ+ZZg4/IWWNEhRS2tsH2+/A5VZTiSfNWs0AwAY1bMaew81BUYFThpkVmixR3U7HN4z0Z8kIAJP6rbru7VlElkVBkQ0CsDDAM5gjLm6K2NsvfN/C4BnAYxX/Z4x9hBjbBxjbFxtbXqVLNOFWEm0S7HGf4/qdmlXEG3LqG9S27NlEBFOGNoFd54zSrnfZLUv+vOqEn78SMtM5Fs8EFFCwb+w4od9HeEQjSSGlgqomJo/gsXfkztxDv18/um5YIrTctLvZ6Hc0dj8Y5MJoxa+nEOa50CXDe7H3J+cgOeumpiwvbK0CIf34JFOfhORQDz5bL/2noyUggF0OUlysqCJ6VIH4YhugY6XALIoDIioN4BnAFzEGFspba8goirxGcDJAJQRSfmEyyf1x7UnD26ReN9ChL8qpw5//sZROO8odRlxsUIV7RdVWJ9Ek6J0Isf8ZiIACclUgqnKQkde1f7gpMF8P5G2nLWKwtN89ZLCGJKcoFbla+PpvwZyr2aB5mZvyWdhJkmmberoXlyA6WrymBYv7FJVltCsR0CEmOp6W4vks817DmlzFgZ1qXT9KjpNNVMlZ0QF16lDWmYhnE5o6ZMA3gMwhIjWEdFlRHQlEV3pDLkFQGcA9/tCSLsCmE1EHwOYC2AGY+zlNM6hRVBWHMXVxw9qkRCvQsQPThyc9hz+5iYqtHecov4mL8nApPZRPEkxvm2W1Af3ngvGuGPk+P+HL+HBdccM6Bwv0RAhZVh0VVlRvClMdTv89mvcxnzZsf2UNI3UmCRkGst8DM5EOWpy8wz49wf+50hMH98b33CqpZpA3Jc9B9WRM5lgsKJOkS5yTSSfAfGcFT+KohHXnBfmw0oXfTpX4L6vj8Xd5x8RPjgDSLmvIWNsesj+ywFcrti+BkBwd3SLgoNJg/gwCBt1UGSSiCMPKu2dCag0g04VJS4jGta9fTyRUTr3Pp0rUHfHNM9cESKlmWhvfZO7gh3ctRLnHNkL5xypN6d8SxEXDwSX5vCbS04d0c3joD66f2c3gUsMrSgtwu1nj9TOqcJJw7viz7M/C2zRmS6O6tsxcL9IPnt39XZtljvACxWu3rIvQXDKuP60ofjnvPR7Gk8blVjsMluwy1yLNgOhvjcH+ANE168bn12cVVpUK/mfOe0lJw2qweCuVa7pJ8w3EY0Q5FMSpa9PGt7V1VJMzMq6owStuv27+voiW95bs931v6TjY/lS/85Y9cvTMK6vOlcksH+5IarLw02RwlQUBNEHQtfEB+BtauVuca0BVhhYtBmYaAYCcshkNqCKjvnK6B6ou2MaHr9sgjPGLJExQl4HskiM/NPF41wOnw6vDOLhfpPYuQGaR1DoqAmCfp+M/yEdjHJ8Fz0DiuVdfHQfDOpSibOOCM57aW1I2UxkYZFvECvTIJ9B707l+GLHAZwywrwpjR8m9aXchi8BjFblQFYhGolH8PiziJOpyKtDkDDykxbUAjWdkt9hEPWo0m3Beuc5o7Bovb4gnUg+CxJ63Tu0w2v/b3JadOQjrDCwyCvUVKbQYN5BSTSCmspSXHeKvhXlQxcfib+//wUO61iuHRMGk/LXKp+BH27SWahmwKOJ9h9qwv0zV3v2iV+mpxmY+wwAYN5PTsTBhmYcd9dbnu3pagZBECaek4arKwSY4ryjDtNGowG85Mb8m050s8MLCVYYWOQN3vzh5LRewkiEMP+mEwPHDO3WHredeXjKxwB4x6owJNMmNSw9IhrhDuQpv5mZUBfIrcibRjVcCji+Shj4y1TLdGYLPavb4Z3rpqbc6yAZ1Bjc37YIKwws8gbCMdcWEM9YT78GEncgM2VROrcibxom9UCfRRL8PduNmw7rlLo2ZxEOKwwsLHyY+5MTzMJzAuAWNjQYa2Qm0vlBhDBIy2eQ2r50xlrkH6wwsLDwQVesMDmYlzwPy7GIEBIarC+4+STP/OnIriAaTcp2uPNYadCqYUNLLSyyAGG2MbGchPVWiEYIS9fHyy8P7FLploaOO5DT8BkEHD6ZZkzZ9BlYZB9WGFhYZAHJlDwPzzMgT9E3OedA+CSyFU3UrkQvDC73lb2oCBhrkf+wwsDCIguIKZLOdDBxIMuNZnp3jjtSxU/TMROZhMqq8JNpw7D6l6e53/3d2CxaF6zPwMIiC3BX6gZ81kQYNEmawT0XxAuXUQYcyGGyoFNFCU7QtO4sihJm/3gqXlu2GVOHJI6xaD2wwsDCIgtIpmd2smYiWUvIhJkoLPxVOKt16NWxHJdOVFdKtWg9sGYiC4ssIN64JnxsmGawftdBN6rnngvGePa50URB0qCFmqNYtG5YYWBhkSKCetOalKMQCLPZr96yz/3sb1IvfmlQmy/llq0WhQFrJrKwSAHv33ACKsv0r0+cSYdzaYNunS6WbtiDKZJtPhPlKCwsACsMLCxSQjdNH12BqMPhgyqoxseaL9n9OQmuA7llKjxbtGFYM5GFRRYgmLZJb4Vkavr8j68HdzLRRC3VWN2idcIKAwuLLCAaNRcGYRnIAt3aJzZ7N+nHvG7XQYcWqz5Y6GGFgYVFFiAYfHNzOAMuCukDIEpG33dhYmN0kej1rUnq/sYA8PFa3szlvU+3h9JiUbiwPgMLiyygKGLegjMMb/5wMg42NisL6HVoV4y6O6YZzbNxt75nr4WF1QwsLLKA452MXV2D92RQVVackUqq2/Y1pD2HRduFFQYWFlnAsYNqsOqXp7k9dXOJ288eCQD4v68nmpksLASsmcjCIkvIZk/gZDB9fG9MH987cMy904/AwDbUac4ieeTH02phUYAY2zv3WoPAV0f3wPAe7XNNhkUOkbIwIKJHiGgLES3R7L+QiBYR0WIimkNEo6V9pxLRCiJaTUTXp0qDhUVrxr+/cwzW/Or0XJNhYQEgPc3gMQCnBuz/DMBkxthIALcBeAgAiCgK4D4ApwEYDmA6EQ1Pgw4Li1YJIrKtIi3yBikLA8bYLAA7AvbPYYztdL6+D6CX83k8gNWMsTWMsQYA/wBwRqp0WFhYWFikj5byGVwG4CXnc08Aa6V965xtFhYWFhY5QtajiYhoKrgwODaF314B4AoA6N07OBrCwsLCwiJ1ZFUzIKJRAB4GcAZjTOTCrwdwmDSsl7MtAYyxhxhj4xhj42pra7NJqoWFhUVBI2vCgIh6A3gGwEWMsZXSrnkABhFRPyIqAXABgBeyRYeFhYWFRThSNhMR0ZMApgCoIaJ1AG4FUAwAjLEHANwCoDOA+50eq03OKr+JiK4G8AqAKIBHGGNL0zoLCwsLC4u0QIG9U/MIRLQVwOeKXTUAtrUwOTpYWtTIF1ryhQ4gf2jJFzoAS4sKmaCjD2Ms1M7eaoSBDkQ0nzE2Ltd0AJYWHfKFlnyhA8gfWvKFDsDSkms6bDkKCwsLCwsrDCwsLCws2oYweCjXBEiwtKiRL7TkCx1A/tCSL3QAlhYVWoyOVu8zsLCwsLBIH21BM7CwsLCwSBOtQhiQk6iQD8gnWvIF+XRN8omWfEE+XRNLSyLyhY5WIQwAlOSaAAnVAEBEOe8SR0TV0udcPlD2/ihg748SlYBbyj7XaCc+5Pj+WGEQBiI6nYheBnAPEV2UY1o6ENErAF4GAMZYUw5pOY2I3gZwHxHd4NDT4s4fe3+0tNj746WDiKgLEc0Er1UGxlhzDumZRkSvA7iXiC506MnV/XkewF1ENKWlj+9HXgoDIioiohsB/AzA3QDeAXA6EX0lh2QdBLALwOFE9DUgN6sbIhoP4KcAfgseaTCWiA5vYRrs/dHA3p9EOIy23vkbRUSnOXS2OP8hopPB7889AOYCOJ6IerQwDcVE9FuHjgcA7AZv8jWhJenwIy+FgbOqWwPgAsbYy+CF7DYgR+quw1Q6gjfpOR/AHxw6m3OgXk4EMIsx9gJ4X4hmAJ+KF6sl6LH3JxD2/vjgnHsvAAsBXA9etwyMsVgOyJkM4BXG2H8AzAdQzBjb0JIEMMYaAawAMJ0x9hK4tlQN/qzkDDm3qwoQ0SUANjDGXnM2PQOgiYiKGWN7iagXgPIWouUaACMBvAfgUYep7AEwjTF2IvHezrcAeIYxtoSIKFtqpkTLXMbYnwC8DuBlIioDcBb4S/9HcMZzczZocOiw9yeYFnt/4rRcA6AHgHmMsX8zxmJEtAHAYAC/AbCRiK4E8AZjbFUL0bKAMfYUgBkAXiWiYgCXAFhBRH9yaH0oW88KEZ0LYC1j7ANn02Pg96eEMbaBiKrAC3vmDoyxnP6Br+ieBrARwCIAUWd7RBpTBuA5AENagJ5vgK8wTwXwNoAbAQwA0AXAL5wx3wTQBGC+8724hWi5GXwF0RHA7wB8xRk3DMASACPs/bH3J1f3B9wR+gMA7wI4F8AnzjXqBGAcgFudcdcC2A/gP873ohai5XLwBfBAAI8AONYZezp4J8a+WaCji/NsbHDuQURxfzoCeANAt2zen7C/nJuJGO+T/Cr4C/MhHBXSh44AyhhjK4joMCI6J4sknQDg14yr1z8EUArga+A26dOI6FUA1wB4E/EqqtlyVvppKQbwPeeaDZaOvxx8lVyaaQLs/UmKloK+P4xztqkAbmKMPQ3OjMcAOAnAJgCTiOhFAJeCM+k1zk8zbh7R0HI4gPMZY6sB9AMXoACwGMBmABnXCBhjWwA8D75g2Ajg22KXNKwPgN2MsU1E1IuIjs80HSbIqTCQ7Kd/ZYztAnA/gLOJqA/jqqUwY/UHUEVE/wtu/8x42zPJmfURgC8DAGNsPvhL3B+8bedr4OaAMYyxkwFMIaJ+zoPXErTMAdCHiIaDM7uHiagcwE3gD/q6DNNh709ytBTM/fH7PqRrMh/AJABwhOQKAKMBHAF+/vMYYyPAm1pNIaKe6d6fJGk5gogGga/E73TGXQreh31nluj4A4Bl4EJ7GhF1Z4wx6f70BBAlou+Bm7G6pUNHqmhRYUBE3Zz/ESAezsUYq3f+zwNX137pfBcrurEAjgZX76Yx3jwnXVpGOHZdOMcSzqx3AUSI6Djn+1Lwh7gKwC2MsZukaXozxj5rQVqWOLQMZYz9DvzhfhrAcABnO6uQTNKRy/szkYgGSLTk8v6Y0pLt++OnI2f3B1KMvnMscU1Wgwuekc73WeD3ZguAKxljtzrjdwCYyBhTtrzNEi1vg/tNOoALziLi4a4jwDsy7skGHYyxRudezAHXEq9xtov7cxKAr4Dfn9MZY0+kSUdqaAlbFPiq4A0Af/JtJ0i2M2dbb3A77AgAXcGdKocDmJQhWkYBmA3gWQA9pe3CltcJ3KZ5H+L21wcA/Mj5HPXT3MK0/BHADc7nEgDVWaQjF/dnLPgK6hCAsTm+P6nQko37o6MjF/fnSwD+DeBRACdL513k/B8I4HZws5nY9gK4IBD3h3JIy38AfNf5XA6gSxbpIPlcnXM/Dty53wtAV+n3J2bimqTzl1XNgDh+D+CvAP7CGPuWtC/COGJE1I6IKgGAMfYFOFNaDL6q6MMYW8IYeydDZN0E4GnG2FnMWZUQUZTFVxN7weOySwH8xok6qAaw3aGvmWUuJC4VWjqC2zfBGGtg3DyQLTpa7P4Qj71+EDw2/17wtqhTJFpa7P6kSUvG7o8BHS36/hBPjLofnJmtAPA/ADo673KTc/zV4OaZAeBhpAAXYnXO/mbmcMAc0VIPx1fBGDvA0tfWguhgjDFGRKVEVOqc+yxwbXYJgJlENIgx9j5j7PV06MgIsi1twGNo/yp9HwCvJ/1n4BdylPN9Orjj7U5kMAoE3CTWH7znsth2EjgjEauGXwD4F4ChALqDh3+9D+BBONK+LdFiSMdtLXR/KgFcCKCd8/0bAP4EKdLEeVZa4v7kBS2GdNzaEvfHmfsaAA86n3sBeBLcMS32/wLAnwH0da7LC+BO7QeRIW0t32gxoOPnAB6HE6kE4Epwk9mvM31/0j6XjE/IVZ7B0vf24BLzFnAb6zPgmsJY8IiLJwAM9P2+XxZpWQXu9HsOfKX1VwA3OA+Nn5YIgKq2REsG6MjK/YHPdADgMgAPiH3gpqwnAAzI9v3JJS0ZoCOb788YADvABdBmADPBQzTPB3CM4lmpRAbMZPlESwboOFH+nk9/mZuIryZngKvONwGokPZdA+BjcHtZKbhUvB1ArTQmkyu7IFpuBLAAwFed78eBh34dLY3J5MohL2jJAB1Zvz+QbODg9t7NADqKfS15f1qalgzQkc37UyntGw/O7M5xvl8Grq2MbqH7kxNaMkBHxu5Ptv4y6TOoAF9Vfs/5LCIswBi7F8AUxtgsxtgh8BXoOAAHANd/kMlYYy0tAP4LvuLt5HyfDx4DXS/Rksk0+XyhJV06sn5/GEfMiTarc8ZMFvskWrJ+f3JAS7p0ZPP+TBI7GGNzwUNTRQ7Fm+CMcqdESzbvT65oSZeOnJaaMEFawoCILiaiyUTUnnHH40MAngJnIhNIKgDFeHKMwJGI121BJm6YAS09nWMtAvAjAFcRUQ24w2ck4g7INkNLvtBhSEsPZxw5xxMJWkIgUVujJV/oSJKWUvAQye86Pz0BfBEhwlvbDC35QkdLIem2l84D2A3cFhYD8Cm4pPw+Y2ybM2YigPPAE0z+5mwrBY91/g14Jt4PGWMr0yI+OVrmM8Yel377/8Cdp4MA/IAxtqwt0JIvdKRAi/ysRBmvN/Q3AKsZYz9Nh458oiVf6EiBFvdZIaIR4DbybgAaAVzNGPukLdCSL3TkBMnYlBCPnx0M4G9iG3iG3TO+sT8A9+h3QDwa4hgAZ6Ziz8ogLVXS9ox48/OFlnyhI01ayqXtJW2JlnyhIw1aqqV3uR2A/m2JlnyhI1d/xhcJwK/AHb+TwbPl/iLtj4DbmCdL2yrBa6nPA3d69cjUg5MGLXPbIi35QoelJb/pyAAt4l3u2ZZoyRc6cv0X6jMgosng8bkdwdO7bwNXg6YSb+QBxm1iP3X+BKaB29AWAhjJMlAzPAO0fNzWaMkXOiwt+U1HhmgR73LaJSTyhZZ8oSMvYCA1J4HX7RDf7wfwHfAEmA8lydkN3LnS19l2BoDjMim5LC35S4elJb/psLTkNx358GdyscrBoxiEPe1CALc7nxeCl+wFeKjok1kl1tKSt3RYWvKbDktLftORD3+hZiLG63ccYvE42ZMAbHU+XwpgGBH9FzwNewGQWMo1U7C05C8dlpb8psPSkt905AWSkKBRcHXpJTjp1OAZkdXgteRbzIFiaclfOiwt+U2HpSW/6cjlXzJJZzHwTk7bAIxypOXNAGKMsdmsZR0olpb8pcPSkt90WFrym47cIUnp+SXwizYbwGW5lGKWlvylw9KS33RYWvKbjlz9JZWBTES9AFwE4HeM1xjKGSwt+UuHpSW/6bC05DcduULS5SgsLCwsLNoeWrQHsoWFhYVFfsIKAwsLCwuL/99eHQgAAAAACNqfepGSSAYAyACAZABAMgAgGQCQDACoBkRX77IcKuwFAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_oil['Close_oil'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10edccb70>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAD6CAYAAABDPiuvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJztnXecVNXZx7/PLiy99yYdFURR1oZSRLERJVFjiSZqrK+v4qvGBGOJQRNNLElMjEpMrLFXBBQbiKIoi4g0EaT33uvunvePe2f3zuzM3Dsz987cnX2+n89+dsqZe38z597znPOc5zxHjDEoiqIoNZuCXAtQFEVRco8aA0VRFEWNgaIoiqLGQFEURUGNgaIoioIaA0VRFAU1BoqiKApqDBRFURTUGCiKoihArVwL8ErLli1Nly5dci1DURSlWjFjxoyNxphWbuUCMQYicjrwN6AQeNIYc3+cMucDdwMGmGWM+VmyY3bp0oWSkpIA1CqKouQvIrLMSznfjYGIFAKPAsOAlcB0ERlrjJnnKNMTuA04wRizRURa+61DURRF8U4QcwbHAIuMMYuNMfuBl4ARMWWuAh41xmwBMMasD0CHoiiK4pEgjEEHYIXj+Ur7NSe9gF4iMlVEptlupSqIyNUiUiIiJRs2bAhAqqIoigK5iyaqBfQEhgAXAf8SkaaxhYwxY4wxxcaY4latXOc/FEVRlDQJwhisAjo5nne0X3OyEhhrjDlgjFkCfI9lHBRFUZQcEIQxmA70FJGuIlIEXAiMjSnzFtaoABFpieU2WhyAFkVRFMUDvhsDY0wpcD0wEZgPvGKMmSsio0XkbLvYRGCTiMwDJgG3GmM2+a0lSKYv3czW3ftzLUNRFMUXpLpse1lcXGzCss7AGEPX2ybQp31jxo8cmGs5Sg5YsnEXzesX0aR+7VxLUZSkiMgMY0yxWzlNR5EBc1dvz7UEJUec9OBkznzk01zLUBTfUGOgKGmyauueXEtQFN9QY6AoiqKoMVAURVHUGCiKoiioMVAURVFQY6AoiqKgxkBRFEVBjYGiKIqCGgNFURQFNQaKoigKagwURVEU1BgoiqIoqDFQFEVRUGOgKIqioMZAURRFQY2BoiiKghoDRVEUBTUGiqIoCmoMFEVRFNQYKIqiKKgxUPKQsnLD9+t25FqGolQr1BgoeccjHy3k1L9MYf6a7bmWoijVBjUGSt7xzYqtAKzdvjfHShS/MMbw948Wsn5H4jr992dLeG/Omiyqyi/UGCiKEnq+XbmNhz74nv976ZuEZe4ZN49rn/86i6ryCzUGSv5ici1A8YvScqsy9xwoy7GS/CUQYyAip4vIAhFZJCKjkpQ7V0SMiBQHoUOpmYjkWoGiVD98NwYiUgg8CpwB9AYuEpHecco1Am4EvvRbQ1A8PXUJM5ZtzrUMRVEU3wliZHAMsMgYs9gYsx94CRgRp9w9wJ+AajPLd/c78zj3sS9yLUPxiFE/kaJ4Jghj0AFY4Xi+0n6tAhE5CuhkjBkfwPkDx2gbE2rUS6QoqZP1CWQRKQAeBm7xUPZqESkRkZINGzYEL05RFKWGEoQxWAV0cjzvaL8WoRFwGDBZRJYCxwFj400iG2PGGGOKjTHFrVq1CkCqoiiKAsEYg+lATxHpKiJFwIXA2MibxphtxpiWxpguxpguwDTgbGNMSQBalBqMuvMUxTu+GwNjTClwPTARmA+8YoyZKyKjReRsv8+nKLGIxpYqSsrUCuKgxpgJwISY1+5KUHZIEBoURVEU7+gKZCVvUTeRonhHjYFShT37y7jhxZlJk4KFGXUSpUZZuWF/aXmuZSg5Ro2BUoWxs1bxzqzVPDhxQa6lKFngp49/Tq873s21DCXHqDFQQsXqrXsoLfOnl6peIm98vXxrriV4Rl1/waHGQAkNm3ftZ8D9H3Pv+PkZHUeDifIPrdPgUWOQgP2l5RjthmSVbXsOADB5wfocK1GUmocagziUlxt63fEuv39nXq6lKBmgxlxRvKPGIA5ldiPy/LRlOVYSHGu27eGWV2axrzS8m4WMmfIDI1+cmcYn1aegKKmixqCGcudbc3n965V8sqBqAkC3DvV7c9bw4bx1ASmr5I8TvmPsrNWBn0dRlIBWICv5gcTpYS9Yu6Nin9ml9w/PtqSUUCeRonhHRwY1nFQazLe/WcVpf50SmJYImTbiGnmiKKmjxqCGkk6D+d3aHf4LcaBtuKLkDjUGSl5x34T5fGDPZ2gwkaJ4p8Yag0nfreeNr1fmWkbOybcG84kpi3MtQVGqJTV2Avnyp6cDcM5RHROWybN2MopvV1afFARKdigrNxQWqLOuplJjRwbJqAm3w7rt+xK+l89GUEnMzOVbci0hIfk2gg0jagyqOZt27stwpW3iz+YqKifbN/7+0nLemrkqNCuWZ63YSpdR41mzbU/Sctv3HvB10WB1iMKqDhqrK2oMqjHfrd1O/3s/5IWvludaii/k6kb/+8cL+b+Xv2Hi3LW5ERDDc/bK908Xbkxa7vC73+f8J6ZlQ1JoCIm9zkvUGFRjfli/C4Cpi5I3Gkpy1m23NvGJJMqrTsxaUTPmfnREEDxqDJSMeH7aMhat35lrGQnIXTey790TeeKTH3J2/vTQFrcmo8ZAyYg73prDWX//LNcyfCGRC2LJxl2MmZJaw75jbyn3vfudD6oUJTvU2NDSfOCzRVaSuUz8qH74YPcc8Dfzqclyjz5eDiYnFzzxBet37ONnx3amYZ0s3jLqH1eyiI4MQspFY6Zx77h57NpXyt44je223Qd48asVgZw7V5N0bo1yqvj1PXbtK/XnQB6p7s6aLxdv4tOFVbPhKuFGRwYh5YvFm/hi8Sae/GwJHZrWY+qooVHv7/dpn+Bk1LRJOzfbEZbQ06Dwq74vGGNFOIU9q60SjY4MHMxasZUev53Ahp2JF2QFjTGGSd9Fb/u4amvyePOMzhfYkVMn2+6hCJFG8LY3ZvNqSdXRlvhoFQfc9xEvfFk1FHjygvVszOF1pyhqDBz8Z+oSSssNUxdtSlouyCbrpekrKlJleMXZYb3tjW85/O6JPqvKLn65i9Kpp9vfnOPr8WJZvW0vv31zdtRrB8rKueyp6Vzy5Jcx5wuTqVbyHXUTeWDvgTIOufO9rJxrdYajgFTnEYL2fOw9UIYxUK+oMNgTBUjQ3rJyuxIWb7DWjeTKPVfDvIJKDIGMDETkdBFZICKLRGRUnPdvFpF5IvKtiHwkIp2D0OGF0/4yhQH3fZS0zI692Z1AzCY79x3grL9/xsJ1Vfcq8MNQHPvHjzj0rtQMaVh7xHk+ZaDUcHw3BiJSCDwKnAH0Bi4Skd4xxWYCxcaYw4HXgD/7rcMrC9btYPW2vbk6vS9s3LmPXne8yzdprEadsnAjs1dt4y8ffl/xmp8901RW9eYqmsj1+9aQLvPb36ymy6jxrNi8O+r1jTv3sWTjrhypUrJFECODY4BFxpjFxpj9wEvACGcBY8wkY0zkipsGJM4jnUPKjeGw303kpQxy/2zdvZ85q7b5qKoqJcu2sL+0nH9/tiT1D9sN5pKNu9m+12q4tQecgDz/XZ7+fCkAc1ZtY8ayzRWvH/fHjzjpwcm5EaVkjSCMQQfA6bheab+WiCuAd+O9ISJXi0iJiJRs2JD9uGVjYOe+Uh764Hv3wgk497HP+VEAK3Tj9WYzCX2cv2Y75z32ues5nOzYW/1y+aRD0AODRNWWK6N8+1tzOPexLypGCKXleW4FFSDH0UQicglQDDwQ731jzBhjTLExprhVq1bZFecTP2zI3vDaecteOOYLnrF7el75fl1qOYb+MWlRSuXDS0j8QBL5l1s9m3ftB4KfK5u7ehtdRo3ny8XJo/ecqFkKjiCMwSqgk+N5R/u1KETkFOB24GxjjAZYu/DAxO8ovvfD5IUcd8q0xZv53di5gWoqKwv3rZnORPT+svKERjSsE9vJeHrqErreNj6UC+Y+t0O4I3tWJyMk5jqvCcIYTAd6ikhXESkCLgTGOguIyJHAE1iGYH2cYwTGsk3uPfUw3jiPTvrBdVGSW2NVWlbO1t37Uz73vNXbWb5pt3tBn8j1zx9rRP1cdJZtRo+bhzGQjqenOho/JX18NwbGmFLgemAiMB94xRgzV0RGi8jZdrEHgIbAqyLyjYiMTXA4X3l39hoGPzCZDxP0RMJwy3ttCNNpMH/z+mz6jf6AMkfLEO+Gdy6Ken7aMs585FMGPTAp9ROmSM7i6z2eN9dGKhPC2MGJEF5lNYtAFp0ZYyYAE2Jeu8vx+JQgzuvGnNVWVM93a7cHcvxlm3bRuUWDQI7tBbf7fewsy1tXlkI38Y63Klfkhrg9iYtfhjVXRurfny3hvblrefryY9I+hoiEtuIy/V1Xb91D3dqFNG9Q5I+gGo6mo3CQyS3z3py1DH5gcpT/80AAyeRKlm5m5Isz+T6FRWJvf7OKy5/6Ku57mbQT89ZkZlT3l5bTZdR4/uISrfX6jJX8+b3c7w2QaZPq1juPbRsXrt/J5AXuUXRBufBCakMqGHD/xxz9B5d5NMUzNcYYlJcbHp3kbYOSdHzEc+1Rx3xHA3nfBG8N2Mad+/j3Z0s8DeXPe/wLxs5azcUxeWySceNL3zDJQ6MSi1tqi89/SB4F4ly8FC8N9J79Vmrup6Za6yPemlklzgCAW16dxT8n527XMC9Xg7NBDtol88OGnazfXrlQctADk1xHe2Fu1zP5uVIZ5SrJqRHGYP2OvTyWwhaE6dzM8T7yn6lVF4EtXLejSv6h/3vpG+4ZN4/5a6r29lPS4PGWd5ZLxe6laiMH/rlynmH0O/NcyzvXc8TbwyFVvNai1+9VbgyvTF/B/tLoEd9H89cx6IFJvDJ9BYvW74w7WbtovXvdGuPt2jv5oU845o/RKVTKE3zO7au9OXOl6/mC5j9Tl0QZNyU35H2iOmMMx/whee4hP0l0813+1Fc8eenRDPvLFCA613skZUMQbqVobUKu+oibdnmPYjKGrCUGTIWx36xm9Lh5rN4WbcznrrZGg79+/du4n5s4dy3XPDeDf/zsyKTH319WzsVPfkmnZvVT1uZmQ5zvfzBvHb3aNGTH3lJuenlWyucKgq+Xb+H0w9r5cqzzn/iCV6453pdjJeKXT0+nTeM63HfO4YGeJ5vk/chg+eaq/lS/R/ElSze7LsCatGBDlbDO0rJydu8v9W2CMp3vFfuZ0/86xR8xVc+U8J3te0vjJspLldgeu5N/f7aEyQsyi2LeahvtzTGGze13X7DW+m7zPcyxfP7DJl8nrCPHco4Gr3q2hCEPTq6Sgyhf+GrJZvdCGfLxd+td3ajGGMqrkRsr741BvKX0r86oOjR2JlRLtfqe+WJZxeNkN3L/mEVjN778Db3vqtx7IOMJyjQ+8+6ctVHPv1ubeaPsFWcDFRkxZcLocdHrA5wul3vGzeOyp6Zz7mOfV5mMds1TZ1dq5HipttXxjMW7s9ekeBSXc6RY+9meHJ69chsT50Zfa865Obd5umRyt+7eX5FXq6J8CGa///TeArr9dkLgI36/yHtjEO+iiDdaWLJxV9rrDNK98MZ/azUIkRshUz+5q6vAvqVCcJ8Ewszl7llbZyzb4nkyesnGXVE9u8jvFttwJWqIY9s3ZwDDfe8GFx21fe+BKj54P+t89/7SlLLRApz1j8+45rkZCd/3eu/FK9dv9Accfvf7Ua8FuTugVyIr2dUYhASvo7SoCbgUb5xM77PNu6yVxaNe/zbDvY2TKzkQYPqIV0tW8GKS7K4L1u3gt2/OZsSjUxn37Wrfzrt221622G4bPxu8Ret3cNKDk3nk44UVryVq9BOdN9JwJfrcjGWbM3ZdxdMw9MHJFRPMyfIcpeuOGvTnSRzx+/fdC6ZAmFZ5r922l58+/nkVd2A81mzLvdHxi7w3BuksPLrv3fkpnWOJIxmd14v62S+WVjzetc8aESzdtJs1Id5bYVOSdBi3vvYtt70xO+H7Kzbv4YUvlzNrxVauf2EmkLihSjTSiheeetx9H9H/3g+szyU8e3LiVVmkHkqWbqlQGelYFMSODBJF8iS5FkTg3Me+4LKnprNyi78NysadViOWeeci+fEffn8BT366mEkL1nPC/R/7EgEWBsZMWcz0pVt442v3SKt/TUkjbXxIyXtjkCjkLpZ12/dW3Dhbdqc2BHbGOnudMLrr7Ur/trMHkomvM95Hk02qpsorJe43x7CHP6G0rJxfetjHOVGPOdFmQ4l6apGfPN3fbvlmb41x5FpytvEn/unjhEaoYmTgIuucf06N/pxPneSXpldOcBpjjXaif0MXP72L7kc+XsS94+cz8sWZrNq6J+UtW50T6o96zICb6eBv74Eyhj38SdJJ5lTmX/Ipf1Peh5Z6NQbX/ffrtM9R5jjHQx98zw0n90z7WOO+TX9i8UAcQ9TrjrhbRQTGwvU7WbFlDx9/F0z+wd37S6lf5O9luzTJLl6fLdpY8bhizsDx/sotexK7iZK0tc63tsekik51H2svXPv8DD75fgMtGzpTN/jTkEV84qke7TVHIIfbLn1+OZEWrtvJwvU7GT1uLuNuGBi3jEkwAkyF0rJy9lSzkVLejwyCnizdsfcApTFD8VT3EfCLKd9nfwOgeATl/X34g+/pfddEdsZxF0H6dR1/o6Cqr62y3Tmx5d16h9noOz7xyeKkk6af2NdGxMXjJxF3XzYieFIdfaSDiTMCTJU7HSP/ZD/L7v2l7CtNbDS27t6ftcgoNQZp4EyD3ffu91kakxsm6H0Ews4zXyx1LbNyy+6UXVhv2ukqttuRLK+WRPegUxmyO7ci9XrPj58dHf1Vcd6EE8jJ5gz8NZl/+fB7Trj/Y7r/doJ74UoVSd9dt30vfe6qXPyXKCqmYi1DzO9QXm4Y/c48T2njvbJhxz4+/m5dxmkokrULkbe81FAi4zrBY+hw77smMvyR+Dshfr9uB/1Gf8DL0/0fKcYj742BVzdRKgx96BPXMr9+LRwrO3PBU1OXupY58U+TqqRU8Eqk8fl0YaULp8uo8VV2antr5iq6jBpfJQYd4Ed//6zCIMRrmF+LsxYlgvO8kKTn73XSwEf8zNVz5bMl7Npf2Wu9P0E47G67TGy02qINO/nP1CVRIaWphqTGY86q7cxamdytlMlPXuEmKnA3B+/Mih8Zl0pvftH6+DsMRl73kqzQD/LeGGxJYzMXP/Ay2Zor8mWdgVvnOpKcL9GcwFp7ojreYcYmuMmh6kri9dvjR1lFRj7pJAmMx77SsiouyWzyiksP9cxHPo16Xqug6vqZZD3mqYs2VoQJJyPoINTKQIHchrtG5iyC6NDGI+8nkGMn55Tqz1dLNjOiXwfmrfaWQjvRXr4ilitjcZIJZC+8niQEcenGXcx2uKQy4eA7/M3XlGpbtyPBXE0iahdafU2v61simXiX3HdmVENsJf+LPka6zfS42ZaRT5YaJDK4ysQU+NF8RwYm2cpokfcjg+qUGyRbRNJtV1dufOkbAPYmmXhz8uvX4ieQA2sxXJA8N21Z3NdTaWie/WJpaGL47xnnnn02QiRr7T7H3JAXN5azyA8bdnLKw59w99jK84q499oTvT35uw1VzhHLhh3WiDGTaCInm9KctI8ESnw4f50vubvcyHtjoPnOq3Le41/kWoIveL1ZE60SzYYX4N+fxV+UlMpo5K635zL0wck+KaoknY5Sou+TDOfEs1sIKUBpuVV+wdodzLLLO0dXO/aV4sGdHxcvQQYfzrfCossymniofLh4Y/w5ATde+LJyRf/0pVvS1+KRvHcTZcvfpmSXA2XlLPO4w1eiNm/Sdxs4rU9bH1UFR6KFeJnwsMsOc+kQb+LUGTVmDNw3YT4XHnNQwmOUl1uduNMSZNDdva8saaSWX5RlMD/j/BXSbYKcnZ10jV8qqDFQqiU9b898Md1z05YldOPUBBYmiGLJhHfiLJp0Lr5atmkXr3+9kiemLE54jNLycibMTDzRXCDpj+piI86S4VfLkagNcrv2nN/RL5dVMmqAmyjXChSl5vCDi4Hx0jkrLTPc8mri0Oyd+7Izf5KJi9m5MDLRYe58a07SYxQ6hwNZGBnkvTHQkYGiZA+3UG4vDazbfMrrX690jSTz47a/d3z8hJWpLpZMtw2KdhPpyCBjwrDJhaLUFCbMXpv0fS934xwPobjOnFHZJtVQ4XTboGg3UVqHSIm8NwYaTaQo2WO7ywrj7z2ESHppbJMtCgwb6TZBTjdRNiLf8t8YqC1QlKzhtn/C3gPuLpZkqUCqI+m4iYwxUWkovPxumZL3xiAbWQ4VRQkXXnrSM5dvYdH6Hazdtped+0o9r7uYtnhTwvfWb9/LKzEJFFMdGRhj+PPEBVGvPfdF8FFvgYSWisjpwN+AQuBJY8z9Me/XAZ4F+gObgAuMMUuD0FKTQwcVpabipTP+k39+HvVcBBrE7JXx6KRFjOjXno7N6le8FrsaeMayLUxesJ5JC9YzZ1XVie1U5gyMMfxxwnz+9Wn04r5sODh8NwYiUgg8CgwDVgLTRWSsMca5jv0KYIsxpoeIXAj8CbjAby2g6SgURbFYHrNI8clfFLP7QBk795ayc98Bdu4tZfPu/Tw/rXLl7wMTF/DAxAUc27U55xzVgTP6tqvIuRTh3Mc+p0Cgf+dm3HrawZx0cOuopH3et941jB43j6emLuXS4zvz3y+XU2q3X9lox4IYGRwDLDLGLAYQkZeAEYDTGIwA7rYfvwb8Q0TEBBD6o6GlilLziN2K9r9fLuO+mBTcJx3SOjqWH2t9gNMYfPrrk3hz5irenLmK37w+m7venhuVawng7xcdycCeLWlav4h4eGmDyssNvxs7l+emLeOXJ3Tlzh8dyqszVlJqpwfPxq5pQRiDDoDTabYSODZRGWNMqYhsA1oAvseL6cBAUWoekQZ45Zbd/Ob1b5m6aBMDe7aM2osi3rRCbAhnp+b1GXlyT24Y2oNZK7fx5tcreSbGf3/WEe1dtLjrvf2tObz41XKuGdSNUWccgohUpAAHWL7ZW+qVTAh1OgoRuRq4GuCggxLnMlEURXGyZOMuvlqymXvHz8cYw33n9OXCozvR9bbKneDiTTInWtwlIvTr1JR+nZry9fKtKa018DIyePGr5Vw3pDu3nnZwRUbW2FFL0ARhDFYBnRzPO9qvxSuzUkRqAU2wJpKjMMaMAcYAFBcXax9fURRPXPFMCQDHd2vBn887nE7N61cpEy8NtpeVvr3aNErJGHjxfo8c2oObhvWK0pTtzXWCCC2dDvQUka4iUgRcCIyNKTMWuNR+fB7wcRDzBYqi1FxGj+jDf688Nq4hSISXznirRnVS0vH8tOWu+0DffOrBVRp/p2Fyc0X5ge/GwBhTClwPTATmA68YY+aKyGgROdsu9m+ghYgsAm4GRvmtQ1GU6snNw3plfIwrT+zKL47v4mkfYydeXDOpem9mr9rGkAcn88unpzN5wXrPkUHO83RqVi+1k6ZBIHMGxpgJwISY1+5yPN4L/DSIcyuKUr3xw1Xeu33jtD7nxTWTqi//qcuOZuaKrbzw5XIue2o6XVs24BfHd07pPO2a1E3pnOmQ9yuQFUUJD4d3bOJaxg9feYM6wcXGpKqvVaM63DysF1NHncRfL+hH0/q1+f077tuHZiNTadT5sno2RVHymt+f3Sfp+9mKkGkYoDFI9yvUqVXIj4/swJvXncDY609wP4+jdc7GhKoaA0VRfMOtM9v/oGYZH8MLRbWCa9r86LEf3rGpa5nBvVpVPC4qDL6pVmOgKEreEeQAJFujm/6dKw1n/QBHOhHy3hjULyrMtQRFUVLAn83us+tvDwLnCCQb3ybvjUGdAIeLilKTOKZrc+b8/rSkZdwarWwtJgpy7tXZSN8zIvkcCaS/Befph7WteNyrTaP0DpICed9SHtbBPXpBUbLN787qnWsJKVMgmU/Mdm/V0LWMW0N+qYewzCAjcZxeorP7dXAtb9I0gXVqVXo1Dm6rxiBjvFx8SvXhV6dmviApDPjdVJ1zpHujlA3cmj0v7na3Io3q1s74GJnYiuiN6tM/TtjIe2Og5BdhSVrywlWxiXhT54HzDo963rtdegulAK47qUemcnzBrX78qD4RuKC4k2uZZPTIoJN4SLvKXroXwxSWa9YNNQZKTjmmS/OUyjvvq+YN4uePzwZ+uB9/GtOgXTO4Gx/cNIh3rj+Rxy85yvNxzjmyAz1aN6S4c/ywzQ9vHuzq6/dCZGL3V6f2on2Tuhwcx4+daI6udqHYx/BwHpdCx3drQd3ayZsut0noZJlE2zZOvtrXuetZJmQ7K6kbagyUnPLKtcenVN55Dz90/hFVetep0KWFPzd1OsRbxVqroICebRrRt2MTTj+snedjRfLv1EsQOdejdUNfF2FdP7Qnn992Mk3rW73iywZ0qaIlEck6yaf3sSZMu7RokPQYA3q0dF0FnOjtPnaaiib1Evfox408Memxu7SoT5/2jT1fe4nqZc7d3gx0toJg1BjUAJw3a9gZ3rcdA7q3SPi+czKuSb3aVXrXqfDyNcfz1OVHey5/Qo/EuuJx2YAuPHZx/B5+vdpWAxHpMUP6fuzIJii3nXFoegdIQNeW0Y1yrL5UvB+tG1m97WQuk0cuOpIZd5zCqX3a8tb/Rq/Q9Wuu6KZTrOMkMwYtGybPSioijB850NO1d9ePeieMBEpkJJy8e+NAPv3NSa7l/CDvjcFPsjCx1qZxailts02QqzEzIVHyrUZ1E/diT+tTGW4Xbw3JwJ4to54nuumX3j+cNo3rctLBraNef/nq46qUHXlyT47u0oz7fpLaKKSwQDijb/we/jlHWdfljDuHcVy35K6yy0/okvT9SG882e+WDrG2KV1j9fwVx9Kvk7Xi1mnMe7a2/Pav/88AZtxxCkW1CmhhN8SR8hGGxNSTW7RQorcLsnwrnN0veerpj24ZzH8uK074/qHtGlcY0qAJZyvhI0d0cl/2nS6XHBfc7mt++RM7t6iftBeUTa63JzkjPtm431CgmWMv2QaOBn/p/cM51DHJekjbqhOuT15azNd3DgOgQ9N6FY+90LtdY47tVrX3f9MpPXn12gEc5NGtFGnoE9Vg8wZF1LLTCzSuW7tZ0JL+AAAdE0lEQVSifmLLd2vVgKsGduV3Z/VhwsiBCc8XGRm4TVSO+Xn/iuP+5vRDkhfGvRPh9Qo90WGgnRpfueZ4Prx5MP07N6swAonoFOOnd7s93OYMsrbeweX97q0aMvSQNlnR4kbeG4NMmXJr4iHaiT2si9yfFZOJycRnaAxcNbCbj2rSJ3IDR3KuJLohIw3Gfef0Ze7o01M6R51ahTRvUMSc35/GR7cMTqlm2jetmjO+ZcM6VfzTRx6UuIMxekQfDkkxJjzyfWN7sx/fMoTbh1vrEZLFqhd7nITv1spy+wjxN2iJHaE0rlebF646ln/arq5E17lzX6ofHV45EmpUtxZ/vaAfFSclus4b1a1Fj9buUT23n3koTepHd2i6tqp0YQ3r3YaRQ6OjqbI9AsgHatRPdtePUl/o4xa1kAkHedyB6enLj8noPEW1CjytlMwWkUbPrSebiYltWKcWdWsXpuTaqOVhNPb1ncN48aqqrqQIvzi+i+sxYl0glT+DN7ERY9O1ZQPOPqI9Z3vcBcv5e597VLT7dOqooXEb+wHdW3qefD6tTxvqF1WWnX33afw41k3rU5yl08U4qFcrbj714Kj33Tpo4YrjCQc1yhi4+V7jkvSqscPl0ryy/nNZ5eTl3y7sl7BcnQwMUqRH+fPju2Q8dzDqjOSuhd+e6e56AIcxiNPblQSvJ+ORi47kiZ/35/FL+qf0uQiJJnkBfnZstCuweYMi6tZOPvHn1hANipnXqPicx+vojesG8NXtJzPpV0N45KIjvX0o6jyCiFS467q3akCHOKMiv+YMMiXudeIipnG94BO7JSOXYc/pUqOMQTqbZiS7sRMdrqeHoS9EXzAj+nWge6v4IXWZLEZydsQyvZfdes5XD+oe9dyZgjeaZL+pJHSbJOLsI9pzWp+2UblcnMd7/or0Foj98MczuemUnq7lYt1CsY3XH3/Sl0d/dhRHd7HWAZzaJ1pnrzbW9ZJsb11nPdYvqhV3UtHNiMa++/yVyUecQfvVvd6P8QYTbp9s1yT5NpFBf7fquKV7bs1nNSCd3lBk8vfQdo2Zv2a75885L5/IaYtqFVTpib5z/Ymc9Y/PvB3TcdDWjeuwYvOehGVbNarDhh37qrx+w9AeLN+829P5nJzSuw2ffL+hyuutG9XhguJO/DxOjpkgOp8nJuiJu5FsEt/5uz5+SX+GPDi5SpnItRMZXQw/PH5k0U2n9GJwr9Yc5SHXfyZUGNkE70ei4ooKC9hfVl7xemQLyUtjXGCVI7zIcd1rz3mNWw2m988c3KZRRbhrlCEJoOHt3KI+R/oQfOLHrm3ZokaNDNIhWVU633PGi0fC3tx8uW69h1tPOzjujkh9k2wduPT+4Qnfe+nq5Au8pt9+StzXbzn1YP52YeruiEQUiPCn8w73vIo3WY/ZK5/cOoSSO+J/v152z/6U3ulFdcTGi6faNtUqLOCYrqmtxE6HyMghUfsUiaS6IWYytmXDOiy9f3iV3yeV7xk5pfMztVw2bDknZl5j4k2DeNyOiIrn1koFtyb6k1tP4q8+XvPVAR0ZxKFvhybMXrXNtVzE6guRXpF1pXduUZ95a7bTtkl0I9a9VQN+2LAr8QFjbq7/9SHfjNPgpHID/ffKY7n4yS8zOnfjNOLenQ1VpKf58S2D2XugPMEnvNE5yarW7q0aMm/0aVGTn2D9dgl7dh4awrD2ChP14C88uhPHd2vBmm177XLB0bS+e7hzpBMQz+j0aN2QQb1aMSXOyDNZh0hJjI4M4tCzTaXPP3Zj7cM6NE4YFx7h4mM789TlR/Njl/S22fAqej1H7ERp6xR74xfbn7/tjEN4/X8GcP85fTnDkVLhw5sHeTpOpUmtpFHd2r6MDpIRawgyIVfeYtckcW7RWyJ0caw6dvsesVFhyWxfxDD66Us/qLnVual+3vlwUmNHBnf+qDf3jJuXtMwD5x1exV/fo1VDDpQatu05UHHxi0iUZSgQqqxsjUfkvmhhTyQ7L+r/GdK96gcS8Nb/nsD67Xv5ZsVWz5+JJTZ8MPLdnJPaie7jX53ai+uHWhOt1wy2dPfv3IwDtt+5sEDo0bqRp5tWoocG1YJEv4uf8v10i0d+4qb24r7Y1b2pkk6+fk8J6+xS6e4HEI9+nax5masGBbv2pjoaqBo7MvjlCV14/6b4vdXYYfQtw6LzoqR7cSZyG1RdbDSYW2LipiMMj5PeoF+nppzapy2/jrOqNN1GpI0ddnjFie43TcLvld6pq0XKXz8bqGydr0frhgw9pDUPnHcEYM0FTLvtZG5LEDLsVn+jRxzGwJ4tKya+k40MIhFx7TL09WdK8wZFLL1/OAO6pxdU4BW3yfowUmNHBuB9KzlnnL+I0LNNI75ft7OKC6myUOXDCSMHsnLLbq5+bobrebwMoR9NEhMfj2SpepPRqG7tKr7XTBskb73B1Mqnyie3DmHVlsQRVU6M8R5NdnqftpxlBwzkypjFnvbxS47iYEfKjtqFBVFrWwDaJsgPFe94sfRq04jnrjiW8d+ucdV2zaBuDOzZsmKxmJf5FLfFiUGv/PeDkE4bxaXGGoN4F+PlJ3Th0LaN+XLJZqDyZvhp/07869MlbNixj9P6tGVQr5ZcfOxBCSdknRdp7/aNqVWY2hWR6cSjSOoNUpBx0W7H/vTXJ/HopEW8NH0FSLC97s4tGiSdTIbK3y8VFZEol+gDpaYtGelUTyppsJ2kevlFVtInC40tKBAO69CETTurhi4n1JGaDI7t2rzi3lVSp8Yag3gc1r4J5/bvWOWCatagqErY5YDuLVmRYux97MUd2+il0wTeeHLyRVFDD8nMH+wkUYPkdb/ZRN+vU/P63HbGocxetY3rT+rBPyYtAnIXjRNvEjuWZI1zJGIyyH144+HX2VI1PH07NmHSr4Z42h8iCDMf0fvsFcewZ39ZAGeoGfhqDESkOfAy0AVYCpxvjNkSU6Yf8BjQGCgD/mCMedlPHekSe6Gme3PFDr3dby5J8iw5reOkz440ZlNHDU05Kigd3PZLiG3U47WRTerXZnwkM2eO5wzEHhp4XRQVm0jw58d1YeWWPb6EBlcXnHsffHzLYNfynlyGbplJY96vU6swahP5XFIdVyD7PYE8CvjIGNMT+Mh+Hstu4BfGmD7A6cBfRSS4PNMpEFuBXqvTGUz07o0Dq2wK4iz33T2OLJwxJ/jdWb3p0LQe7Zpmlr88cti2jetSO2Zhz7s3DkyaP93LcWNx26Sjyu/q8YfNlbu1YoGUx/LNYvLQ1CsqZPSIw3zdXcyLFr+an0wHNN1aNaRbgj2G02kjq2PDGqE6zGtE8NsYjACesR8/A/w4toAx5ntjzEL78WpgPZAoiU1O8HozxLtGExmCCPGSnEXON/SQNkwdNdRT7yaS9ydp7qQ4rx3arjFDD2lT0VBdeWJX13OlS5URgcfPheXWT9YGRd7ysnhKqYqXe8xrQ1qdjUWY8NsYtDHGREIL1gJJ1/eLyDFAEfCDzzr8wWsPViqTj8XLZ5NoQjSTS7i9PXpINtma7PiTfjWECSMHxnUzJTxehvdcdbllU+kZZ6vfl80G74iOTenboQl3DPd3G01ILzggG+s4lDTmDETkQ6Bqeki43fnEGGNEJGHNi0g74DngUmNM3FwDInI1cDXAQQcFt6tYhMqEW17LV369h356BLefeWgVt4yT2J5OULHIXiZAWzWqQ6tGdfhs0YYoLX7rABh+eHSOpkzdEC9ffRxrt+/N7CBJeOqyY3jmi6VR+aYSUV0MXCrUKyrknRuSbwqfOd5DS5XskLIxMMbEz/YFiMg6EWlnjFljN/brE5RrDIwHbjfGTEtyrjHAGIDi4uLg77s0zyAIRbUKksZsJ/18Bhd9pj7JVD7vNH6vXHM85z/xRdLyBQVCyR2npL3tpltCtaA4sWfLtDOd5pJq4S5JQWIkBLiTx02gwkQ1qIkq+O0mGgtcaj++FHg7toCIFAFvAs8aY17z+fwZUTXUM3mVRnLKXz80edRIR3v/1utOik4xEVQ8fSSc1O+OlbOt8Zpls2XDOlVGS+45dKrPrZQ1N1GWzpMtvHSAzj2qAy9ffRwjXDaVDzXVaHTj9zqD+4FXROQKYBlwPoCIFAPXGmOutF8bBLQQkcvsz11mjPnGZy1p47WnXq+o0FOGxIZ1aiUt53fEwT9+dhQbduyjwMM2jhGy0dik+i3D7CaoTgaruiIigY8ClUp8NQbGmE3AyXFeLwGutB8/Dzzv53mrK5m0J8k+W7d2YSiH1l6/bnVqZsOapjqsNG9QRM/WDflNnDxa6VKdrpcwU+NWIHdqXi/hbl+RBjbi1mnRIPgFW5DhnIFPbVE6h6ldKBwoSyNjpceTVacY7aDx0nGoDo1ircICPrjZfVGaF9QQ+0uNMwavXzuAuS5bUV43pDu92zXm5EP9S+UQjzDdvJ4WNcW0SB/ePJj5a3akfq4M8+6HgWogUckl1fACqXHGoHXjurRunDzqp1ZhQdpbIKZDLvs3mXSuvCR9izpXisevDh2/aiAx7wlz56E6XMMRaux+BrE4UxBni0wmIX95YldaNqwT+OjFSbYWnYX43s4h+qsowVLjRgaJiJuCOGAqtwtMvfvQq02jhBu8Z6IlG1Sn3lIiwtgbbZmlOS4lP9GRgeKZbLV/lw3oDMBx1SCsMFuGLdnK9ghN6tdm/ujTXcvlC73bWxv3dGvl3VWpJEZHBkrWcetV9+/c3NP6jZpE3w5NPJVzyyCbT/y0f0f6dWrqecfCbBLCgaMrOjIIAWFwm3hZDR1G10iuyPYeyBpGWRURCaUhcFKdak2NQQ5pVNcamA2shnlwMiG/2rW8+jJKDaZGuInuPqs345Js2t0gR0PrpvWL+PTXJ6Wd4M5PvCzwynZvONToT6HkGTXCGFx2QlcuOyH+Ji4vXHUsXVKIlfebsKSN0IY+PfJrlKPUZGq8m2hA95a0b1ov1zJyRqwvetKvhjB+ZPxc9sMyXIjXv3MzAI48KBS7nCpKYNz5o0MpKiygXpydDcNKjRgZKN5Jtm1nn/beIloSMeTg1sy8c1iVPYOrIzqOUpJxwdEHccHRwW/I5Sc1fmSgWGQrUigfDIET9RIp+YIagxqONmaKooAaA0VJC11zoeQbagwUJQM0mkjJF9QYKCnRpnEdClPYTlNRlOqBRhMpKTH1N0M1kgZoWr82ADee3CvHShTFH9QYKClRy0P2zJpA3dqFmkxPySv0zlYURVHUGNR0dAJUURRQY6Ao1YZLj++cawlKHqPGQAEy249ZyQ6/H3GYzlMogaHGoIajXiJFUUCNgaIoioIagxrPwF6tADi7X/scK1EUJZf4agxEpLmIfCAiC+3/zZKUbSwiK0XkH35qUFKje6uGLL1/OP07N8+1FEVRcojfI4NRwEfGmJ7AR/bzRNwDTPH5/IqiKEoa+G0MRgDP2I+fAX4cr5CI9AfaAO/7fH5FURQlDfw2Bm2MMZGd59diNfhRiEgB8BDwK7eDicjVIlIiIiUbNmzwV6miKIpSQcq5iUTkQ6BtnLdudz4xxhgRiRe8fh0wwRizMnb/3ViMMWOAMQDFxcUaCK8oihIQKRsDY8wpid4TkXUi0s4Ys0ZE2gHr4xQ7HhgoItcBDYEiEdlpjEk2v6AoiqIEiN9ZS8cClwL32//fji1gjLk48lhELgOK1RC48/gl/dlzoDTXMhRFyVP8Ngb3A6+IyBXAMuB8ABEpBq41xlzp8/lqDKcfFs8zpyiK4g++GgNjzCbg5DivlwBVDIEx5mngaT81VAeG923HOUd1yLUMJU8Z0L0Fn/+wKdcylGqGVJcEZcXFxaakpCTXMhQl9BhjMAYKdHtSBRCRGcaYYrdyutOZouQZIqL7VCgpo7mJFEVRFDUGiqIoihoDRVEUBTUGiqIoCmoMFEVRFNQYKIqiKFSjdQYisgFrVXMsLYGNWZaTCNUSn7BoCYsOCI+WsOgA1RIPP3R0Nsa0citUbYxBIkSkxMuCimygWuITFi1h0QHh0RIWHaBacq1D3USKoiiKGgNFURQlP4zBmFwLcKBa4hMWLWHRAeHREhYdoFrikTUd1X7OQFEURcmcfBgZKIqiKBlSLYyBuG2WnEXCpCUshOk3CZOWsBCm30S1VCUsOqqFMQCKci3AQVMAEcl5+m8Raep4nMsLSusnDlo/cWkIICKFuRYC1Is8yHH9qDFwQ0TOFJH3gL+JyM9zrKWJiEwE3gMwxuRsQ2IROUNEPgEeFZHbbD1Zn/zR+kmoResnWoeISGsRmQw8CWCMKcuhnuEi8iHwiIhcbOvJVf28DTwgIkOyff5YQmkMRKSWiPwW+D3wV+BT4EwROSuHsvYAW4HDROSnkJvejYgcA9wNPIQVaXCUiByWZQ1aPwnQ+qmK3dDutf8OF5EzbJ1Zb39E5FSs+vkb8BUwVETaZ1lDbRF5yNbxOLANuEhEjs2mjlhCaQzsXt1i4EJjzHvAWGA1ORru2o1KM2AacAHwd1tnWQ6GlycAU4wxY4EVQBnwQ+TGyoYerZ+kaP3EYH/3jsA3wCjgLltneQ7kDAYmGmPeAUqA2saY1dkUYIw5ACwALjLGvIs1WmqKda3kjJz7VSOIyKXAamPMB/ZLbwClIlLbGLNDRDoC9bOkZSTQF/gCeMpuVLYDw40xp4jItyJyF/CGMWaOiEhQw0yHlq+MMf8CPgTeE5G6wE+wbvrHsBqeO4PQYOvQ+kmuReunUstIoD0w3RjzujGmXERWA72AB4E1InIt8JExZmGWtHxtjHkFGA+8LyK1gUuBBSLyL1vrmKCuFRE5D1hhjPnSfulprPopMsasFpFGQAu/z5sS1ubZufvD6tG9BqwBvgUK7dcLHGXqAm8BB2dBz2VYPczTgU+A3wLdgdbAvXaZXwKlQIn9vHaWtNyJ1YNoBjwMnGWXOxSYA/TR+tH6yVX9YE2E3gRMBc4D5tu/UXOgGPidXe5XwC7gHft5rSxpuRKrA9wD+A9wol32TOBdoEsAOlrb18Zquw4K4tRPM+AjoG2Q9eP2l3M3kTFmC/A+1g0zA3sIGUMzoK4xZoGIdBKRcwOUdDLwJ2MNr28B6gA/xfJJnyEi7wMjgY+pzKIa1GRlrJbawA32b9bLcf7vsHrJdfwWoPWTkpYaXT/GatlOAu4wxryG1Rj3A4YBa4GBIjIBuByrkV5sf9R390gCLYcBFxhjFgFdsQwowGxgHeD7iMAYsx54G6vDsAa4JvKWo1hnYJsxZq2IdBSRoX7r8EJOjYHDf/qsMWYr8E/gHBHpbKyhZcSN1Q1oJCL/h+X/dE3HmoaWyG8xE/gRgDGmBOsm7gacCHyA5Q7oZ4w5FRgiIl3tCy8bWj4HOotIb6zG7kkRqQ/cgXWhr/RZh9ZPalpqTP3Ezn04fpMSYCCAbSQXAEcAR2J9/+nGmD7AhVj10yHT+klRy5Ei0hOrJ/5nu9zlQAdgS0A6/g7MwzLaw0WknTHGOOqnA1AoIjdgubHaZqIjXbJqDESkrf2/ACrDuYwxe+3/07GGa3+wn0d6dEcBx2MN74YbYx73QUsf26+Lfa7IZNZUoEBEBtnP52JdxI2Au4wxdzgOc5AxZkkWtcyxtRxijHkY6+J+DegNnGP3QvzUkcv6OUFEuju05LJ+vGoJun5ideSsfnDE6Nvnivwmi7AMT1/7+RSsulkPXGuM+Z1dfjNwgjFmVRa1fII1b9IEy3DWEivctQ/wc2PM9iB0GGMO2HXxOdYocaT9eqR+hgFnYdXPmcaYFzLUkR7Z8EVh9Qo+Av4V87rg8J3Zrx2E5YftA7TBmlQ5DBjok5bDgc+AN4EOjtcjvrzmWD7NR6n0vz4O3Go/LozVnGUtjwG32Y+LgKYB6shF/RyF1YPaBxyV4/pJR0sQ9ZNIRy7q5zjgdeAp4FTH965l/+8B3IflNou8NhbLEETqR3Ko5R3gOvtxfaB1gDrE+V3t7z4Ia3K/I9DG8flT/PhNMvkLdGQgFn8BngWeMcZc5XivwFiUi0g9EWkIYIxZjtUozcbqVXQ2xswxxnzqk6w7gNeMMT8xdq9ERApNZW9iB1Zcdh3gQTvqoCmwydZXZvwLiUtHSzMs/ybGmP3Gcg8EpSNr9SNW7PUTWLH5jwATgSEOLVmrnwy1+FY/HnRk9f4Ra2HUP7EaswXAJUAz+14utc+/CMs90x0rjBQsI7bUfr/M2C1gjrTsxZ6rMMbsNpmP1pLpMMYYIyJ1RKSO/d2nYI1m5wCTRaSnMWaaMebDTHT4QtDWBiuG9lnH8+5Ez6T/HuuHPNx+fhHWxNuf8TEKBMsl1g34j+O1YVgNSaTXcC/wKnAI0A4r/Gsa8AS2tc8nLR513JOl+mkIXAzUs59fBvwLR6SJfa1ko35CocWjjt9lo37sY48EnrAfdwRexJqYjrx/L/BvoIv9u4zFmtR+Ap9Ga2HT4kHHaOA57Egl4Fosl9mf/K6fjL+L7we0hjy9HM8bY1nMu7B8rG9gjRSOwoq4eAHoEfP5rgFqWYg16fcWVk/rWeA2+6KJ1VIANMonLT7oCKR+iHEdAFcAj0few3JlvQB0D7p+cqnFBx1B3j/9gM1YBmgdMBkrRPMCYECca6UhPrjJwqTFBx2nOJ+H6c+/A1m9yfFYQ+c7gAaO90YCs7D8ZXWwrOJ9QCtHGT97dsm0/Bb4Gjjbfj4IK/TreEcZP3sOodDig47A6weHDxzL37sOaBZ5L5v1k20tPugIsn4aOt47BquxO9d+fgXWaOWILNVPTrT4oMO3+gnqz885gwZYvcob7MeRCAuMMY8AQ4wxU4wx+7B6oMXAbqiYP/Az1jihFmAcVo+3uf28BCsGeq9Di5/L5MOiJVMdgdePsSi3o82W2mUGR95zaAm8fnKgJVMdQdbPwMgbxpivsEJTI2soPsZqKLc4tARZP7nSkqmOnKaa8EJGxkBEfiEig0WksbEmHscAr2A1IseKIwGUsRbHROhPZd4W/KgwD1o62Of6FrgV+F8RaYk14dOXygnIvNESFh0etbS3y4l9vsgCrYhBknzTEhYdKWqpgxUieZ390ZOxOhGR8Na80RIWHdki5W0v7QuwLZYvrBz4ActS3miM2WiXOQE4H2uByfP2a3WwYp0fxFqJd4sx5vuMxKempcQY85zjszdjTZ72BG4yxszLBy1h0ZGGFue1UmisfEPPA4uMMXdnoiNMWsKiIw0tFdeKiPTB8pG3BQ4A1xtj5ueDlrDoyAmp+JSojJ/tBTwfeQ1rhd0bMWVvwprRb0JlNMQA4Mfp+LN81NLI8bovs/lh0RIWHRlqqe94vSiftIRFRwZamjru5XpAt3zSEhYdufrz/CMBf8Sa+B2MtVruGcf7BVg+5sGO1xpi5VKfjjXp1d6vCycDLV/lo5aw6FAt4dbhg5bIvdwhn7SERUeu/1znDERkMFZ8bjOs5d33YA2DThJrIw+M5RO72/6LMBzLh/YN0Nf4kDPcBy2z8k1LWHSolnDr8ElL5F7OOIVEWLSERUco8GA1B2Ll7Yg8/yfwP1gLYGY4LGdbrMmVLvZrI4BBflou1RJeHaol3DpUS7h1hOHPy49VHyuKIeJPuxi4z378DVbKXrBCRV8MVKxqCa0O1RJuHaol3DrC8OfqJjJW/o59pjJOdhiwwX58OXCoiIzDWob9NVRN5eoXqiW8OlRLuHWolnDrCAUpWNBCrOHSu9jLqbFWRDbFyiWftQkU1RJeHaol3DpUS7h15PIvlUVn5Vg7OW0EDret5Z1AuTHmM5PdCRTVEl4dqiXcOlRLuHXkjhSt53FYP9pnwBW5tGKqJbw6VEu4daiWcOvI1V9KK5BFpCPwc+BhY+UYyhmqJbw6VEu4daiWcOvIFSmno1AURVHyj6zugawoiqKEEzUGiqIoihoDRVEURY2BoiiKghoDRVEUBTUGiqIoCmoMFEVRFNQYKIqiKMD/A6wR4AJCvpM/AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create return (open-close) normalized\n",
    "\n",
    "df['return_next'] = df['Open'].shift(-1) - df['Close'].shift(-1)\n",
    "return_range = df['return_next'].max() - df['return_next'].min()\n",
    "df['return_next'] = df['return_next'] / return_range\n",
    "\n",
    "df['return'] = df['Open'] - df['Close']\n",
    "return_range = df['return'].max() - df['return'].min()\n",
    "df['return'] = df['return'] / return_range\n",
    "\n",
    "# If return > 0, green; otherwise, red\n",
    "\n",
    "df['return'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gmt time</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>datetime</th>\n",
       "      <th>return_next</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-31 14:00:00</th>\n",
       "      <td>31.05.2018 14:00:00.000</td>\n",
       "      <td>1.29518</td>\n",
       "      <td>1.29770</td>\n",
       "      <td>1.29430</td>\n",
       "      <td>1.29718</td>\n",
       "      <td>9.841480e+09</td>\n",
       "      <td>2018-05-31 14:00:00</td>\n",
       "      <td>0.005464</td>\n",
       "      <td>-0.064288</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 15:00:00</th>\n",
       "      <td>31.05.2018 15:00:00.000</td>\n",
       "      <td>1.29716</td>\n",
       "      <td>1.29903</td>\n",
       "      <td>1.29586</td>\n",
       "      <td>1.29699</td>\n",
       "      <td>1.169860e+10</td>\n",
       "      <td>2018-05-31 15:00:00</td>\n",
       "      <td>0.045002</td>\n",
       "      <td>0.005464</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 16:00:00</th>\n",
       "      <td>31.05.2018 16:00:00.000</td>\n",
       "      <td>1.29699</td>\n",
       "      <td>1.29750</td>\n",
       "      <td>1.29555</td>\n",
       "      <td>1.29559</td>\n",
       "      <td>7.277670e+09</td>\n",
       "      <td>2018-05-31 16:00:00</td>\n",
       "      <td>-0.072324</td>\n",
       "      <td>0.045002</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 17:00:00</th>\n",
       "      <td>31.05.2018 17:00:00.000</td>\n",
       "      <td>1.29560</td>\n",
       "      <td>1.29803</td>\n",
       "      <td>1.29424</td>\n",
       "      <td>1.29785</td>\n",
       "      <td>7.262840e+09</td>\n",
       "      <td>2018-05-31 17:00:00</td>\n",
       "      <td>0.071038</td>\n",
       "      <td>-0.072324</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 18:00:00</th>\n",
       "      <td>31.05.2018 18:00:00.000</td>\n",
       "      <td>1.29785</td>\n",
       "      <td>1.29884</td>\n",
       "      <td>1.29538</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>6.044530e+09</td>\n",
       "      <td>2018-05-31 18:00:00</td>\n",
       "      <td>-0.032144</td>\n",
       "      <td>0.071038</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 19:00:00</th>\n",
       "      <td>31.05.2018 19:00:00.000</td>\n",
       "      <td>1.29566</td>\n",
       "      <td>1.29671</td>\n",
       "      <td>1.29541</td>\n",
       "      <td>1.29666</td>\n",
       "      <td>5.330320e+09</td>\n",
       "      <td>2018-05-31 19:00:00</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>-0.032144</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 20:00:00</th>\n",
       "      <td>31.05.2018 20:00:00.000</td>\n",
       "      <td>1.29668</td>\n",
       "      <td>1.29682</td>\n",
       "      <td>1.29502</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>3.116830e+09</td>\n",
       "      <td>2018-05-31 20:00:00</td>\n",
       "      <td>-0.000321</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 21:00:00</th>\n",
       "      <td>31.05.2018 21:00:00.000</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29577</td>\n",
       "      <td>1.29498</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>1.456650e+09</td>\n",
       "      <td>2018-05-31 21:00:00</td>\n",
       "      <td>0.049502</td>\n",
       "      <td>-0.000321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 22:00:00</th>\n",
       "      <td>31.05.2018 22:00:00.000</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29407</td>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.370330e+09</td>\n",
       "      <td>2018-05-31 22:00:00</td>\n",
       "      <td>-0.008679</td>\n",
       "      <td>0.049502</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 23:00:00</th>\n",
       "      <td>31.05.2018 23:00:00.000</td>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.29469</td>\n",
       "      <td>1.29406</td>\n",
       "      <td>1.29436</td>\n",
       "      <td>1.259850e+09</td>\n",
       "      <td>2018-05-31 23:00:00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.008679</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    Gmt time     Open     High      Low  \\\n",
       "2018-05-31 14:00:00  31.05.2018 14:00:00.000  1.29518  1.29770  1.29430   \n",
       "2018-05-31 15:00:00  31.05.2018 15:00:00.000  1.29716  1.29903  1.29586   \n",
       "2018-05-31 16:00:00  31.05.2018 16:00:00.000  1.29699  1.29750  1.29555   \n",
       "2018-05-31 17:00:00  31.05.2018 17:00:00.000  1.29560  1.29803  1.29424   \n",
       "2018-05-31 18:00:00  31.05.2018 18:00:00.000  1.29785  1.29884  1.29538   \n",
       "2018-05-31 19:00:00  31.05.2018 19:00:00.000  1.29566  1.29671  1.29541   \n",
       "2018-05-31 20:00:00  31.05.2018 20:00:00.000  1.29668  1.29682  1.29502   \n",
       "2018-05-31 21:00:00  31.05.2018 21:00:00.000  1.29563  1.29577  1.29498   \n",
       "2018-05-31 22:00:00  31.05.2018 22:00:00.000  1.29563  1.29563  1.29407   \n",
       "2018-05-31 23:00:00  31.05.2018 23:00:00.000  1.29409  1.29469  1.29406   \n",
       "\n",
       "                       Close        Volume            datetime  return_next  \\\n",
       "2018-05-31 14:00:00  1.29718  9.841480e+09 2018-05-31 14:00:00     0.005464   \n",
       "2018-05-31 15:00:00  1.29699  1.169860e+10 2018-05-31 15:00:00     0.045002   \n",
       "2018-05-31 16:00:00  1.29559  7.277670e+09 2018-05-31 16:00:00    -0.072324   \n",
       "2018-05-31 17:00:00  1.29785  7.262840e+09 2018-05-31 17:00:00     0.071038   \n",
       "2018-05-31 18:00:00  1.29564  6.044530e+09 2018-05-31 18:00:00    -0.032144   \n",
       "2018-05-31 19:00:00  1.29666  5.330320e+09 2018-05-31 19:00:00     0.033751   \n",
       "2018-05-31 20:00:00  1.29563  3.116830e+09 2018-05-31 20:00:00    -0.000321   \n",
       "2018-05-31 21:00:00  1.29564  1.456650e+09 2018-05-31 21:00:00     0.049502   \n",
       "2018-05-31 22:00:00  1.29409  1.370330e+09 2018-05-31 22:00:00    -0.008679   \n",
       "2018-05-31 23:00:00  1.29436  1.259850e+09 2018-05-31 23:00:00          NaN   \n",
       "\n",
       "                       return  label  \n",
       "2018-05-31 14:00:00 -0.064288      1  \n",
       "2018-05-31 15:00:00  0.005464      1  \n",
       "2018-05-31 16:00:00  0.045002      0  \n",
       "2018-05-31 17:00:00 -0.072324      1  \n",
       "2018-05-31 18:00:00  0.071038      0  \n",
       "2018-05-31 19:00:00 -0.032144      1  \n",
       "2018-05-31 20:00:00  0.033751      0  \n",
       "2018-05-31 21:00:00 -0.000321      1  \n",
       "2018-05-31 22:00:00  0.049502      0  \n",
       "2018-05-31 23:00:00 -0.008679      0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Make label, 1 as rising price, 0 as falling price - prediction for the following timestamp\n",
    "# ****************** consider when return ==0 next version\n",
    "\n",
    "df['label'] = df['return_next'].apply(lambda x: 1 if x>0.0 else 0)\n",
    "df.tail(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gmt time</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>datetime</th>\n",
       "      <th>return_next</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-31 18:00:00</th>\n",
       "      <td>31.05.2018 18:00:00.000</td>\n",
       "      <td>1.29785</td>\n",
       "      <td>1.29884</td>\n",
       "      <td>1.29538</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>6.044530e+09</td>\n",
       "      <td>2018-05-31 18:00:00</td>\n",
       "      <td>-0.032144</td>\n",
       "      <td>0.071038</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 19:00:00</th>\n",
       "      <td>31.05.2018 19:00:00.000</td>\n",
       "      <td>1.29566</td>\n",
       "      <td>1.29671</td>\n",
       "      <td>1.29541</td>\n",
       "      <td>1.29666</td>\n",
       "      <td>5.330320e+09</td>\n",
       "      <td>2018-05-31 19:00:00</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>-0.032144</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 20:00:00</th>\n",
       "      <td>31.05.2018 20:00:00.000</td>\n",
       "      <td>1.29668</td>\n",
       "      <td>1.29682</td>\n",
       "      <td>1.29502</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>3.116830e+09</td>\n",
       "      <td>2018-05-31 20:00:00</td>\n",
       "      <td>-0.000321</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 21:00:00</th>\n",
       "      <td>31.05.2018 21:00:00.000</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29577</td>\n",
       "      <td>1.29498</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>1.456650e+09</td>\n",
       "      <td>2018-05-31 21:00:00</td>\n",
       "      <td>0.049502</td>\n",
       "      <td>-0.000321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 22:00:00</th>\n",
       "      <td>31.05.2018 22:00:00.000</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29407</td>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.370330e+09</td>\n",
       "      <td>2018-05-31 22:00:00</td>\n",
       "      <td>-0.008679</td>\n",
       "      <td>0.049502</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    Gmt time     Open     High      Low  \\\n",
       "2018-05-31 18:00:00  31.05.2018 18:00:00.000  1.29785  1.29884  1.29538   \n",
       "2018-05-31 19:00:00  31.05.2018 19:00:00.000  1.29566  1.29671  1.29541   \n",
       "2018-05-31 20:00:00  31.05.2018 20:00:00.000  1.29668  1.29682  1.29502   \n",
       "2018-05-31 21:00:00  31.05.2018 21:00:00.000  1.29563  1.29577  1.29498   \n",
       "2018-05-31 22:00:00  31.05.2018 22:00:00.000  1.29563  1.29563  1.29407   \n",
       "\n",
       "                       Close        Volume            datetime  return_next  \\\n",
       "2018-05-31 18:00:00  1.29564  6.044530e+09 2018-05-31 18:00:00    -0.032144   \n",
       "2018-05-31 19:00:00  1.29666  5.330320e+09 2018-05-31 19:00:00     0.033751   \n",
       "2018-05-31 20:00:00  1.29563  3.116830e+09 2018-05-31 20:00:00    -0.000321   \n",
       "2018-05-31 21:00:00  1.29564  1.456650e+09 2018-05-31 21:00:00     0.049502   \n",
       "2018-05-31 22:00:00  1.29409  1.370330e+09 2018-05-31 22:00:00    -0.008679   \n",
       "\n",
       "                       return  label  \n",
       "2018-05-31 18:00:00  0.071038      0  \n",
       "2018-05-31 19:00:00 -0.032144      1  \n",
       "2018-05-31 20:00:00  0.033751      0  \n",
       "2018-05-31 21:00:00 -0.000321      1  \n",
       "2018-05-31 22:00:00  0.049502      0  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The last register shows a NaN\n",
    "\n",
    "df.dropna(inplace=True)\n",
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-31 18:00:00</th>\n",
       "      <td>1.29785</td>\n",
       "      <td>1.29884</td>\n",
       "      <td>1.29538</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>6.044530e+09</td>\n",
       "      <td>0.071038</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 19:00:00</th>\n",
       "      <td>1.29566</td>\n",
       "      <td>1.29671</td>\n",
       "      <td>1.29541</td>\n",
       "      <td>1.29666</td>\n",
       "      <td>5.330320e+09</td>\n",
       "      <td>-0.032144</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 20:00:00</th>\n",
       "      <td>1.29668</td>\n",
       "      <td>1.29682</td>\n",
       "      <td>1.29502</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>3.116830e+09</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 21:00:00</th>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29577</td>\n",
       "      <td>1.29498</td>\n",
       "      <td>1.29564</td>\n",
       "      <td>1.456650e+09</td>\n",
       "      <td>-0.000321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 22:00:00</th>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29563</td>\n",
       "      <td>1.29407</td>\n",
       "      <td>1.29409</td>\n",
       "      <td>1.370330e+09</td>\n",
       "      <td>0.049502</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close        Volume  \\\n",
       "2018-05-31 18:00:00  1.29785  1.29884  1.29538  1.29564  6.044530e+09   \n",
       "2018-05-31 19:00:00  1.29566  1.29671  1.29541  1.29666  5.330320e+09   \n",
       "2018-05-31 20:00:00  1.29668  1.29682  1.29502  1.29563  3.116830e+09   \n",
       "2018-05-31 21:00:00  1.29563  1.29577  1.29498  1.29564  1.456650e+09   \n",
       "2018-05-31 22:00:00  1.29563  1.29563  1.29407  1.29409  1.370330e+09   \n",
       "\n",
       "                       return  label  \n",
       "2018-05-31 18:00:00  0.071038      0  \n",
       "2018-05-31 19:00:00 -0.032144      1  \n",
       "2018-05-31 20:00:00  0.033751      0  \n",
       "2018-05-31 21:00:00 -0.000321      1  \n",
       "2018-05-31 22:00:00  0.049502      0  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = df.copy()\n",
    "\n",
    "df = df.loc[:,['Open', 'High', 'Low', 'Close', 'Volume', 'return', 'label']]\n",
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016-01-01 00:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 01:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 02:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 03:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 04:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close  Volume  return  \\\n",
       "2016-01-01 00:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 01:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 02:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 03:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 04:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "\n",
       "                     label  Open_oil  High_oil  Low_oil  Close_oil  Volume_oil  \n",
       "2016-01-01 00:00:00      0   1.38326   1.38326  1.38326    1.38326         0.0  \n",
       "2016-01-01 01:00:00      0   1.38326   1.38326  1.38326    1.38326         0.0  \n",
       "2016-01-01 02:00:00      0   1.38326   1.38326  1.38326    1.38326         0.0  \n",
       "2016-01-01 03:00:00      0   1.38326   1.38326  1.38326    1.38326         0.0  \n",
       "2016-01-01 04:00:00      0   1.38326   1.38326  1.38326    1.38326         0.0  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.concat([df,df_oil],axis=1, join='inner')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016-01-01 00:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 01:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 02:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 03:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 04:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close  Volume  return  \\\n",
       "2016-01-01 00:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 01:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 02:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 03:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 04:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "\n",
       "                     label  Open_oil  High_oil  Low_oil  Close_oil  \\\n",
       "2016-01-01 00:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 01:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 02:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 03:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 04:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "\n",
       "                     Volume_oil  year  month  day  hour  \n",
       "2016-01-01 00:00:00         0.0  2016      1    1     0  \n",
       "2016-01-01 01:00:00         0.0  2016      1    1     1  \n",
       "2016-01-01 02:00:00         0.0  2016      1    1     2  \n",
       "2016-01-01 03:00:00         0.0  2016      1    1     3  \n",
       "2016-01-01 04:00:00         0.0  2016      1    1     4  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating datetime features\n",
    "\n",
    "\n",
    "df['year'] = df.index.year\n",
    "df['month'] = df.index.month\n",
    "df['day'] = df.index.day\n",
    "\n",
    "df['hour'] = df.index.hour\n",
    "\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Open          float64\n",
       "High          float64\n",
       "Low           float64\n",
       "Close         float64\n",
       "Volume        float64\n",
       "return        float64\n",
       "label           int64\n",
       "Open_oil      float64\n",
       "High_oil      float64\n",
       "Low_oil       float64\n",
       "Close_oil     float64\n",
       "Volume_oil    float64\n",
       "year            int64\n",
       "month           int64\n",
       "day             int64\n",
       "hour            int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes\n",
    "\n",
    "# we have to change the volume type later"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Choosing a timeframe to trade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-30 03:00:00</th>\n",
       "      <td>1.30247</td>\n",
       "      <td>1.30350</td>\n",
       "      <td>1.30238</td>\n",
       "      <td>1.30303</td>\n",
       "      <td>2.623410e+09</td>\n",
       "      <td>-0.018001</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30247</td>\n",
       "      <td>1.30350</td>\n",
       "      <td>1.30238</td>\n",
       "      <td>1.30303</td>\n",
       "      <td>2.623410e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 04:00:00</th>\n",
       "      <td>1.30306</td>\n",
       "      <td>1.30314</td>\n",
       "      <td>1.30230</td>\n",
       "      <td>1.30260</td>\n",
       "      <td>1.692350e+09</td>\n",
       "      <td>0.014786</td>\n",
       "      <td>0</td>\n",
       "      <td>1.30306</td>\n",
       "      <td>1.30314</td>\n",
       "      <td>1.30230</td>\n",
       "      <td>1.30260</td>\n",
       "      <td>1.692350e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 05:00:00</th>\n",
       "      <td>1.30259</td>\n",
       "      <td>1.30398</td>\n",
       "      <td>1.30234</td>\n",
       "      <td>1.30262</td>\n",
       "      <td>3.642960e+09</td>\n",
       "      <td>-0.000964</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30259</td>\n",
       "      <td>1.30398</td>\n",
       "      <td>1.30234</td>\n",
       "      <td>1.30262</td>\n",
       "      <td>3.642960e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 06:00:00</th>\n",
       "      <td>1.30266</td>\n",
       "      <td>1.30302</td>\n",
       "      <td>1.30143</td>\n",
       "      <td>1.30222</td>\n",
       "      <td>7.352290e+09</td>\n",
       "      <td>0.014143</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30266</td>\n",
       "      <td>1.30302</td>\n",
       "      <td>1.30143</td>\n",
       "      <td>1.30222</td>\n",
       "      <td>7.352290e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 07:00:00</th>\n",
       "      <td>1.30218</td>\n",
       "      <td>1.30279</td>\n",
       "      <td>1.30109</td>\n",
       "      <td>1.30148</td>\n",
       "      <td>7.621840e+09</td>\n",
       "      <td>0.022501</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30218</td>\n",
       "      <td>1.30279</td>\n",
       "      <td>1.30109</td>\n",
       "      <td>1.30148</td>\n",
       "      <td>7.621840e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 08:00:00</th>\n",
       "      <td>1.30152</td>\n",
       "      <td>1.30161</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>7.426750e+09</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30152</td>\n",
       "      <td>1.30161</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>7.426750e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 09:00:00</th>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30049</td>\n",
       "      <td>1.29931</td>\n",
       "      <td>1.29958</td>\n",
       "      <td>7.103510e+09</td>\n",
       "      <td>0.028608</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30049</td>\n",
       "      <td>1.29931</td>\n",
       "      <td>1.29958</td>\n",
       "      <td>7.103510e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 10:00:00</th>\n",
       "      <td>1.29959</td>\n",
       "      <td>1.29999</td>\n",
       "      <td>1.29813</td>\n",
       "      <td>1.29861</td>\n",
       "      <td>8.318910e+09</td>\n",
       "      <td>0.031501</td>\n",
       "      <td>1</td>\n",
       "      <td>1.29959</td>\n",
       "      <td>1.29999</td>\n",
       "      <td>1.29813</td>\n",
       "      <td>1.29861</td>\n",
       "      <td>8.318910e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 11:00:00</th>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29885</td>\n",
       "      <td>1.29705</td>\n",
       "      <td>1.29715</td>\n",
       "      <td>6.624660e+09</td>\n",
       "      <td>0.047573</td>\n",
       "      <td>0</td>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29885</td>\n",
       "      <td>1.29705</td>\n",
       "      <td>1.29715</td>\n",
       "      <td>6.624660e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 12:00:00</th>\n",
       "      <td>1.29717</td>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29527</td>\n",
       "      <td>1.29781</td>\n",
       "      <td>1.067015e+10</td>\n",
       "      <td>-0.020572</td>\n",
       "      <td>0</td>\n",
       "      <td>1.29717</td>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29527</td>\n",
       "      <td>1.29781</td>\n",
       "      <td>1.067015e+10</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 03:00:00</th>\n",
       "      <td>1.28779</td>\n",
       "      <td>1.28788</td>\n",
       "      <td>1.28669</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>2.857670e+09</td>\n",
       "      <td>0.019286</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28779</td>\n",
       "      <td>1.28788</td>\n",
       "      <td>1.28669</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>2.857670e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 04:00:00</th>\n",
       "      <td>1.28719</td>\n",
       "      <td>1.28725</td>\n",
       "      <td>1.28647</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.907510e+09</td>\n",
       "      <td>0.013500</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>1.28725</td>\n",
       "      <td>1.28647</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.907510e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 05:00:00</th>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.28707</td>\n",
       "      <td>1.28598</td>\n",
       "      <td>1.28673</td>\n",
       "      <td>3.887940e+09</td>\n",
       "      <td>0.001286</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.28707</td>\n",
       "      <td>1.28598</td>\n",
       "      <td>1.28673</td>\n",
       "      <td>3.887940e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 06:00:00</th>\n",
       "      <td>1.28678</td>\n",
       "      <td>1.28758</td>\n",
       "      <td>1.28654</td>\n",
       "      <td>1.28734</td>\n",
       "      <td>6.998760e+09</td>\n",
       "      <td>-0.018001</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28678</td>\n",
       "      <td>1.28758</td>\n",
       "      <td>1.28654</td>\n",
       "      <td>1.28734</td>\n",
       "      <td>6.998760e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 07:00:00</th>\n",
       "      <td>1.28735</td>\n",
       "      <td>1.28747</td>\n",
       "      <td>1.28452</td>\n",
       "      <td>1.28482</td>\n",
       "      <td>7.630360e+09</td>\n",
       "      <td>0.081324</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28735</td>\n",
       "      <td>1.28747</td>\n",
       "      <td>1.28452</td>\n",
       "      <td>1.28482</td>\n",
       "      <td>7.630360e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 08:00:00</th>\n",
       "      <td>1.28484</td>\n",
       "      <td>1.28538</td>\n",
       "      <td>1.28356</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>5.385360e+09</td>\n",
       "      <td>0.027001</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28484</td>\n",
       "      <td>1.28538</td>\n",
       "      <td>1.28356</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>5.385360e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 09:00:00</th>\n",
       "      <td>1.28400</td>\n",
       "      <td>1.28539</td>\n",
       "      <td>1.28394</td>\n",
       "      <td>1.28470</td>\n",
       "      <td>5.552970e+09</td>\n",
       "      <td>-0.022501</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>1.28539</td>\n",
       "      <td>1.28394</td>\n",
       "      <td>1.28470</td>\n",
       "      <td>5.552970e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 10:00:00</th>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28583</td>\n",
       "      <td>1.28180</td>\n",
       "      <td>1.28259</td>\n",
       "      <td>5.881380e+09</td>\n",
       "      <td>0.068467</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28583</td>\n",
       "      <td>1.28180</td>\n",
       "      <td>1.28259</td>\n",
       "      <td>5.881380e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 11:00:00</th>\n",
       "      <td>1.28264</td>\n",
       "      <td>1.28514</td>\n",
       "      <td>1.28251</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>6.608530e+09</td>\n",
       "      <td>-0.067502</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28264</td>\n",
       "      <td>1.28514</td>\n",
       "      <td>1.28251</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>6.608530e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 12:00:00</th>\n",
       "      <td>1.28474</td>\n",
       "      <td>1.28987</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28985</td>\n",
       "      <td>1.079586e+10</td>\n",
       "      <td>-0.164256</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>1.28987</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28985</td>\n",
       "      <td>1.079586e+10</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close        Volume  \\\n",
       "2018-05-30 03:00:00  1.30247  1.30350  1.30238  1.30303  2.623410e+09   \n",
       "2018-05-30 04:00:00  1.30306  1.30314  1.30230  1.30260  1.692350e+09   \n",
       "2018-05-30 05:00:00  1.30259  1.30398  1.30234  1.30262  3.642960e+09   \n",
       "2018-05-30 06:00:00  1.30266  1.30302  1.30143  1.30222  7.352290e+09   \n",
       "2018-05-30 07:00:00  1.30218  1.30279  1.30109  1.30148  7.621840e+09   \n",
       "2018-05-30 08:00:00  1.30152  1.30161  1.30047  1.30047  7.426750e+09   \n",
       "2018-05-30 09:00:00  1.30047  1.30049  1.29931  1.29958  7.103510e+09   \n",
       "2018-05-30 10:00:00  1.29959  1.29999  1.29813  1.29861  8.318910e+09   \n",
       "2018-05-30 11:00:00  1.29863  1.29885  1.29705  1.29715  6.624660e+09   \n",
       "2018-05-30 12:00:00  1.29717  1.29863  1.29527  1.29781  1.067015e+10   \n",
       "2018-05-31 03:00:00  1.28779  1.28788  1.28669  1.28719  2.857670e+09   \n",
       "2018-05-31 04:00:00  1.28719  1.28725  1.28647  1.28677  1.907510e+09   \n",
       "2018-05-31 05:00:00  1.28677  1.28707  1.28598  1.28673  3.887940e+09   \n",
       "2018-05-31 06:00:00  1.28678  1.28758  1.28654  1.28734  6.998760e+09   \n",
       "2018-05-31 07:00:00  1.28735  1.28747  1.28452  1.28482  7.630360e+09   \n",
       "2018-05-31 08:00:00  1.28484  1.28538  1.28356  1.28400  5.385360e+09   \n",
       "2018-05-31 09:00:00  1.28400  1.28539  1.28394  1.28470  5.552970e+09   \n",
       "2018-05-31 10:00:00  1.28472  1.28583  1.28180  1.28259  5.881380e+09   \n",
       "2018-05-31 11:00:00  1.28264  1.28514  1.28251  1.28474  6.608530e+09   \n",
       "2018-05-31 12:00:00  1.28474  1.28987  1.28472  1.28985  1.079586e+10   \n",
       "\n",
       "                       return  label  Open_oil  High_oil  Low_oil  Close_oil  \\\n",
       "2018-05-30 03:00:00 -0.018001      1   1.30247   1.30350  1.30238    1.30303   \n",
       "2018-05-30 04:00:00  0.014786      0   1.30306   1.30314  1.30230    1.30260   \n",
       "2018-05-30 05:00:00 -0.000964      1   1.30259   1.30398  1.30234    1.30262   \n",
       "2018-05-30 06:00:00  0.014143      1   1.30266   1.30302  1.30143    1.30222   \n",
       "2018-05-30 07:00:00  0.022501      1   1.30218   1.30279  1.30109    1.30148   \n",
       "2018-05-30 08:00:00  0.033751      1   1.30152   1.30161  1.30047    1.30047   \n",
       "2018-05-30 09:00:00  0.028608      1   1.30047   1.30049  1.29931    1.29958   \n",
       "2018-05-30 10:00:00  0.031501      1   1.29959   1.29999  1.29813    1.29861   \n",
       "2018-05-30 11:00:00  0.047573      0   1.29863   1.29885  1.29705    1.29715   \n",
       "2018-05-30 12:00:00 -0.020572      0   1.29717   1.29863  1.29527    1.29781   \n",
       "2018-05-31 03:00:00  0.019286      1   1.28779   1.28788  1.28669    1.28719   \n",
       "2018-05-31 04:00:00  0.013500      1   1.28719   1.28725  1.28647    1.28677   \n",
       "2018-05-31 05:00:00  0.001286      0   1.28677   1.28707  1.28598    1.28673   \n",
       "2018-05-31 06:00:00 -0.018001      1   1.28678   1.28758  1.28654    1.28734   \n",
       "2018-05-31 07:00:00  0.081324      1   1.28735   1.28747  1.28452    1.28482   \n",
       "2018-05-31 08:00:00  0.027001      0   1.28484   1.28538  1.28356    1.28400   \n",
       "2018-05-31 09:00:00 -0.022501      1   1.28400   1.28539  1.28394    1.28470   \n",
       "2018-05-31 10:00:00  0.068467      0   1.28472   1.28583  1.28180    1.28259   \n",
       "2018-05-31 11:00:00 -0.067502      0   1.28264   1.28514  1.28251    1.28474   \n",
       "2018-05-31 12:00:00 -0.164256      0   1.28474   1.28987  1.28472    1.28985   \n",
       "\n",
       "                       Volume_oil  year  month  day  hour  \n",
       "2018-05-30 03:00:00  2.623410e+09  2018      5   30     3  \n",
       "2018-05-30 04:00:00  1.692350e+09  2018      5   30     4  \n",
       "2018-05-30 05:00:00  3.642960e+09  2018      5   30     5  \n",
       "2018-05-30 06:00:00  7.352290e+09  2018      5   30     6  \n",
       "2018-05-30 07:00:00  7.621840e+09  2018      5   30     7  \n",
       "2018-05-30 08:00:00  7.426750e+09  2018      5   30     8  \n",
       "2018-05-30 09:00:00  7.103510e+09  2018      5   30     9  \n",
       "2018-05-30 10:00:00  8.318910e+09  2018      5   30    10  \n",
       "2018-05-30 11:00:00  6.624660e+09  2018      5   30    11  \n",
       "2018-05-30 12:00:00  1.067015e+10  2018      5   30    12  \n",
       "2018-05-31 03:00:00  2.857670e+09  2018      5   31     3  \n",
       "2018-05-31 04:00:00  1.907510e+09  2018      5   31     4  \n",
       "2018-05-31 05:00:00  3.887940e+09  2018      5   31     5  \n",
       "2018-05-31 06:00:00  6.998760e+09  2018      5   31     6  \n",
       "2018-05-31 07:00:00  7.630360e+09  2018      5   31     7  \n",
       "2018-05-31 08:00:00  5.385360e+09  2018      5   31     8  \n",
       "2018-05-31 09:00:00  5.552970e+09  2018      5   31     9  \n",
       "2018-05-31 10:00:00  5.881380e+09  2018      5   31    10  \n",
       "2018-05-31 11:00:00  6.608530e+09  2018      5   31    11  \n",
       "2018-05-31 12:00:00  1.079586e+10  2018      5   31    12  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Limiting the dataframe: from 3:00 to 12:00 (10 timestamps) \n",
    "\n",
    "df = df.between_time(include_start=True,include_end=True, start_time='03:00:00', end_time='12:00:00')\n",
    "df.tail(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-31 03:00:00</th>\n",
       "      <td>1.28779</td>\n",
       "      <td>1.28788</td>\n",
       "      <td>1.28669</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>2.857670e+09</td>\n",
       "      <td>0.019286</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28779</td>\n",
       "      <td>1.28788</td>\n",
       "      <td>1.28669</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>2.857670e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 04:00:00</th>\n",
       "      <td>1.28719</td>\n",
       "      <td>1.28725</td>\n",
       "      <td>1.28647</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.907510e+09</td>\n",
       "      <td>0.013500</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>1.28725</td>\n",
       "      <td>1.28647</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.907510e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 05:00:00</th>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.28707</td>\n",
       "      <td>1.28598</td>\n",
       "      <td>1.28673</td>\n",
       "      <td>3.887940e+09</td>\n",
       "      <td>0.001286</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.28707</td>\n",
       "      <td>1.28598</td>\n",
       "      <td>1.28673</td>\n",
       "      <td>3.887940e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 06:00:00</th>\n",
       "      <td>1.28678</td>\n",
       "      <td>1.28758</td>\n",
       "      <td>1.28654</td>\n",
       "      <td>1.28734</td>\n",
       "      <td>6.998760e+09</td>\n",
       "      <td>-0.018001</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28678</td>\n",
       "      <td>1.28758</td>\n",
       "      <td>1.28654</td>\n",
       "      <td>1.28734</td>\n",
       "      <td>6.998760e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 07:00:00</th>\n",
       "      <td>1.28735</td>\n",
       "      <td>1.28747</td>\n",
       "      <td>1.28452</td>\n",
       "      <td>1.28482</td>\n",
       "      <td>7.630360e+09</td>\n",
       "      <td>0.081324</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28735</td>\n",
       "      <td>1.28747</td>\n",
       "      <td>1.28452</td>\n",
       "      <td>1.28482</td>\n",
       "      <td>7.630360e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close        Volume  \\\n",
       "2018-05-31 03:00:00  1.28779  1.28788  1.28669  1.28719  2.857670e+09   \n",
       "2018-05-31 04:00:00  1.28719  1.28725  1.28647  1.28677  1.907510e+09   \n",
       "2018-05-31 05:00:00  1.28677  1.28707  1.28598  1.28673  3.887940e+09   \n",
       "2018-05-31 06:00:00  1.28678  1.28758  1.28654  1.28734  6.998760e+09   \n",
       "2018-05-31 07:00:00  1.28735  1.28747  1.28452  1.28482  7.630360e+09   \n",
       "\n",
       "                       return  label  Open_oil  High_oil  Low_oil  Close_oil  \\\n",
       "2018-05-31 03:00:00  0.019286      1   1.28779   1.28788  1.28669    1.28719   \n",
       "2018-05-31 04:00:00  0.013500      1   1.28719   1.28725  1.28647    1.28677   \n",
       "2018-05-31 05:00:00  0.001286      0   1.28677   1.28707  1.28598    1.28673   \n",
       "2018-05-31 06:00:00 -0.018001      1   1.28678   1.28758  1.28654    1.28734   \n",
       "2018-05-31 07:00:00  0.081324      1   1.28735   1.28747  1.28452    1.28482   \n",
       "\n",
       "                       Volume_oil  year  month  day  hour  \n",
       "2018-05-31 03:00:00  2.857670e+09  2018      5   31     3  \n",
       "2018-05-31 04:00:00  1.907510e+09  2018      5   31     4  \n",
       "2018-05-31 05:00:00  3.887940e+09  2018      5   31     5  \n",
       "2018-05-31 06:00:00  6.998760e+09  2018      5   31     6  \n",
       "2018-05-31 07:00:00  7.630360e+09  2018      5   31     7  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-31 08:00:00</th>\n",
       "      <td>1.28484</td>\n",
       "      <td>1.28538</td>\n",
       "      <td>1.28356</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>5.385360e+09</td>\n",
       "      <td>0.027001</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28484</td>\n",
       "      <td>1.28538</td>\n",
       "      <td>1.28356</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>5.385360e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 09:00:00</th>\n",
       "      <td>1.28400</td>\n",
       "      <td>1.28539</td>\n",
       "      <td>1.28394</td>\n",
       "      <td>1.28470</td>\n",
       "      <td>5.552970e+09</td>\n",
       "      <td>-0.022501</td>\n",
       "      <td>1</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>1.28539</td>\n",
       "      <td>1.28394</td>\n",
       "      <td>1.28470</td>\n",
       "      <td>5.552970e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 10:00:00</th>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28583</td>\n",
       "      <td>1.28180</td>\n",
       "      <td>1.28259</td>\n",
       "      <td>5.881380e+09</td>\n",
       "      <td>0.068467</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28583</td>\n",
       "      <td>1.28180</td>\n",
       "      <td>1.28259</td>\n",
       "      <td>5.881380e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 11:00:00</th>\n",
       "      <td>1.28264</td>\n",
       "      <td>1.28514</td>\n",
       "      <td>1.28251</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>6.608530e+09</td>\n",
       "      <td>-0.067502</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28264</td>\n",
       "      <td>1.28514</td>\n",
       "      <td>1.28251</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>6.608530e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 12:00:00</th>\n",
       "      <td>1.28474</td>\n",
       "      <td>1.28987</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28985</td>\n",
       "      <td>1.079586e+10</td>\n",
       "      <td>-0.164256</td>\n",
       "      <td>0</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>1.28987</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28985</td>\n",
       "      <td>1.079586e+10</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close        Volume  \\\n",
       "2018-05-31 08:00:00  1.28484  1.28538  1.28356  1.28400  5.385360e+09   \n",
       "2018-05-31 09:00:00  1.28400  1.28539  1.28394  1.28470  5.552970e+09   \n",
       "2018-05-31 10:00:00  1.28472  1.28583  1.28180  1.28259  5.881380e+09   \n",
       "2018-05-31 11:00:00  1.28264  1.28514  1.28251  1.28474  6.608530e+09   \n",
       "2018-05-31 12:00:00  1.28474  1.28987  1.28472  1.28985  1.079586e+10   \n",
       "\n",
       "                       return  label  Open_oil  High_oil  Low_oil  Close_oil  \\\n",
       "2018-05-31 08:00:00  0.027001      0   1.28484   1.28538  1.28356    1.28400   \n",
       "2018-05-31 09:00:00 -0.022501      1   1.28400   1.28539  1.28394    1.28470   \n",
       "2018-05-31 10:00:00  0.068467      0   1.28472   1.28583  1.28180    1.28259   \n",
       "2018-05-31 11:00:00 -0.067502      0   1.28264   1.28514  1.28251    1.28474   \n",
       "2018-05-31 12:00:00 -0.164256      0   1.28474   1.28987  1.28472    1.28985   \n",
       "\n",
       "                       Volume_oil  year  month  day  hour  \n",
       "2018-05-31 08:00:00  5.385360e+09  2018      5   31     8  \n",
       "2018-05-31 09:00:00  5.552970e+09  2018      5   31     9  \n",
       "2018-05-31 10:00:00  5.881380e+09  2018      5   31    10  \n",
       "2018-05-31 11:00:00  6.608530e+09  2018      5   31    11  \n",
       "2018-05-31 12:00:00  1.079586e+10  2018      5   31    12  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016-01-01 03:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 04:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 05:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 06:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-01 07:00:00</th>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>1.38326</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close  Volume  return  \\\n",
       "2016-01-01 03:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 04:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 05:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 06:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "2016-01-01 07:00:00  1.38326  1.38326  1.38326  1.38326     0.0     0.0   \n",
       "\n",
       "                     label  Open_oil  High_oil  Low_oil  Close_oil  \\\n",
       "2016-01-01 03:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 04:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 05:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 06:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "2016-01-01 07:00:00      0   1.38326   1.38326  1.38326    1.38326   \n",
       "\n",
       "                     Volume_oil  year  month  day  hour  \n",
       "2016-01-01 03:00:00         0.0  2016      1    1     3  \n",
       "2016-01-01 04:00:00         0.0  2016      1    1     4  \n",
       "2016-01-01 05:00:00         0.0  2016      1    1     5  \n",
       "2016-01-01 06:00:00         0.0  2016      1    1     6  \n",
       "2016-01-01 07:00:00         0.0  2016      1    1     7  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>label</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-30 08:00:00</th>\n",
       "      <td>1.30152</td>\n",
       "      <td>1.30161</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>7.426750e+09</td>\n",
       "      <td>0.033751</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30152</td>\n",
       "      <td>1.30161</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>7.426750e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 09:00:00</th>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30049</td>\n",
       "      <td>1.29931</td>\n",
       "      <td>1.29958</td>\n",
       "      <td>7.103510e+09</td>\n",
       "      <td>0.028608</td>\n",
       "      <td>1</td>\n",
       "      <td>1.30047</td>\n",
       "      <td>1.30049</td>\n",
       "      <td>1.29931</td>\n",
       "      <td>1.29958</td>\n",
       "      <td>7.103510e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 10:00:00</th>\n",
       "      <td>1.29959</td>\n",
       "      <td>1.29999</td>\n",
       "      <td>1.29813</td>\n",
       "      <td>1.29861</td>\n",
       "      <td>8.318910e+09</td>\n",
       "      <td>0.031501</td>\n",
       "      <td>1</td>\n",
       "      <td>1.29959</td>\n",
       "      <td>1.29999</td>\n",
       "      <td>1.29813</td>\n",
       "      <td>1.29861</td>\n",
       "      <td>8.318910e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 11:00:00</th>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29885</td>\n",
       "      <td>1.29705</td>\n",
       "      <td>1.29715</td>\n",
       "      <td>6.624660e+09</td>\n",
       "      <td>0.047573</td>\n",
       "      <td>0</td>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29885</td>\n",
       "      <td>1.29705</td>\n",
       "      <td>1.29715</td>\n",
       "      <td>6.624660e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-30 12:00:00</th>\n",
       "      <td>1.29717</td>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29527</td>\n",
       "      <td>1.29781</td>\n",
       "      <td>1.067015e+10</td>\n",
       "      <td>-0.020572</td>\n",
       "      <td>0</td>\n",
       "      <td>1.29717</td>\n",
       "      <td>1.29863</td>\n",
       "      <td>1.29527</td>\n",
       "      <td>1.29781</td>\n",
       "      <td>1.067015e+10</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close        Volume  \\\n",
       "2018-05-30 08:00:00  1.30152  1.30161  1.30047  1.30047  7.426750e+09   \n",
       "2018-05-30 09:00:00  1.30047  1.30049  1.29931  1.29958  7.103510e+09   \n",
       "2018-05-30 10:00:00  1.29959  1.29999  1.29813  1.29861  8.318910e+09   \n",
       "2018-05-30 11:00:00  1.29863  1.29885  1.29705  1.29715  6.624660e+09   \n",
       "2018-05-30 12:00:00  1.29717  1.29863  1.29527  1.29781  1.067015e+10   \n",
       "\n",
       "                       return  label  Open_oil  High_oil  Low_oil  Close_oil  \\\n",
       "2018-05-30 08:00:00  0.033751      1   1.30152   1.30161  1.30047    1.30047   \n",
       "2018-05-30 09:00:00  0.028608      1   1.30047   1.30049  1.29931    1.29958   \n",
       "2018-05-30 10:00:00  0.031501      1   1.29959   1.29999  1.29813    1.29861   \n",
       "2018-05-30 11:00:00  0.047573      0   1.29863   1.29885  1.29705    1.29715   \n",
       "2018-05-30 12:00:00 -0.020572      0   1.29717   1.29863  1.29527    1.29781   \n",
       "\n",
       "                       Volume_oil  year  month  day  hour  \n",
       "2018-05-30 08:00:00  7.426750e+09  2018      5   30     8  \n",
       "2018-05-30 09:00:00  7.103510e+09  2018      5   30     9  \n",
       "2018-05-30 10:00:00  8.318910e+09  2018      5   30    10  \n",
       "2018-05-30 11:00:00  6.624660e+09  2018      5   30    11  \n",
       "2018-05-30 12:00:00  1.067015e+10  2018      5   30    12  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Make training and test dataset\n",
    "\n",
    "df_test = df.loc['2018-05-31' ,:]  \n",
    "display(df_test.head())\n",
    "display(df_test.tail())\n",
    "\n",
    "df_train = df.loc[:'2018-05-30',:] \n",
    "display(df_train.head())\n",
    "display(df_train.tail())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Open', 'High', 'Low', 'Close', 'Volume', 'return', 'label', 'Open_oil',\n",
       "       'High_oil', 'Low_oil', 'Close_oil', 'Volume_oil', 'year', 'month',\n",
       "       'day', 'hour'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Applying a model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Separate features and output\n",
    "\n",
    "y_train = df_train['label']\n",
    "X_train = df_train.loc[:,['Open', 'High', 'Low', 'Close', 'Volume', 'return', 'Open_oil', 'High_oil', 'Low_oil', 'Close_oil', 'Volume_oil',\n",
    "       'year', 'month', 'day', 'hour']]\n",
    "\n",
    "y_test = df_test['label']\n",
    "X_test = df_test.loc[:,['Open', 'High', 'Low', 'Close', 'Volume', 'return', 'Open_oil', 'High_oil', 'Low_oil', 'Close_oil', 'Volume_oil',\n",
    "       'year', 'month', 'day', 'hour']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10, 15)\n",
      "(10,)\n",
      "% of Class0 : 0.500000 y_test\n",
      "% of Class1 : 0.500000 y_test\n"
     ]
    }
   ],
   "source": [
    "print(X_test.shape)\n",
    "print(y_test.shape)\n",
    "\n",
    "print('%% of Class0 : %f y_test' % (np.count_nonzero(y_test == 0)/float(len(y_test))))\n",
    "print('%% of Class1 : %f y_test' % (np.count_nonzero(y_test == 1)/float(len(y_test))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(6040, 15)\n",
      "(6040,)\n",
      "% of Class0 : 0.513742 y_train\n",
      "% of Class1 : 0.486258 y_train\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(y_train.shape)\n",
    "\n",
    "print('%% of Class0 : %f y_train' % (np.count_nonzero(y_train == 0)/float(len(y_train))))\n",
    "print('%% of Class1 : %f y_train' % (np.count_nonzero(y_train == 1)/float(len(y_train))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      Iter       Train Loss   Remaining Time \n",
      "         1           1.3840            1.65s\n",
      "         2           1.3827            1.14s\n",
      "         3           1.3815            1.14s\n",
      "         4           1.3805            0.96s\n",
      "         5           1.3796            0.89s\n",
      "         6           1.3786            0.82s\n",
      "         7           1.3777            0.75s\n",
      "         8           1.3769            0.72s\n",
      "         9           1.3756            0.70s\n",
      "        10           1.3749            0.66s\n",
      "        11           1.3734            0.65s\n",
      "        12           1.3725            0.62s\n",
      "        13           1.3713            0.60s\n",
      "        14           1.3704            0.60s\n",
      "        15           1.3695            0.58s\n",
      "        16           1.3687            0.56s\n",
      "        17           1.3674            0.55s\n",
      "        18           1.3664            0.54s\n",
      "        19           1.3655            0.53s\n",
      "        20           1.3651            0.52s\n",
      "        21           1.3648            0.51s\n",
      "        22           1.3643            0.50s\n",
      "        23           1.3634            0.49s\n",
      "        24           1.3626            0.49s\n",
      "        25           1.3623            0.49s\n",
      "        26           1.3617            0.48s\n",
      "        27           1.3613            0.47s\n",
      "        28           1.3607            0.48s\n",
      "        29           1.3596            0.49s\n",
      "        30           1.3590            0.48s\n",
      "        31           1.3578            0.48s\n",
      "        32           1.3565            0.49s\n",
      "        33           1.3557            0.49s\n",
      "        34           1.3554            0.49s\n",
      "        35           1.3547            0.49s\n",
      "        36           1.3537            0.48s\n",
      "        37           1.3525            0.48s\n",
      "        38           1.3517            0.47s\n",
      "        39           1.3510            0.46s\n",
      "        40           1.3504            0.45s\n",
      "        41           1.3493            0.44s\n",
      "        42           1.3482            0.43s\n",
      "        43           1.3478            0.42s\n",
      "        44           1.3465            0.41s\n",
      "        45           1.3460            0.40s\n",
      "        46           1.3449            0.39s\n",
      "        47           1.3437            0.38s\n",
      "        48           1.3435            0.37s\n",
      "        49           1.3428            0.36s\n",
      "        50           1.3419            0.36s\n",
      "        51           1.3410            0.35s\n",
      "        52           1.3401            0.34s\n",
      "        53           1.3396            0.34s\n",
      "        54           1.3392            0.33s\n",
      "        55           1.3387            0.32s\n",
      "        56           1.3377            0.32s\n",
      "        57           1.3369            0.31s\n",
      "        58           1.3367            0.30s\n",
      "        59           1.3357            0.29s\n",
      "        60           1.3343            0.29s\n",
      "        61           1.3330            0.28s\n",
      "        62           1.3321            0.27s\n",
      "        63           1.3311            0.26s\n",
      "        64           1.3303            0.26s\n",
      "        65           1.3299            0.25s\n",
      "        66           1.3290            0.24s\n",
      "        67           1.3283            0.23s\n",
      "        68           1.3279            0.22s\n",
      "        69           1.3277            0.22s\n",
      "        70           1.3269            0.21s\n",
      "        71           1.3258            0.20s\n",
      "        72           1.3249            0.19s\n",
      "        73           1.3241            0.19s\n",
      "        74           1.3236            0.18s\n",
      "        75           1.3233            0.17s\n",
      "        76           1.3221            0.16s\n",
      "        77           1.3213            0.16s\n",
      "        78           1.3201            0.15s\n",
      "        79           1.3194            0.14s\n",
      "        80           1.3186            0.14s\n",
      "        81           1.3175            0.13s\n",
      "        82           1.3167            0.12s\n",
      "        83           1.3163            0.12s\n",
      "        84           1.3153            0.11s\n",
      "        85           1.3143            0.10s\n",
      "        86           1.3134            0.10s\n",
      "        87           1.3124            0.09s\n",
      "        88           1.3115            0.08s\n",
      "        89           1.3106            0.08s\n",
      "        90           1.3102            0.07s\n",
      "        91           1.3094            0.06s\n",
      "        92           1.3090            0.06s\n",
      "        93           1.3080            0.05s\n",
      "        94           1.3072            0.04s\n",
      "        95           1.3062            0.04s\n",
      "        96           1.3054            0.03s\n",
      "        97           1.3046            0.02s\n",
      "        98           1.3041            0.01s\n",
      "        99           1.3033            0.01s\n",
      "       100           1.3025            0.00s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
       "              learning_rate=0.1, loss='deviance', max_depth=3,\n",
       "              max_features=None, max_leaf_nodes=None,\n",
       "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "              presort='auto', random_state=None, subsample=1.0, verbose=3,\n",
       "              warm_start=False)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Define Model and fit\n",
    "\n",
    "\n",
    "clf = ensemble.GradientBoostingClassifier(verbose=3)\n",
    "\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## How accurate is this model?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing Accuracy: 0.630298\n"
     ]
    }
   ],
   "source": [
    "accuracy = clf.score(X_train, y_train)\n",
    "print('Testing Accuracy: %f' % accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Predict test data\n",
    "\n",
    "pred = clf.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>return</th>\n",
       "      <th>Open_oil</th>\n",
       "      <th>High_oil</th>\n",
       "      <th>Low_oil</th>\n",
       "      <th>Close_oil</th>\n",
       "      <th>Volume_oil</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th>label</th>\n",
       "      <th>pred</th>\n",
       "      <th>won</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-05-31 03:00:00</th>\n",
       "      <td>1.28779</td>\n",
       "      <td>1.28788</td>\n",
       "      <td>1.28669</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>2.857670e+09</td>\n",
       "      <td>0.019286</td>\n",
       "      <td>1.28779</td>\n",
       "      <td>1.28788</td>\n",
       "      <td>1.28669</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>2.857670e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 04:00:00</th>\n",
       "      <td>1.28719</td>\n",
       "      <td>1.28725</td>\n",
       "      <td>1.28647</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.907510e+09</td>\n",
       "      <td>0.013500</td>\n",
       "      <td>1.28719</td>\n",
       "      <td>1.28725</td>\n",
       "      <td>1.28647</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.907510e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 05:00:00</th>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.28707</td>\n",
       "      <td>1.28598</td>\n",
       "      <td>1.28673</td>\n",
       "      <td>3.887940e+09</td>\n",
       "      <td>0.001286</td>\n",
       "      <td>1.28677</td>\n",
       "      <td>1.28707</td>\n",
       "      <td>1.28598</td>\n",
       "      <td>1.28673</td>\n",
       "      <td>3.887940e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 06:00:00</th>\n",
       "      <td>1.28678</td>\n",
       "      <td>1.28758</td>\n",
       "      <td>1.28654</td>\n",
       "      <td>1.28734</td>\n",
       "      <td>6.998760e+09</td>\n",
       "      <td>-0.018001</td>\n",
       "      <td>1.28678</td>\n",
       "      <td>1.28758</td>\n",
       "      <td>1.28654</td>\n",
       "      <td>1.28734</td>\n",
       "      <td>6.998760e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 07:00:00</th>\n",
       "      <td>1.28735</td>\n",
       "      <td>1.28747</td>\n",
       "      <td>1.28452</td>\n",
       "      <td>1.28482</td>\n",
       "      <td>7.630360e+09</td>\n",
       "      <td>0.081324</td>\n",
       "      <td>1.28735</td>\n",
       "      <td>1.28747</td>\n",
       "      <td>1.28452</td>\n",
       "      <td>1.28482</td>\n",
       "      <td>7.630360e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 08:00:00</th>\n",
       "      <td>1.28484</td>\n",
       "      <td>1.28538</td>\n",
       "      <td>1.28356</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>5.385360e+09</td>\n",
       "      <td>0.027001</td>\n",
       "      <td>1.28484</td>\n",
       "      <td>1.28538</td>\n",
       "      <td>1.28356</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>5.385360e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 09:00:00</th>\n",
       "      <td>1.28400</td>\n",
       "      <td>1.28539</td>\n",
       "      <td>1.28394</td>\n",
       "      <td>1.28470</td>\n",
       "      <td>5.552970e+09</td>\n",
       "      <td>-0.022501</td>\n",
       "      <td>1.28400</td>\n",
       "      <td>1.28539</td>\n",
       "      <td>1.28394</td>\n",
       "      <td>1.28470</td>\n",
       "      <td>5.552970e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 10:00:00</th>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28583</td>\n",
       "      <td>1.28180</td>\n",
       "      <td>1.28259</td>\n",
       "      <td>5.881380e+09</td>\n",
       "      <td>0.068467</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28583</td>\n",
       "      <td>1.28180</td>\n",
       "      <td>1.28259</td>\n",
       "      <td>5.881380e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 11:00:00</th>\n",
       "      <td>1.28264</td>\n",
       "      <td>1.28514</td>\n",
       "      <td>1.28251</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>6.608530e+09</td>\n",
       "      <td>-0.067502</td>\n",
       "      <td>1.28264</td>\n",
       "      <td>1.28514</td>\n",
       "      <td>1.28251</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>6.608530e+09</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-05-31 12:00:00</th>\n",
       "      <td>1.28474</td>\n",
       "      <td>1.28987</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28985</td>\n",
       "      <td>1.079586e+10</td>\n",
       "      <td>-0.164256</td>\n",
       "      <td>1.28474</td>\n",
       "      <td>1.28987</td>\n",
       "      <td>1.28472</td>\n",
       "      <td>1.28985</td>\n",
       "      <td>1.079586e+10</td>\n",
       "      <td>2018</td>\n",
       "      <td>5</td>\n",
       "      <td>31</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Open     High      Low    Close        Volume  \\\n",
       "2018-05-31 03:00:00  1.28779  1.28788  1.28669  1.28719  2.857670e+09   \n",
       "2018-05-31 04:00:00  1.28719  1.28725  1.28647  1.28677  1.907510e+09   \n",
       "2018-05-31 05:00:00  1.28677  1.28707  1.28598  1.28673  3.887940e+09   \n",
       "2018-05-31 06:00:00  1.28678  1.28758  1.28654  1.28734  6.998760e+09   \n",
       "2018-05-31 07:00:00  1.28735  1.28747  1.28452  1.28482  7.630360e+09   \n",
       "2018-05-31 08:00:00  1.28484  1.28538  1.28356  1.28400  5.385360e+09   \n",
       "2018-05-31 09:00:00  1.28400  1.28539  1.28394  1.28470  5.552970e+09   \n",
       "2018-05-31 10:00:00  1.28472  1.28583  1.28180  1.28259  5.881380e+09   \n",
       "2018-05-31 11:00:00  1.28264  1.28514  1.28251  1.28474  6.608530e+09   \n",
       "2018-05-31 12:00:00  1.28474  1.28987  1.28472  1.28985  1.079586e+10   \n",
       "\n",
       "                       return  Open_oil  High_oil  Low_oil  Close_oil  \\\n",
       "2018-05-31 03:00:00  0.019286   1.28779   1.28788  1.28669    1.28719   \n",
       "2018-05-31 04:00:00  0.013500   1.28719   1.28725  1.28647    1.28677   \n",
       "2018-05-31 05:00:00  0.001286   1.28677   1.28707  1.28598    1.28673   \n",
       "2018-05-31 06:00:00 -0.018001   1.28678   1.28758  1.28654    1.28734   \n",
       "2018-05-31 07:00:00  0.081324   1.28735   1.28747  1.28452    1.28482   \n",
       "2018-05-31 08:00:00  0.027001   1.28484   1.28538  1.28356    1.28400   \n",
       "2018-05-31 09:00:00 -0.022501   1.28400   1.28539  1.28394    1.28470   \n",
       "2018-05-31 10:00:00  0.068467   1.28472   1.28583  1.28180    1.28259   \n",
       "2018-05-31 11:00:00 -0.067502   1.28264   1.28514  1.28251    1.28474   \n",
       "2018-05-31 12:00:00 -0.164256   1.28474   1.28987  1.28472    1.28985   \n",
       "\n",
       "                       Volume_oil  year  month  day  hour  label  pred    won  \n",
       "2018-05-31 03:00:00  2.857670e+09  2018      5   31     3      1     0  False  \n",
       "2018-05-31 04:00:00  1.907510e+09  2018      5   31     4      1     0  False  \n",
       "2018-05-31 05:00:00  3.887940e+09  2018      5   31     5      0     0   True  \n",
       "2018-05-31 06:00:00  6.998760e+09  2018      5   31     6      1     1   True  \n",
       "2018-05-31 07:00:00  7.630360e+09  2018      5   31     7      1     1   True  \n",
       "2018-05-31 08:00:00  5.385360e+09  2018      5   31     8      0     0   True  \n",
       "2018-05-31 09:00:00  5.552970e+09  2018      5   31     9      1     0  False  \n",
       "2018-05-31 10:00:00  5.881380e+09  2018      5   31    10      0     0   True  \n",
       "2018-05-31 11:00:00  6.608530e+09  2018      5   31    11      0     0   True  \n",
       "2018-05-31 12:00:00  1.079586e+10  2018      5   31    12      0     1  False  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Comparing predicted and label\n",
    "\n",
    "df_trade = pd.DataFrame(X_test)\n",
    "df_trade['label']  = y_test\n",
    "df_trade['pred']   = pred\n",
    "df_trade['won']    = df_trade['label'] == df_trade['pred']\n",
    "\n",
    "df_trade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n",
      "6\n"
     ]
    }
   ],
   "source": [
    "# Didnt won and won\n",
    "\n",
    "print(df_trade[df_trade['won'] == False]['won'].count())\n",
    "print(df_trade[df_trade['won'] == True]['won'].count())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# % of won\n",
    "\n",
    "df_trade[df_trade['won'] == True]['won'].count()/df_trade['won'].count()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}