notebooks/07_First_Forex_Prediction_15Min.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Making the first prediction in the Forex Market \n",
    "\n",
    "\n",
    "![deep_network_model](../images/deep_network_model_avatar.jpg)\n",
    "\n",
    "\n",
    "The progress of the model can be saved during and after training. This means that a model can be resumed where it left off and avoid long training times. Saving also means that you can share your model and others can recreate your work.\n",
    "\n",
    "We will illustrate how to create a multiple fully connected hidden layer NN, save it and make predictions with trained model after reload it.\n",
    "\n",
    "We will use the EURUSD data for this exercise.\n",
    "\n",
    "We will build a four-hidden layer neural network  to predict the next close price, from the other four features of the precedent period (open, high, low and close).\n",
    "\n",
    "This is a practical exercise to learn how to make predictions with TensorFlow, but it is a naive approach to the real forecasting problem. Don't worry we will be climbing toward better approaches. Also, in the meantime, you will be able to elaborate on your own systems."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load configuration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below, it  is an example of .env file for your customization:\n",
    "~~~~\n",
    "PROJ_DIR=.\n",
    "DATA_DIR=../data/\n",
    "RAW_DIR=../data/raw/\n",
    "INTERIM_DIR=../data/interim/\n",
    "PROCESSED_DIR=../data/processed/\n",
    "FIGURES_DIR=../reports/figures/\n",
    "MODEL_DIR=../models/\n",
    "EXTERNAL_DIR=../data/external/\n",
    "PRODUCTION_DIR=/home/PRODUCTION/\n",
    "~~~~"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/parrondo/anaconda3/envs/deeptrading/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
      "  return f(*args, **kwds)\n",
      "/home/parrondo/anaconda3/envs/deeptrading/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
      "  return f(*args, **kwds)\n",
      "/home/parrondo/anaconda3/envs/deeptrading/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
      "  return f(*args, **kwds)\n",
      "/home/parrondo/anaconda3/envs/deeptrading/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
      "  return f(*args, **kwds)\n",
      "/home/parrondo/anaconda3/envs/deeptrading/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
      "  return f(*args, **kwds)\n",
      "/home/parrondo/anaconda3/envs/deeptrading/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "## 1. Import libraries and modules. Load env variables\n",
    "import os\n",
    "import os\n",
    "import sys\n",
    "module_path = os.path.abspath(os.path.join('..'))\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "#import seaborn as sns\n",
    "import zipfile\n",
    "import sqlite3\n",
    "from datetime import date, datetime, timezone\n",
    "from dotenv import find_dotenv, load_dotenv\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow.python.framework import ops\n",
    "from tensorflow.python.saved_model import builder as saved_model_builder\n",
    "from tensorflow.python.saved_model import signature_constants\n",
    "from tensorflow.python.saved_model import signature_def_utils\n",
    "from tensorflow.python.saved_model import tag_constants\n",
    "from tensorflow.python.saved_model.utils import build_tensor_info\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "#from IPython.display import Image\n",
    "\n",
    "#Reads the key,value pair from .env and adds them to environment variable \n",
    "load_dotenv(find_dotenv())\n",
    "\n",
    "# Check the env variables exist. Check as many variables as you need\n",
    "raw_msg = \"Set your raw data absolute path in the .env file at project root\"\n",
    "assert \"RAW_DIR\" in os.environ, raw_msg\n",
    "data_msg = \"Set your processed data absolute path in the .env file at project root\"\n",
    "assert \"DATA_DIR\" in os.environ, data_msg\n",
    "interim_msg = \"Set your interim data absolute path in the .env file at project root\"\n",
    "assert \"INTERIM_DIR\" in os.environ, interim_msg\n",
    "\n",
    "# Load env variables\n",
    "proj_dir = os.path.expanduser(os.environ.get(\"PROJ_DIR\"))\n",
    "data_dir = os.path.expanduser(os.environ.get(\"DATA_DIR\"))\n",
    "raw_dir = os.path.expanduser(os.environ.get(\"RAW_DIR\"))\n",
    "interim_dir = os.path.expanduser(os.environ.get(\"INTERIM_DIR\"))\n",
    "processed_dir = os.path.expanduser(os.environ.get(\"PROCESSED_DIR\"))\n",
    "figures_dir = os.path.expanduser(os.environ.get(\"FIGURES_DIR\"))\n",
    "model_dir = os.path.expanduser(os.environ.get(\"MODEL_DIR\"))\n",
    "external_dir = os.path.expanduser(os.environ.get(\"EXTERNAL_DIR\"))\n",
    "production_dir = os.path.expanduser(os.environ.get(\"PRODUCTION_DIR\"))\n",
    "\n",
    "# Import our project modules\n",
    "#\n",
    "#Add src/app to the PATH\n",
    "#Ram sys.path.append(os.path.join(proj_dir,\"src/app\"))\n",
    "\n",
    "#Add src/data to the PATH\n",
    "sys.path.append(os.path.join(proj_dir,\"src/data\"))\n",
    "#Ram import make_dataset as md\n",
    "\n",
    "#Add src/visualization to the PATH\n",
    "#Ram sys.path.append(os.path.join(proj_dir,\"src/visualization\"))\n",
    "#Ram import visualize as vs\n",
    "\n",
    "#Data files\n",
    "#raw_data = \n",
    "#interim_data = \n",
    "\n",
    "#Global configuration variables\n",
    "# Send models to production env. folder (True: send. False: Do not send)\n",
    "to_production = True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ingest raw data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Download raw datasets\n",
    "\n",
    "\n",
    "The great people at [HistData.com](http://www.histdata.com/) have setup the infrastructure necessary to provide you FOREX data for free. This is awesome and if possible, you should donate or purchase some of their services to help them. There exist several tools contained in the internet to download the data, but all of them need your carefully atention. For example:\n",
    "* https://github.com/philipperemy/FX-1-Minute-Data\n",
    "* https://github.com/xeb/forex-histdata-etl\n",
    "* https://gist.github.com/EvianZhow/93b30edb5e1ac44f3dd2de7ef9a543d9\n",
    "\n",
    "I include forex time series zipped files needed in this tutorial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "## 2. Download raw data sets\n",
    "\n",
    "#This point must be adapted for each project\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Download .csv data file from HistData.com and save in ../data/raw dir"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Basic pre-process data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Machine Learning time series algorithms usuallly require data to be into a single text file in tabular format, with each row representing a timestamp of the input dataset and each column one of its features.\n",
    "\n",
    "\"Prepare” data for Machine Learning is a complex task depending on where the data is stored and where it is obtained from. And doubtless, it is one of the most time consumption task. Often the Forex data is not available in a single file. They may be distributed across different sources like multiple compressed CSV files, spreadsheets or plain text files, normalized in database tables, or even in NoSql database like MongoDB. So we need a tool to stage, filter, transform when necessary, and finally export to a single flat, text CSV file.\n",
    "\n",
    "If your Forex data is small and the changes are simple such as adding a derived field or new events you can use a spreadsheet, make the necessary changes, and then export it to a CSV file. Certainly, not too professional. But when the changes are more complex; e.g., joining several sources, filtering a subset of the data, or managing a large amount of timestamp rows, you might need a more powerful tool like an RDBMS. MySQL is a great one and it’s free, and opensourced. In this tutorial, we have selected SQLite which is enough for our purpose and data size. Here we treat several compressed .csv files distributed in different folders, which is very usual in real trading. If the data size that we are managing is in the terabytes, then we should consider Hadoop. But trustus, that is another story.\n",
    "\n",
    "![deep_network_model](../images/sqlite_transformations.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# All available instruments to trade with\n",
    "symbol = \"EURUSD\" # type=str, symbol list using format \"EURUSD\" \"EURGBP\"\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Clean database table\n",
    "DATABASE_FILE = processed_dir+\"Data.db\"\n",
    "def initialize_db(self):\n",
    "    with sqlite3.connect(DATABASE_FILE) as connection:\n",
    "        cursor = connection.cursor()\n",
    "        cursor.execute('CREATE TABLE IF NOT EXISTS History (timestamp INTEGER,'\n",
    "                       'symbol VARCHAR(20), high FLOAT, low FLOAT,'\n",
    "                       'open FLOAT, close FLOAT, volume FLOAT, '\n",
    "                       'quoteVolume FLOAT, weightedAverage FLOAT,'\n",
    "                       'PRIMARY KEY (timestamp, symbol));')\n",
    "        connection.commit()\n",
    "\n",
    "initialize_db(DATABASE_FILE)\n",
    "conn = sqlite3.connect(DATABASE_FILE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the dataframe\n",
    "columns = [\"timestamp\", \"symbol\", \"open\", \"high\", \"low\", \"close\", \"volume\", \"quoteVolume\", \"weightedAverage\"]\n",
    "dtype = {\"timestamp\":\"INTEGER\",\n",
    "         \"symbol\":\"VARCHAR(20)\",\n",
    "         \"open\":\"FLOAT\",\n",
    "         \"high\":\"FLOAT\",\n",
    "         \"low\":\"FLOAT\",\n",
    "         \"close\":\"FLOAT\",\n",
    "         \"volume\":\"FLOAT\",\n",
    "         \"quoteVolume\":\"FLOAT\",\n",
    "         \"weightedAverage\":\"FLOAT\"}\n",
    "df0 = pd.DataFrame(columns=columns)\n",
    "\n",
    "# Write dataframe to sqlite database\n",
    "df0.to_sql(\"History\", conn, if_exists=\"replace\", index=False, dtype=dtype)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "inputfile:  DAT_ASCII_EURUSD_M1_2017.csv\n"
     ]
    }
   ],
   "source": [
    "#\n",
    "# Database population\n",
    "#\n",
    "#\n",
    "# All price instrument in cash currency base\n",
    "#\n",
    "\n",
    "\n",
    "# Initialicing dataframes           \n",
    "df1 = pd.DataFrame().iloc[0:0]\n",
    "df2 = pd.DataFrame().iloc[0:0]\n",
    "\n",
    "# Managing diferent raw data files from several folders\n",
    "compressedfile = os.path.join(raw_dir,symbol.lower(),'HISTDATA_COM_ASCII_'+symbol+'_M1_2017.zip')\n",
    "zf = zipfile.ZipFile(compressedfile) # having .csv zipped file\n",
    "inputfile = 'DAT_ASCII_'+symbol+'_M1_2017.csv'\n",
    "print(\"inputfile: \",inputfile)\n",
    "#df1 = pd.read_csv(inputfile, names=['date', 'open', 'high', 'low', 'close', 'volume'],index_col=0, parse_dates=True, delimiter=\";\")\n",
    "\n",
    "df1 = pd.read_csv(zf.open(inputfile), header=None,\n",
    "                  names=['timestamp', 'open', 'high', 'low', 'close', 'volume'],\n",
    "                  index_col=0, parse_dates=True,sep=';') # reads the csv and creates the dataframe called \"df1\"\n",
    "\n",
    "# Resampling data from 1Min to desired Period\n",
    "df2 =  df1[\"open\"].resample('15Min').ohlc()\n",
    "        \n",
    "# Convert pandas timestamps in Unix timestamps:\n",
    "df2.index = df2.index.astype(np.int64) // 10**9\n",
    "\n",
    "# Insert new columns with the instrument name and their values\n",
    "df2.insert(loc=0, column='symbol', value=symbol)\n",
    "\n",
    "#Only for compatibility with stocks code (optional, you may want to remove this fields from database)\n",
    "df2['volume']=1000.\n",
    "df2['quoteVolume']=1000.\n",
    "df2['weightedAverage']=1.\n",
    "\n",
    "# Reset index to save in database\n",
    "df2=df2.reset_index()\n",
    " \n",
    "#Filling gaps forward\n",
    "df2 = df2.fillna(method='pad')\n",
    "        \n",
    "# Save to database (Populate database)\n",
    "df2.to_sql(\"History\", conn, if_exists=\"append\", index=False, chunksize=1000)\n",
    "\n",
    "# Liberate memory\n",
    "del df1\n",
    "del df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Committing changes and closing the connection to the database file\n",
    "conn.commit()\n",
    "conn.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data from database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load dataset (In this case reading the database)\n",
    "DATABASE_FILE=processed_dir+\"Data.db\"\n",
    "conn = sqlite3.connect(DATABASE_FILE)\n",
    "df = pd.read_sql_query(\"select * from History;\", conn)\n",
    "df=df.drop([\"index\"], axis=1, errors=\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>timestamp</th>\n",
       "      <th>symbol</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>quoteVolume</th>\n",
       "      <th>weightedAverage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1483322400</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05206</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1483323300</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05172</td>\n",
       "      <td>1.05178</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1483324200</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05185</td>\n",
       "      <td>1.05209</td>\n",
       "      <td>1.05168</td>\n",
       "      <td>1.05169</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1483325100</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05130</td>\n",
       "      <td>1.05150</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1483326000</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05175</td>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05156</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1483326900</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05083</td>\n",
       "      <td>1.05122</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1483327800</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05118</td>\n",
       "      <td>1.05144</td>\n",
       "      <td>1.05107</td>\n",
       "      <td>1.05107</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1483328700</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.04929</td>\n",
       "      <td>1.04929</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1483329600</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04889</td>\n",
       "      <td>1.04904</td>\n",
       "      <td>1.04765</td>\n",
       "      <td>1.04802</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1483330500</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04826</td>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04816</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1483331400</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04890</td>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04874</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1483332300</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04876</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04858</td>\n",
       "      <td>1.04868</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1483333200</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04866</td>\n",
       "      <td>1.04885</td>\n",
       "      <td>1.04852</td>\n",
       "      <td>1.04879</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1483334100</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04803</td>\n",
       "      <td>1.04803</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1483335000</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1483335900</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04822</td>\n",
       "      <td>1.04835</td>\n",
       "      <td>1.04791</td>\n",
       "      <td>1.04803</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1483336800</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04805</td>\n",
       "      <td>1.04812</td>\n",
       "      <td>1.04801</td>\n",
       "      <td>1.04801</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1483337700</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04784</td>\n",
       "      <td>1.04798</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1483338600</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04776</td>\n",
       "      <td>1.04794</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1483339500</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04795</td>\n",
       "      <td>1.04798</td>\n",
       "      <td>1.04768</td>\n",
       "      <td>1.04782</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1483340400</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04726</td>\n",
       "      <td>1.04726</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1483341300</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04722</td>\n",
       "      <td>1.04746</td>\n",
       "      <td>1.04712</td>\n",
       "      <td>1.04714</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1483342200</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04724</td>\n",
       "      <td>1.04749</td>\n",
       "      <td>1.04699</td>\n",
       "      <td>1.04711</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1483343100</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04717</td>\n",
       "      <td>1.04720</td>\n",
       "      <td>1.04653</td>\n",
       "      <td>1.04659</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1483344000</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04615</td>\n",
       "      <td>1.04649</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1483344900</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04680</td>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04657</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1483345800</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04654</td>\n",
       "      <td>1.04664</td>\n",
       "      <td>1.04644</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1483346700</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04658</td>\n",
       "      <td>1.04673</td>\n",
       "      <td>1.04657</td>\n",
       "      <td>1.04668</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1483347600</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04679</td>\n",
       "      <td>1.04649</td>\n",
       "      <td>1.04677</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1483348500</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.04676</td>\n",
       "      <td>1.04695</td>\n",
       "      <td>1.04661</td>\n",
       "      <td>1.04695</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34686</th>\n",
       "      <td>1514539800</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19911</td>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19907</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34687</th>\n",
       "      <td>1514540700</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19946</td>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19926</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34688</th>\n",
       "      <td>1514541600</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19962</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19952</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34689</th>\n",
       "      <td>1514542500</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19951</td>\n",
       "      <td>1.19955</td>\n",
       "      <td>1.19880</td>\n",
       "      <td>1.19911</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34690</th>\n",
       "      <td>1514543400</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34691</th>\n",
       "      <td>1514544300</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19977</td>\n",
       "      <td>1.20069</td>\n",
       "      <td>1.19973</td>\n",
       "      <td>1.20069</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34692</th>\n",
       "      <td>1514545200</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20023</td>\n",
       "      <td>1.20082</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34693</th>\n",
       "      <td>1514546100</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20100</td>\n",
       "      <td>1.20128</td>\n",
       "      <td>1.20058</td>\n",
       "      <td>1.20070</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34694</th>\n",
       "      <td>1514547000</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34695</th>\n",
       "      <td>1514547900</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34696</th>\n",
       "      <td>1514548800</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20220</td>\n",
       "      <td>1.20247</td>\n",
       "      <td>1.20193</td>\n",
       "      <td>1.20208</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34697</th>\n",
       "      <td>1514549700</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20197</td>\n",
       "      <td>1.20238</td>\n",
       "      <td>1.20195</td>\n",
       "      <td>1.20229</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34698</th>\n",
       "      <td>1514550600</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20243</td>\n",
       "      <td>1.20255</td>\n",
       "      <td>1.20201</td>\n",
       "      <td>1.20210</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34699</th>\n",
       "      <td>1514551500</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20222</td>\n",
       "      <td>1.20236</td>\n",
       "      <td>1.20196</td>\n",
       "      <td>1.20197</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34700</th>\n",
       "      <td>1514552400</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20214</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20200</td>\n",
       "      <td>1.20225</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34701</th>\n",
       "      <td>1514553300</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20225</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20189</td>\n",
       "      <td>1.20189</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34702</th>\n",
       "      <td>1514554200</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20150</td>\n",
       "      <td>1.20155</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34703</th>\n",
       "      <td>1514555100</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20164</td>\n",
       "      <td>1.20167</td>\n",
       "      <td>1.20124</td>\n",
       "      <td>1.20133</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34704</th>\n",
       "      <td>1514556000</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20134</td>\n",
       "      <td>1.20138</td>\n",
       "      <td>1.20094</td>\n",
       "      <td>1.20094</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34705</th>\n",
       "      <td>1514556900</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20103</td>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20100</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34706</th>\n",
       "      <td>1514557800</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20098</td>\n",
       "      <td>1.20099</td>\n",
       "      <td>1.20078</td>\n",
       "      <td>1.20081</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34707</th>\n",
       "      <td>1514558700</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20082</td>\n",
       "      <td>1.20114</td>\n",
       "      <td>1.20071</td>\n",
       "      <td>1.20106</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34708</th>\n",
       "      <td>1514559600</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20092</td>\n",
       "      <td>1.20104</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20081</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34709</th>\n",
       "      <td>1514560500</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20049</td>\n",
       "      <td>1.20053</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34710</th>\n",
       "      <td>1514561400</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20051</td>\n",
       "      <td>1.20052</td>\n",
       "      <td>1.20019</td>\n",
       "      <td>1.20024</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34711</th>\n",
       "      <td>1514562300</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20025</td>\n",
       "      <td>1.20028</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.19983</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34712</th>\n",
       "      <td>1514563200</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20035</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20024</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34713</th>\n",
       "      <td>1514564100</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.19958</td>\n",
       "      <td>1.19963</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34714</th>\n",
       "      <td>1514565000</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19988</td>\n",
       "      <td>1.19963</td>\n",
       "      <td>1.19966</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34715</th>\n",
       "      <td>1514565900</td>\n",
       "      <td>EURUSD</td>\n",
       "      <td>1.19966</td>\n",
       "      <td>1.20009</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19982</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>34716 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        timestamp  symbol     open     high      low    close  volume  \\\n",
       "0      1483322400  EURUSD  1.05155  1.05213  1.05155  1.05206  1000.0   \n",
       "1      1483323300  EURUSD  1.05213  1.05213  1.05172  1.05178  1000.0   \n",
       "2      1483324200  EURUSD  1.05185  1.05209  1.05168  1.05169  1000.0   \n",
       "3      1483325100  EURUSD  1.05188  1.05188  1.05130  1.05150  1000.0   \n",
       "4      1483326000  EURUSD  1.05152  1.05175  1.05149  1.05156  1000.0   \n",
       "5      1483326900  EURUSD  1.05149  1.05152  1.05083  1.05122  1000.0   \n",
       "6      1483327800  EURUSD  1.05118  1.05144  1.05107  1.05107  1000.0   \n",
       "7      1483328700  EURUSD  1.05101  1.05101  1.04929  1.04929  1000.0   \n",
       "8      1483329600  EURUSD  1.04889  1.04904  1.04765  1.04802  1000.0   \n",
       "9      1483330500  EURUSD  1.04793  1.04826  1.04793  1.04816  1000.0   \n",
       "10     1483331400  EURUSD  1.04818  1.04890  1.04818  1.04874  1000.0   \n",
       "11     1483332300  EURUSD  1.04876  1.04884  1.04858  1.04868  1000.0   \n",
       "12     1483333200  EURUSD  1.04866  1.04885  1.04852  1.04879  1000.0   \n",
       "13     1483334100  EURUSD  1.04884  1.04884  1.04803  1.04803  1000.0   \n",
       "14     1483335000  EURUSD  1.04802  1.04831  1.04802  1.04831  1000.0   \n",
       "15     1483335900  EURUSD  1.04822  1.04835  1.04791  1.04803  1000.0   \n",
       "16     1483336800  EURUSD  1.04805  1.04812  1.04801  1.04801  1000.0   \n",
       "17     1483337700  EURUSD  1.04806  1.04806  1.04784  1.04798  1000.0   \n",
       "18     1483338600  EURUSD  1.04800  1.04800  1.04776  1.04794  1000.0   \n",
       "19     1483339500  EURUSD  1.04795  1.04798  1.04768  1.04782  1000.0   \n",
       "20     1483340400  EURUSD  1.04782  1.04782  1.04726  1.04726  1000.0   \n",
       "21     1483341300  EURUSD  1.04722  1.04746  1.04712  1.04714  1000.0   \n",
       "22     1483342200  EURUSD  1.04724  1.04749  1.04699  1.04711  1000.0   \n",
       "23     1483343100  EURUSD  1.04717  1.04720  1.04653  1.04659  1000.0   \n",
       "24     1483344000  EURUSD  1.04655  1.04655  1.04615  1.04649  1000.0   \n",
       "25     1483344900  EURUSD  1.04645  1.04680  1.04645  1.04657  1000.0   \n",
       "26     1483345800  EURUSD  1.04654  1.04664  1.04644  1.04655  1000.0   \n",
       "27     1483346700  EURUSD  1.04658  1.04673  1.04657  1.04668  1000.0   \n",
       "28     1483347600  EURUSD  1.04655  1.04679  1.04649  1.04677  1000.0   \n",
       "29     1483348500  EURUSD  1.04676  1.04695  1.04661  1.04695  1000.0   \n",
       "...           ...     ...      ...      ...      ...      ...     ...   \n",
       "34686  1514539800  EURUSD  1.19882  1.19911  1.19882  1.19907  1000.0   \n",
       "34687  1514540700  EURUSD  1.19894  1.19946  1.19894  1.19926  1000.0   \n",
       "34688  1514541600  EURUSD  1.19927  1.19962  1.19927  1.19952  1000.0   \n",
       "34689  1514542500  EURUSD  1.19951  1.19955  1.19880  1.19911  1000.0   \n",
       "34690  1514543400  EURUSD  1.19905  1.19981  1.19905  1.19981  1000.0   \n",
       "34691  1514544300  EURUSD  1.19977  1.20069  1.19973  1.20069  1000.0   \n",
       "34692  1514545200  EURUSD  1.20097  1.20097  1.20023  1.20082  1000.0   \n",
       "34693  1514546100  EURUSD  1.20100  1.20128  1.20058  1.20070  1000.0   \n",
       "34694  1514547000  EURUSD  1.20072  1.20141  1.20072  1.20141  1000.0   \n",
       "34695  1514547900  EURUSD  1.20142  1.20215  1.20142  1.20215  1000.0   \n",
       "34696  1514548800  EURUSD  1.20220  1.20247  1.20193  1.20208  1000.0   \n",
       "34697  1514549700  EURUSD  1.20197  1.20238  1.20195  1.20229  1000.0   \n",
       "34698  1514550600  EURUSD  1.20243  1.20255  1.20201  1.20210  1000.0   \n",
       "34699  1514551500  EURUSD  1.20222  1.20236  1.20196  1.20197  1000.0   \n",
       "34700  1514552400  EURUSD  1.20214  1.20231  1.20200  1.20225  1000.0   \n",
       "34701  1514553300  EURUSD  1.20225  1.20231  1.20189  1.20189  1000.0   \n",
       "34702  1514554200  EURUSD  1.20181  1.20181  1.20150  1.20155  1000.0   \n",
       "34703  1514555100  EURUSD  1.20164  1.20167  1.20124  1.20133  1000.0   \n",
       "34704  1514556000  EURUSD  1.20134  1.20138  1.20094  1.20094  1000.0   \n",
       "34705  1514556900  EURUSD  1.20087  1.20103  1.20087  1.20100  1000.0   \n",
       "34706  1514557800  EURUSD  1.20098  1.20099  1.20078  1.20081  1000.0   \n",
       "34707  1514558700  EURUSD  1.20082  1.20114  1.20071  1.20106  1000.0   \n",
       "34708  1514559600  EURUSD  1.20092  1.20104  1.20072  1.20081  1000.0   \n",
       "34709  1514560500  EURUSD  1.20080  1.20080  1.20049  1.20053  1000.0   \n",
       "34710  1514561400  EURUSD  1.20051  1.20052  1.20019  1.20024  1000.0   \n",
       "34711  1514562300  EURUSD  1.20025  1.20028  1.19978  1.19983  1000.0   \n",
       "34712  1514563200  EURUSD  1.19978  1.20035  1.19978  1.20024  1000.0   \n",
       "34713  1514564100  EURUSD  1.20033  1.20033  1.19958  1.19963  1000.0   \n",
       "34714  1514565000  EURUSD  1.19981  1.19988  1.19963  1.19966  1000.0   \n",
       "34715  1514565900  EURUSD  1.19966  1.20009  1.19927  1.19982  1000.0   \n",
       "\n",
       "       quoteVolume  weightedAverage  \n",
       "0           1000.0              1.0  \n",
       "1           1000.0              1.0  \n",
       "2           1000.0              1.0  \n",
       "3           1000.0              1.0  \n",
       "4           1000.0              1.0  \n",
       "5           1000.0              1.0  \n",
       "6           1000.0              1.0  \n",
       "7           1000.0              1.0  \n",
       "8           1000.0              1.0  \n",
       "9           1000.0              1.0  \n",
       "10          1000.0              1.0  \n",
       "11          1000.0              1.0  \n",
       "12          1000.0              1.0  \n",
       "13          1000.0              1.0  \n",
       "14          1000.0              1.0  \n",
       "15          1000.0              1.0  \n",
       "16          1000.0              1.0  \n",
       "17          1000.0              1.0  \n",
       "18          1000.0              1.0  \n",
       "19          1000.0              1.0  \n",
       "20          1000.0              1.0  \n",
       "21          1000.0              1.0  \n",
       "22          1000.0              1.0  \n",
       "23          1000.0              1.0  \n",
       "24          1000.0              1.0  \n",
       "25          1000.0              1.0  \n",
       "26          1000.0              1.0  \n",
       "27          1000.0              1.0  \n",
       "28          1000.0              1.0  \n",
       "29          1000.0              1.0  \n",
       "...            ...              ...  \n",
       "34686       1000.0              1.0  \n",
       "34687       1000.0              1.0  \n",
       "34688       1000.0              1.0  \n",
       "34689       1000.0              1.0  \n",
       "34690       1000.0              1.0  \n",
       "34691       1000.0              1.0  \n",
       "34692       1000.0              1.0  \n",
       "34693       1000.0              1.0  \n",
       "34694       1000.0              1.0  \n",
       "34695       1000.0              1.0  \n",
       "34696       1000.0              1.0  \n",
       "34697       1000.0              1.0  \n",
       "34698       1000.0              1.0  \n",
       "34699       1000.0              1.0  \n",
       "34700       1000.0              1.0  \n",
       "34701       1000.0              1.0  \n",
       "34702       1000.0              1.0  \n",
       "34703       1000.0              1.0  \n",
       "34704       1000.0              1.0  \n",
       "34705       1000.0              1.0  \n",
       "34706       1000.0              1.0  \n",
       "34707       1000.0              1.0  \n",
       "34708       1000.0              1.0  \n",
       "34709       1000.0              1.0  \n",
       "34710       1000.0              1.0  \n",
       "34711       1000.0              1.0  \n",
       "34712       1000.0              1.0  \n",
       "34713       1000.0              1.0  \n",
       "34714       1000.0              1.0  \n",
       "34715       1000.0              1.0  \n",
       "\n",
       "[34716 rows x 9 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1 = df.loc[(df['symbol'] == \"EURUSD\"),['timestamp','open','high','low','close']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>timestamp</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1483322400</td>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1483323300</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05172</td>\n",
       "      <td>1.05178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1483324200</td>\n",
       "      <td>1.05185</td>\n",
       "      <td>1.05209</td>\n",
       "      <td>1.05168</td>\n",
       "      <td>1.05169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1483325100</td>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05130</td>\n",
       "      <td>1.05150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1483326000</td>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05175</td>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1483326900</td>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05083</td>\n",
       "      <td>1.05122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1483327800</td>\n",
       "      <td>1.05118</td>\n",
       "      <td>1.05144</td>\n",
       "      <td>1.05107</td>\n",
       "      <td>1.05107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1483328700</td>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.04929</td>\n",
       "      <td>1.04929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1483329600</td>\n",
       "      <td>1.04889</td>\n",
       "      <td>1.04904</td>\n",
       "      <td>1.04765</td>\n",
       "      <td>1.04802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1483330500</td>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04826</td>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1483331400</td>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04890</td>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1483332300</td>\n",
       "      <td>1.04876</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04858</td>\n",
       "      <td>1.04868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1483333200</td>\n",
       "      <td>1.04866</td>\n",
       "      <td>1.04885</td>\n",
       "      <td>1.04852</td>\n",
       "      <td>1.04879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1483334100</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04803</td>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1483335000</td>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1483335900</td>\n",
       "      <td>1.04822</td>\n",
       "      <td>1.04835</td>\n",
       "      <td>1.04791</td>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1483336800</td>\n",
       "      <td>1.04805</td>\n",
       "      <td>1.04812</td>\n",
       "      <td>1.04801</td>\n",
       "      <td>1.04801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1483337700</td>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04784</td>\n",
       "      <td>1.04798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1483338600</td>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04776</td>\n",
       "      <td>1.04794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1483339500</td>\n",
       "      <td>1.04795</td>\n",
       "      <td>1.04798</td>\n",
       "      <td>1.04768</td>\n",
       "      <td>1.04782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1483340400</td>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04726</td>\n",
       "      <td>1.04726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1483341300</td>\n",
       "      <td>1.04722</td>\n",
       "      <td>1.04746</td>\n",
       "      <td>1.04712</td>\n",
       "      <td>1.04714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1483342200</td>\n",
       "      <td>1.04724</td>\n",
       "      <td>1.04749</td>\n",
       "      <td>1.04699</td>\n",
       "      <td>1.04711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1483343100</td>\n",
       "      <td>1.04717</td>\n",
       "      <td>1.04720</td>\n",
       "      <td>1.04653</td>\n",
       "      <td>1.04659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1483344000</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04615</td>\n",
       "      <td>1.04649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1483344900</td>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04680</td>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1483345800</td>\n",
       "      <td>1.04654</td>\n",
       "      <td>1.04664</td>\n",
       "      <td>1.04644</td>\n",
       "      <td>1.04655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1483346700</td>\n",
       "      <td>1.04658</td>\n",
       "      <td>1.04673</td>\n",
       "      <td>1.04657</td>\n",
       "      <td>1.04668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1483347600</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04679</td>\n",
       "      <td>1.04649</td>\n",
       "      <td>1.04677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1483348500</td>\n",
       "      <td>1.04676</td>\n",
       "      <td>1.04695</td>\n",
       "      <td>1.04661</td>\n",
       "      <td>1.04695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34686</th>\n",
       "      <td>1514539800</td>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19911</td>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34687</th>\n",
       "      <td>1514540700</td>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19946</td>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34688</th>\n",
       "      <td>1514541600</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19962</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34689</th>\n",
       "      <td>1514542500</td>\n",
       "      <td>1.19951</td>\n",
       "      <td>1.19955</td>\n",
       "      <td>1.19880</td>\n",
       "      <td>1.19911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34690</th>\n",
       "      <td>1514543400</td>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34691</th>\n",
       "      <td>1514544300</td>\n",
       "      <td>1.19977</td>\n",
       "      <td>1.20069</td>\n",
       "      <td>1.19973</td>\n",
       "      <td>1.20069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34692</th>\n",
       "      <td>1514545200</td>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20023</td>\n",
       "      <td>1.20082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34693</th>\n",
       "      <td>1514546100</td>\n",
       "      <td>1.20100</td>\n",
       "      <td>1.20128</td>\n",
       "      <td>1.20058</td>\n",
       "      <td>1.20070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34694</th>\n",
       "      <td>1514547000</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34695</th>\n",
       "      <td>1514547900</td>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34696</th>\n",
       "      <td>1514548800</td>\n",
       "      <td>1.20220</td>\n",
       "      <td>1.20247</td>\n",
       "      <td>1.20193</td>\n",
       "      <td>1.20208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34697</th>\n",
       "      <td>1514549700</td>\n",
       "      <td>1.20197</td>\n",
       "      <td>1.20238</td>\n",
       "      <td>1.20195</td>\n",
       "      <td>1.20229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34698</th>\n",
       "      <td>1514550600</td>\n",
       "      <td>1.20243</td>\n",
       "      <td>1.20255</td>\n",
       "      <td>1.20201</td>\n",
       "      <td>1.20210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34699</th>\n",
       "      <td>1514551500</td>\n",
       "      <td>1.20222</td>\n",
       "      <td>1.20236</td>\n",
       "      <td>1.20196</td>\n",
       "      <td>1.20197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34700</th>\n",
       "      <td>1514552400</td>\n",
       "      <td>1.20214</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20200</td>\n",
       "      <td>1.20225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34701</th>\n",
       "      <td>1514553300</td>\n",
       "      <td>1.20225</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20189</td>\n",
       "      <td>1.20189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34702</th>\n",
       "      <td>1514554200</td>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20150</td>\n",
       "      <td>1.20155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34703</th>\n",
       "      <td>1514555100</td>\n",
       "      <td>1.20164</td>\n",
       "      <td>1.20167</td>\n",
       "      <td>1.20124</td>\n",
       "      <td>1.20133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34704</th>\n",
       "      <td>1514556000</td>\n",
       "      <td>1.20134</td>\n",
       "      <td>1.20138</td>\n",
       "      <td>1.20094</td>\n",
       "      <td>1.20094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34705</th>\n",
       "      <td>1514556900</td>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20103</td>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34706</th>\n",
       "      <td>1514557800</td>\n",
       "      <td>1.20098</td>\n",
       "      <td>1.20099</td>\n",
       "      <td>1.20078</td>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34707</th>\n",
       "      <td>1514558700</td>\n",
       "      <td>1.20082</td>\n",
       "      <td>1.20114</td>\n",
       "      <td>1.20071</td>\n",
       "      <td>1.20106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34708</th>\n",
       "      <td>1514559600</td>\n",
       "      <td>1.20092</td>\n",
       "      <td>1.20104</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34709</th>\n",
       "      <td>1514560500</td>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20049</td>\n",
       "      <td>1.20053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34710</th>\n",
       "      <td>1514561400</td>\n",
       "      <td>1.20051</td>\n",
       "      <td>1.20052</td>\n",
       "      <td>1.20019</td>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34711</th>\n",
       "      <td>1514562300</td>\n",
       "      <td>1.20025</td>\n",
       "      <td>1.20028</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.19983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34712</th>\n",
       "      <td>1514563200</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20035</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34713</th>\n",
       "      <td>1514564100</td>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.19958</td>\n",
       "      <td>1.19963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34714</th>\n",
       "      <td>1514565000</td>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19988</td>\n",
       "      <td>1.19963</td>\n",
       "      <td>1.19966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34715</th>\n",
       "      <td>1514565900</td>\n",
       "      <td>1.19966</td>\n",
       "      <td>1.20009</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19982</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>34716 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        timestamp     open     high      low    close\n",
       "0      1483322400  1.05155  1.05213  1.05155  1.05206\n",
       "1      1483323300  1.05213  1.05213  1.05172  1.05178\n",
       "2      1483324200  1.05185  1.05209  1.05168  1.05169\n",
       "3      1483325100  1.05188  1.05188  1.05130  1.05150\n",
       "4      1483326000  1.05152  1.05175  1.05149  1.05156\n",
       "5      1483326900  1.05149  1.05152  1.05083  1.05122\n",
       "6      1483327800  1.05118  1.05144  1.05107  1.05107\n",
       "7      1483328700  1.05101  1.05101  1.04929  1.04929\n",
       "8      1483329600  1.04889  1.04904  1.04765  1.04802\n",
       "9      1483330500  1.04793  1.04826  1.04793  1.04816\n",
       "10     1483331400  1.04818  1.04890  1.04818  1.04874\n",
       "11     1483332300  1.04876  1.04884  1.04858  1.04868\n",
       "12     1483333200  1.04866  1.04885  1.04852  1.04879\n",
       "13     1483334100  1.04884  1.04884  1.04803  1.04803\n",
       "14     1483335000  1.04802  1.04831  1.04802  1.04831\n",
       "15     1483335900  1.04822  1.04835  1.04791  1.04803\n",
       "16     1483336800  1.04805  1.04812  1.04801  1.04801\n",
       "17     1483337700  1.04806  1.04806  1.04784  1.04798\n",
       "18     1483338600  1.04800  1.04800  1.04776  1.04794\n",
       "19     1483339500  1.04795  1.04798  1.04768  1.04782\n",
       "20     1483340400  1.04782  1.04782  1.04726  1.04726\n",
       "21     1483341300  1.04722  1.04746  1.04712  1.04714\n",
       "22     1483342200  1.04724  1.04749  1.04699  1.04711\n",
       "23     1483343100  1.04717  1.04720  1.04653  1.04659\n",
       "24     1483344000  1.04655  1.04655  1.04615  1.04649\n",
       "25     1483344900  1.04645  1.04680  1.04645  1.04657\n",
       "26     1483345800  1.04654  1.04664  1.04644  1.04655\n",
       "27     1483346700  1.04658  1.04673  1.04657  1.04668\n",
       "28     1483347600  1.04655  1.04679  1.04649  1.04677\n",
       "29     1483348500  1.04676  1.04695  1.04661  1.04695\n",
       "...           ...      ...      ...      ...      ...\n",
       "34686  1514539800  1.19882  1.19911  1.19882  1.19907\n",
       "34687  1514540700  1.19894  1.19946  1.19894  1.19926\n",
       "34688  1514541600  1.19927  1.19962  1.19927  1.19952\n",
       "34689  1514542500  1.19951  1.19955  1.19880  1.19911\n",
       "34690  1514543400  1.19905  1.19981  1.19905  1.19981\n",
       "34691  1514544300  1.19977  1.20069  1.19973  1.20069\n",
       "34692  1514545200  1.20097  1.20097  1.20023  1.20082\n",
       "34693  1514546100  1.20100  1.20128  1.20058  1.20070\n",
       "34694  1514547000  1.20072  1.20141  1.20072  1.20141\n",
       "34695  1514547900  1.20142  1.20215  1.20142  1.20215\n",
       "34696  1514548800  1.20220  1.20247  1.20193  1.20208\n",
       "34697  1514549700  1.20197  1.20238  1.20195  1.20229\n",
       "34698  1514550600  1.20243  1.20255  1.20201  1.20210\n",
       "34699  1514551500  1.20222  1.20236  1.20196  1.20197\n",
       "34700  1514552400  1.20214  1.20231  1.20200  1.20225\n",
       "34701  1514553300  1.20225  1.20231  1.20189  1.20189\n",
       "34702  1514554200  1.20181  1.20181  1.20150  1.20155\n",
       "34703  1514555100  1.20164  1.20167  1.20124  1.20133\n",
       "34704  1514556000  1.20134  1.20138  1.20094  1.20094\n",
       "34705  1514556900  1.20087  1.20103  1.20087  1.20100\n",
       "34706  1514557800  1.20098  1.20099  1.20078  1.20081\n",
       "34707  1514558700  1.20082  1.20114  1.20071  1.20106\n",
       "34708  1514559600  1.20092  1.20104  1.20072  1.20081\n",
       "34709  1514560500  1.20080  1.20080  1.20049  1.20053\n",
       "34710  1514561400  1.20051  1.20052  1.20019  1.20024\n",
       "34711  1514562300  1.20025  1.20028  1.19978  1.19983\n",
       "34712  1514563200  1.19978  1.20035  1.19978  1.20024\n",
       "34713  1514564100  1.20033  1.20033  1.19958  1.19963\n",
       "34714  1514565000  1.19981  1.19988  1.19963  1.19966\n",
       "34715  1514565900  1.19966  1.20009  1.19927  1.19982\n",
       "\n",
       "[34716 rows x 5 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dimensions of dataset\n",
      "n= 34716 p= 4\n"
     ]
    }
   ],
   "source": [
    "# Create the apropiate features dataframe (In this case is open, high, low and close prices of symbol)\n",
    "df2 = pd.DataFrame()\n",
    "symbol_features = ['open', 'high', 'low', 'close']\n",
    "for feature in symbol_features:\n",
    "    df1 = df.loc[(df['symbol'] == symbol),['timestamp', feature]]\n",
    "    df1.columns=['timestamp',symbol+feature]\n",
    "    # Setting the timestamp as the index\n",
    "    df1.set_index('timestamp', inplace=True)\n",
    "    \n",
    "    # Convert timestamps to dates but it's not mandatory\n",
    "    #df1.index = pd.to_datetime(df1.index, unit='s')\n",
    "\n",
    "    # Just perform a join and that's it\n",
    "    df2 = df2.join(df1, how='outer')\n",
    "\n",
    "# Filling the remaining gaps backguards (the initial gaps has not before value)\n",
    "df2 = df2.fillna(method='bfill')\n",
    "\n",
    "# Independent variables data\n",
    "X_raw = df2\n",
    "\n",
    "# Dimensions of dataset\n",
    "print(\"Dimensions of dataset\")\n",
    "n = X_raw.shape[0]\n",
    "p = X_raw.shape[1]\n",
    "print(\"n=\",n,\"p=\",p)\n",
    "\n",
    "# Drop timestamp variable (only when necessary)\n",
    "#Ram print(\"Drop timestamp variable\")\n",
    "#Ram X_raw = X_raw.drop(['timestamp'], 1)\n"
   ]
  },
  {
   "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>EURUSDopen</th>\n",
       "      <th>EURUSDhigh</th>\n",
       "      <th>EURUSDlow</th>\n",
       "      <th>EURUSDclose</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>timestamp</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1483322400</th>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483323300</th>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05172</td>\n",
       "      <td>1.05178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483324200</th>\n",
       "      <td>1.05185</td>\n",
       "      <td>1.05209</td>\n",
       "      <td>1.05168</td>\n",
       "      <td>1.05169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483325100</th>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05130</td>\n",
       "      <td>1.05150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483326000</th>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05175</td>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483326900</th>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05083</td>\n",
       "      <td>1.05122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483327800</th>\n",
       "      <td>1.05118</td>\n",
       "      <td>1.05144</td>\n",
       "      <td>1.05107</td>\n",
       "      <td>1.05107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483328700</th>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.04929</td>\n",
       "      <td>1.04929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483329600</th>\n",
       "      <td>1.04889</td>\n",
       "      <td>1.04904</td>\n",
       "      <td>1.04765</td>\n",
       "      <td>1.04802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483330500</th>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04826</td>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483331400</th>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04890</td>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483332300</th>\n",
       "      <td>1.04876</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04858</td>\n",
       "      <td>1.04868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483333200</th>\n",
       "      <td>1.04866</td>\n",
       "      <td>1.04885</td>\n",
       "      <td>1.04852</td>\n",
       "      <td>1.04879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483334100</th>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04803</td>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483335000</th>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483335900</th>\n",
       "      <td>1.04822</td>\n",
       "      <td>1.04835</td>\n",
       "      <td>1.04791</td>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483336800</th>\n",
       "      <td>1.04805</td>\n",
       "      <td>1.04812</td>\n",
       "      <td>1.04801</td>\n",
       "      <td>1.04801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483337700</th>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04784</td>\n",
       "      <td>1.04798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483338600</th>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04776</td>\n",
       "      <td>1.04794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483339500</th>\n",
       "      <td>1.04795</td>\n",
       "      <td>1.04798</td>\n",
       "      <td>1.04768</td>\n",
       "      <td>1.04782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483340400</th>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04726</td>\n",
       "      <td>1.04726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483341300</th>\n",
       "      <td>1.04722</td>\n",
       "      <td>1.04746</td>\n",
       "      <td>1.04712</td>\n",
       "      <td>1.04714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483342200</th>\n",
       "      <td>1.04724</td>\n",
       "      <td>1.04749</td>\n",
       "      <td>1.04699</td>\n",
       "      <td>1.04711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483343100</th>\n",
       "      <td>1.04717</td>\n",
       "      <td>1.04720</td>\n",
       "      <td>1.04653</td>\n",
       "      <td>1.04659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483344000</th>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04615</td>\n",
       "      <td>1.04649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483344900</th>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04680</td>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483345800</th>\n",
       "      <td>1.04654</td>\n",
       "      <td>1.04664</td>\n",
       "      <td>1.04644</td>\n",
       "      <td>1.04655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483346700</th>\n",
       "      <td>1.04658</td>\n",
       "      <td>1.04673</td>\n",
       "      <td>1.04657</td>\n",
       "      <td>1.04668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483347600</th>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04679</td>\n",
       "      <td>1.04649</td>\n",
       "      <td>1.04677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1483348500</th>\n",
       "      <td>1.04676</td>\n",
       "      <td>1.04695</td>\n",
       "      <td>1.04661</td>\n",
       "      <td>1.04695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514539800</th>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19911</td>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514540700</th>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19946</td>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514541600</th>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19962</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514542500</th>\n",
       "      <td>1.19951</td>\n",
       "      <td>1.19955</td>\n",
       "      <td>1.19880</td>\n",
       "      <td>1.19911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514543400</th>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514544300</th>\n",
       "      <td>1.19977</td>\n",
       "      <td>1.20069</td>\n",
       "      <td>1.19973</td>\n",
       "      <td>1.20069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514545200</th>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20023</td>\n",
       "      <td>1.20082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514546100</th>\n",
       "      <td>1.20100</td>\n",
       "      <td>1.20128</td>\n",
       "      <td>1.20058</td>\n",
       "      <td>1.20070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514547000</th>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514547900</th>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514548800</th>\n",
       "      <td>1.20220</td>\n",
       "      <td>1.20247</td>\n",
       "      <td>1.20193</td>\n",
       "      <td>1.20208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514549700</th>\n",
       "      <td>1.20197</td>\n",
       "      <td>1.20238</td>\n",
       "      <td>1.20195</td>\n",
       "      <td>1.20229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514550600</th>\n",
       "      <td>1.20243</td>\n",
       "      <td>1.20255</td>\n",
       "      <td>1.20201</td>\n",
       "      <td>1.20210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514551500</th>\n",
       "      <td>1.20222</td>\n",
       "      <td>1.20236</td>\n",
       "      <td>1.20196</td>\n",
       "      <td>1.20197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514552400</th>\n",
       "      <td>1.20214</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20200</td>\n",
       "      <td>1.20225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514553300</th>\n",
       "      <td>1.20225</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20189</td>\n",
       "      <td>1.20189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514554200</th>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20150</td>\n",
       "      <td>1.20155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514555100</th>\n",
       "      <td>1.20164</td>\n",
       "      <td>1.20167</td>\n",
       "      <td>1.20124</td>\n",
       "      <td>1.20133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514556000</th>\n",
       "      <td>1.20134</td>\n",
       "      <td>1.20138</td>\n",
       "      <td>1.20094</td>\n",
       "      <td>1.20094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514556900</th>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20103</td>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514557800</th>\n",
       "      <td>1.20098</td>\n",
       "      <td>1.20099</td>\n",
       "      <td>1.20078</td>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514558700</th>\n",
       "      <td>1.20082</td>\n",
       "      <td>1.20114</td>\n",
       "      <td>1.20071</td>\n",
       "      <td>1.20106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514559600</th>\n",
       "      <td>1.20092</td>\n",
       "      <td>1.20104</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514560500</th>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20049</td>\n",
       "      <td>1.20053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514561400</th>\n",
       "      <td>1.20051</td>\n",
       "      <td>1.20052</td>\n",
       "      <td>1.20019</td>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514562300</th>\n",
       "      <td>1.20025</td>\n",
       "      <td>1.20028</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.19983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514563200</th>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20035</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514564100</th>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.19958</td>\n",
       "      <td>1.19963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514565000</th>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19988</td>\n",
       "      <td>1.19963</td>\n",
       "      <td>1.19966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514565900</th>\n",
       "      <td>1.19966</td>\n",
       "      <td>1.20009</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19982</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>34716 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            EURUSDopen  EURUSDhigh  EURUSDlow  EURUSDclose\n",
       "timestamp                                                 \n",
       "1483322400     1.05155     1.05213    1.05155      1.05206\n",
       "1483323300     1.05213     1.05213    1.05172      1.05178\n",
       "1483324200     1.05185     1.05209    1.05168      1.05169\n",
       "1483325100     1.05188     1.05188    1.05130      1.05150\n",
       "1483326000     1.05152     1.05175    1.05149      1.05156\n",
       "1483326900     1.05149     1.05152    1.05083      1.05122\n",
       "1483327800     1.05118     1.05144    1.05107      1.05107\n",
       "1483328700     1.05101     1.05101    1.04929      1.04929\n",
       "1483329600     1.04889     1.04904    1.04765      1.04802\n",
       "1483330500     1.04793     1.04826    1.04793      1.04816\n",
       "1483331400     1.04818     1.04890    1.04818      1.04874\n",
       "1483332300     1.04876     1.04884    1.04858      1.04868\n",
       "1483333200     1.04866     1.04885    1.04852      1.04879\n",
       "1483334100     1.04884     1.04884    1.04803      1.04803\n",
       "1483335000     1.04802     1.04831    1.04802      1.04831\n",
       "1483335900     1.04822     1.04835    1.04791      1.04803\n",
       "1483336800     1.04805     1.04812    1.04801      1.04801\n",
       "1483337700     1.04806     1.04806    1.04784      1.04798\n",
       "1483338600     1.04800     1.04800    1.04776      1.04794\n",
       "1483339500     1.04795     1.04798    1.04768      1.04782\n",
       "1483340400     1.04782     1.04782    1.04726      1.04726\n",
       "1483341300     1.04722     1.04746    1.04712      1.04714\n",
       "1483342200     1.04724     1.04749    1.04699      1.04711\n",
       "1483343100     1.04717     1.04720    1.04653      1.04659\n",
       "1483344000     1.04655     1.04655    1.04615      1.04649\n",
       "1483344900     1.04645     1.04680    1.04645      1.04657\n",
       "1483345800     1.04654     1.04664    1.04644      1.04655\n",
       "1483346700     1.04658     1.04673    1.04657      1.04668\n",
       "1483347600     1.04655     1.04679    1.04649      1.04677\n",
       "1483348500     1.04676     1.04695    1.04661      1.04695\n",
       "...                ...         ...        ...          ...\n",
       "1514539800     1.19882     1.19911    1.19882      1.19907\n",
       "1514540700     1.19894     1.19946    1.19894      1.19926\n",
       "1514541600     1.19927     1.19962    1.19927      1.19952\n",
       "1514542500     1.19951     1.19955    1.19880      1.19911\n",
       "1514543400     1.19905     1.19981    1.19905      1.19981\n",
       "1514544300     1.19977     1.20069    1.19973      1.20069\n",
       "1514545200     1.20097     1.20097    1.20023      1.20082\n",
       "1514546100     1.20100     1.20128    1.20058      1.20070\n",
       "1514547000     1.20072     1.20141    1.20072      1.20141\n",
       "1514547900     1.20142     1.20215    1.20142      1.20215\n",
       "1514548800     1.20220     1.20247    1.20193      1.20208\n",
       "1514549700     1.20197     1.20238    1.20195      1.20229\n",
       "1514550600     1.20243     1.20255    1.20201      1.20210\n",
       "1514551500     1.20222     1.20236    1.20196      1.20197\n",
       "1514552400     1.20214     1.20231    1.20200      1.20225\n",
       "1514553300     1.20225     1.20231    1.20189      1.20189\n",
       "1514554200     1.20181     1.20181    1.20150      1.20155\n",
       "1514555100     1.20164     1.20167    1.20124      1.20133\n",
       "1514556000     1.20134     1.20138    1.20094      1.20094\n",
       "1514556900     1.20087     1.20103    1.20087      1.20100\n",
       "1514557800     1.20098     1.20099    1.20078      1.20081\n",
       "1514558700     1.20082     1.20114    1.20071      1.20106\n",
       "1514559600     1.20092     1.20104    1.20072      1.20081\n",
       "1514560500     1.20080     1.20080    1.20049      1.20053\n",
       "1514561400     1.20051     1.20052    1.20019      1.20024\n",
       "1514562300     1.20025     1.20028    1.19978      1.19983\n",
       "1514563200     1.19978     1.20035    1.19978      1.20024\n",
       "1514564100     1.20033     1.20033    1.19958      1.19963\n",
       "1514565000     1.19981     1.19988    1.19963      1.19966\n",
       "1514565900     1.19966     1.20009    1.19927      1.19982\n",
       "\n",
       "[34716 rows x 4 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_raw"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Target\n",
    "# We use as target one of the symbols rate, i.e. \"EURUSD\". That is we try to predict next value of EURUSD\n",
    "lag = -1\n",
    "y_raw = df2.loc[:,\"EURUSDclose\"].shift(periods=lag)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "timestamp\n",
       "1483322400    1.05178\n",
       "1483323300    1.05169\n",
       "1483324200    1.05150\n",
       "1483325100    1.05156\n",
       "1483326000    1.05122\n",
       "1483326900    1.05107\n",
       "1483327800    1.04929\n",
       "1483328700    1.04802\n",
       "1483329600    1.04816\n",
       "1483330500    1.04874\n",
       "1483331400    1.04868\n",
       "1483332300    1.04879\n",
       "1483333200    1.04803\n",
       "1483334100    1.04831\n",
       "1483335000    1.04803\n",
       "1483335900    1.04801\n",
       "1483336800    1.04798\n",
       "1483337700    1.04794\n",
       "1483338600    1.04782\n",
       "1483339500    1.04726\n",
       "1483340400    1.04714\n",
       "1483341300    1.04711\n",
       "1483342200    1.04659\n",
       "1483343100    1.04649\n",
       "1483344000    1.04657\n",
       "1483344900    1.04655\n",
       "1483345800    1.04668\n",
       "1483346700    1.04677\n",
       "1483347600    1.04695\n",
       "1483348500    1.04733\n",
       "               ...   \n",
       "1514539800    1.19926\n",
       "1514540700    1.19952\n",
       "1514541600    1.19911\n",
       "1514542500    1.19981\n",
       "1514543400    1.20069\n",
       "1514544300    1.20082\n",
       "1514545200    1.20070\n",
       "1514546100    1.20141\n",
       "1514547000    1.20215\n",
       "1514547900    1.20208\n",
       "1514548800    1.20229\n",
       "1514549700    1.20210\n",
       "1514550600    1.20197\n",
       "1514551500    1.20225\n",
       "1514552400    1.20189\n",
       "1514553300    1.20155\n",
       "1514554200    1.20133\n",
       "1514555100    1.20094\n",
       "1514556000    1.20100\n",
       "1514556900    1.20081\n",
       "1514557800    1.20106\n",
       "1514558700    1.20081\n",
       "1514559600    1.20053\n",
       "1514560500    1.20024\n",
       "1514561400    1.19983\n",
       "1514562300    1.20024\n",
       "1514563200    1.19963\n",
       "1514564100    1.19966\n",
       "1514565000    1.19982\n",
       "1514565900    1.19982\n",
       "Name: EURUSDclose, Length: 34716, dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#\n",
    "# Removal of Null values**\n",
    "# Now since there still exists 'NaN' values in our target dataframe, and these are Null values,\n",
    "# we have to do something about them. In here, I will just do the naive thing of replacing these NaNs\n",
    "# with previous value because it is only the last value an error is negligible as such:\n",
    "\n",
    "# Filling gaps forward\n",
    "y_raw = y_raw.fillna(method='pad')\n",
    "y_raw\n",
    "\n",
    "# Drop timestamp variable (only when necessary)\n",
    "#Ram print(\"Drop timestamp variable\")\n",
    "#Ram y_raw = data.drop(['timestamp'], 1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f5232989c18>]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#A quick look at the dataframe time series using pyplot.plot(X_raw['EURUSD']):\n",
    "plt.plot(X_raw['EURUSDclose'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Split data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of samples:  34716 \n",
      "Samples in train set:  31244 \n",
      "Samples in test set:  3472\n",
      "X_train.shape =  (31244, 4) y_train.shape = (31244,) \n",
      "X_test.shape =   (3472, 4) y_test.shape =  (3472,)\n"
     ]
    }
   ],
   "source": [
    "# split into train and test sets\n",
    "\n",
    "# Total samples\n",
    "nsamples = n\n",
    "\n",
    "# Splitting into train (90%) and test (10%) sets\n",
    "split = 90 # training split% ; test (100-split)%\n",
    "jindex = nsamples*split//100 # Index for slicing the samples\n",
    "\n",
    "# Samples in train\n",
    "nsamples_train = jindex\n",
    "\n",
    "# Samples in test\n",
    "nsamples_test = nsamples - nsamples_train\n",
    "print(\"Total number of samples: \",nsamples,\"\\nSamples in train set: \", nsamples_train,\n",
    "      \"\\nSamples in test set: \",nsamples_test)\n",
    "\n",
    "# Here are train and test samples\n",
    "X_train = X_raw.values[:jindex, :]\n",
    "y_train = y_raw.values[:jindex]\n",
    "\n",
    "X_test = X_raw.values[jindex:, :]\n",
    "y_test = y_raw.values[jindex:]\n",
    "\n",
    "print(\"X_train.shape = \", X_train.shape, \"y_train.shape =\", y_train.shape, \"\\nX_test.shape =  \",\n",
    "      X_test.shape, \"y_test.shape = \", y_test.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train\n"
     ]
    },
    {
     "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>EURUSDopen</th>\n",
       "      <th>EURUSDhigh</th>\n",
       "      <th>EURUSDlow</th>\n",
       "      <th>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05155</td>\n",
       "      <td>1.05206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05213</td>\n",
       "      <td>1.05172</td>\n",
       "      <td>1.05178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.05185</td>\n",
       "      <td>1.05209</td>\n",
       "      <td>1.05168</td>\n",
       "      <td>1.05169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05188</td>\n",
       "      <td>1.05130</td>\n",
       "      <td>1.05150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05175</td>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.05149</td>\n",
       "      <td>1.05152</td>\n",
       "      <td>1.05083</td>\n",
       "      <td>1.05122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.05118</td>\n",
       "      <td>1.05144</td>\n",
       "      <td>1.05107</td>\n",
       "      <td>1.05107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.05101</td>\n",
       "      <td>1.04929</td>\n",
       "      <td>1.04929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.04889</td>\n",
       "      <td>1.04904</td>\n",
       "      <td>1.04765</td>\n",
       "      <td>1.04802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04826</td>\n",
       "      <td>1.04793</td>\n",
       "      <td>1.04816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04890</td>\n",
       "      <td>1.04818</td>\n",
       "      <td>1.04874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.04876</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04858</td>\n",
       "      <td>1.04868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.04866</td>\n",
       "      <td>1.04885</td>\n",
       "      <td>1.04852</td>\n",
       "      <td>1.04879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04884</td>\n",
       "      <td>1.04803</td>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "      <td>1.04802</td>\n",
       "      <td>1.04831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.04822</td>\n",
       "      <td>1.04835</td>\n",
       "      <td>1.04791</td>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.04805</td>\n",
       "      <td>1.04812</td>\n",
       "      <td>1.04801</td>\n",
       "      <td>1.04801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04806</td>\n",
       "      <td>1.04784</td>\n",
       "      <td>1.04798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04800</td>\n",
       "      <td>1.04776</td>\n",
       "      <td>1.04794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.04795</td>\n",
       "      <td>1.04798</td>\n",
       "      <td>1.04768</td>\n",
       "      <td>1.04782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04782</td>\n",
       "      <td>1.04726</td>\n",
       "      <td>1.04726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.04722</td>\n",
       "      <td>1.04746</td>\n",
       "      <td>1.04712</td>\n",
       "      <td>1.04714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1.04724</td>\n",
       "      <td>1.04749</td>\n",
       "      <td>1.04699</td>\n",
       "      <td>1.04711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.04717</td>\n",
       "      <td>1.04720</td>\n",
       "      <td>1.04653</td>\n",
       "      <td>1.04659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04615</td>\n",
       "      <td>1.04649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04680</td>\n",
       "      <td>1.04645</td>\n",
       "      <td>1.04657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.04654</td>\n",
       "      <td>1.04664</td>\n",
       "      <td>1.04644</td>\n",
       "      <td>1.04655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1.04658</td>\n",
       "      <td>1.04673</td>\n",
       "      <td>1.04657</td>\n",
       "      <td>1.04668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1.04655</td>\n",
       "      <td>1.04679</td>\n",
       "      <td>1.04649</td>\n",
       "      <td>1.04677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1.04676</td>\n",
       "      <td>1.04695</td>\n",
       "      <td>1.04661</td>\n",
       "      <td>1.04695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31214</th>\n",
       "      <td>1.18439</td>\n",
       "      <td>1.18439</td>\n",
       "      <td>1.18389</td>\n",
       "      <td>1.18404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31215</th>\n",
       "      <td>1.18401</td>\n",
       "      <td>1.18419</td>\n",
       "      <td>1.18401</td>\n",
       "      <td>1.18404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31216</th>\n",
       "      <td>1.18465</td>\n",
       "      <td>1.18499</td>\n",
       "      <td>1.18463</td>\n",
       "      <td>1.18478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31217</th>\n",
       "      <td>1.18481</td>\n",
       "      <td>1.18481</td>\n",
       "      <td>1.18423</td>\n",
       "      <td>1.18447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31218</th>\n",
       "      <td>1.18453</td>\n",
       "      <td>1.18510</td>\n",
       "      <td>1.18453</td>\n",
       "      <td>1.18502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31219</th>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18461</td>\n",
       "      <td>1.18463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31220</th>\n",
       "      <td>1.18454</td>\n",
       "      <td>1.18472</td>\n",
       "      <td>1.18410</td>\n",
       "      <td>1.18460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31221</th>\n",
       "      <td>1.18453</td>\n",
       "      <td>1.18487</td>\n",
       "      <td>1.18440</td>\n",
       "      <td>1.18471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31222</th>\n",
       "      <td>1.18476</td>\n",
       "      <td>1.18529</td>\n",
       "      <td>1.18460</td>\n",
       "      <td>1.18496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31223</th>\n",
       "      <td>1.18476</td>\n",
       "      <td>1.18476</td>\n",
       "      <td>1.18421</td>\n",
       "      <td>1.18429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31224</th>\n",
       "      <td>1.18432</td>\n",
       "      <td>1.18478</td>\n",
       "      <td>1.18430</td>\n",
       "      <td>1.18474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31225</th>\n",
       "      <td>1.18484</td>\n",
       "      <td>1.18484</td>\n",
       "      <td>1.18449</td>\n",
       "      <td>1.18449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31226</th>\n",
       "      <td>1.18442</td>\n",
       "      <td>1.18464</td>\n",
       "      <td>1.18430</td>\n",
       "      <td>1.18430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31227</th>\n",
       "      <td>1.18435</td>\n",
       "      <td>1.18489</td>\n",
       "      <td>1.18430</td>\n",
       "      <td>1.18489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31228</th>\n",
       "      <td>1.18492</td>\n",
       "      <td>1.18503</td>\n",
       "      <td>1.18461</td>\n",
       "      <td>1.18477</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31229</th>\n",
       "      <td>1.18465</td>\n",
       "      <td>1.18519</td>\n",
       "      <td>1.18465</td>\n",
       "      <td>1.18484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31230</th>\n",
       "      <td>1.18482</td>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18473</td>\n",
       "      <td>1.18476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31231</th>\n",
       "      <td>1.18484</td>\n",
       "      <td>1.18492</td>\n",
       "      <td>1.18460</td>\n",
       "      <td>1.18466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31232</th>\n",
       "      <td>1.18457</td>\n",
       "      <td>1.18491</td>\n",
       "      <td>1.18457</td>\n",
       "      <td>1.18464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31233</th>\n",
       "      <td>1.18467</td>\n",
       "      <td>1.18486</td>\n",
       "      <td>1.18457</td>\n",
       "      <td>1.18457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31234</th>\n",
       "      <td>1.18460</td>\n",
       "      <td>1.18531</td>\n",
       "      <td>1.18451</td>\n",
       "      <td>1.18531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31235</th>\n",
       "      <td>1.18527</td>\n",
       "      <td>1.18534</td>\n",
       "      <td>1.18460</td>\n",
       "      <td>1.18460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31236</th>\n",
       "      <td>1.18430</td>\n",
       "      <td>1.18439</td>\n",
       "      <td>1.18402</td>\n",
       "      <td>1.18437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31237</th>\n",
       "      <td>1.18442</td>\n",
       "      <td>1.18470</td>\n",
       "      <td>1.18432</td>\n",
       "      <td>1.18470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31238</th>\n",
       "      <td>1.18464</td>\n",
       "      <td>1.18488</td>\n",
       "      <td>1.18461</td>\n",
       "      <td>1.18483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31239</th>\n",
       "      <td>1.18486</td>\n",
       "      <td>1.18497</td>\n",
       "      <td>1.18463</td>\n",
       "      <td>1.18496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31240</th>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18526</td>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31241</th>\n",
       "      <td>1.18498</td>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18474</td>\n",
       "      <td>1.18505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31242</th>\n",
       "      <td>1.18518</td>\n",
       "      <td>1.18539</td>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31243</th>\n",
       "      <td>1.18511</td>\n",
       "      <td>1.18537</td>\n",
       "      <td>1.18490</td>\n",
       "      <td>1.18490</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31244 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       EURUSDopen  EURUSDhigh  EURUSDlow  EURUSDclose\n",
       "0         1.05155     1.05213    1.05155      1.05206\n",
       "1         1.05213     1.05213    1.05172      1.05178\n",
       "2         1.05185     1.05209    1.05168      1.05169\n",
       "3         1.05188     1.05188    1.05130      1.05150\n",
       "4         1.05152     1.05175    1.05149      1.05156\n",
       "5         1.05149     1.05152    1.05083      1.05122\n",
       "6         1.05118     1.05144    1.05107      1.05107\n",
       "7         1.05101     1.05101    1.04929      1.04929\n",
       "8         1.04889     1.04904    1.04765      1.04802\n",
       "9         1.04793     1.04826    1.04793      1.04816\n",
       "10        1.04818     1.04890    1.04818      1.04874\n",
       "11        1.04876     1.04884    1.04858      1.04868\n",
       "12        1.04866     1.04885    1.04852      1.04879\n",
       "13        1.04884     1.04884    1.04803      1.04803\n",
       "14        1.04802     1.04831    1.04802      1.04831\n",
       "15        1.04822     1.04835    1.04791      1.04803\n",
       "16        1.04805     1.04812    1.04801      1.04801\n",
       "17        1.04806     1.04806    1.04784      1.04798\n",
       "18        1.04800     1.04800    1.04776      1.04794\n",
       "19        1.04795     1.04798    1.04768      1.04782\n",
       "20        1.04782     1.04782    1.04726      1.04726\n",
       "21        1.04722     1.04746    1.04712      1.04714\n",
       "22        1.04724     1.04749    1.04699      1.04711\n",
       "23        1.04717     1.04720    1.04653      1.04659\n",
       "24        1.04655     1.04655    1.04615      1.04649\n",
       "25        1.04645     1.04680    1.04645      1.04657\n",
       "26        1.04654     1.04664    1.04644      1.04655\n",
       "27        1.04658     1.04673    1.04657      1.04668\n",
       "28        1.04655     1.04679    1.04649      1.04677\n",
       "29        1.04676     1.04695    1.04661      1.04695\n",
       "...           ...         ...        ...          ...\n",
       "31214     1.18439     1.18439    1.18389      1.18404\n",
       "31215     1.18401     1.18419    1.18401      1.18404\n",
       "31216     1.18465     1.18499    1.18463      1.18478\n",
       "31217     1.18481     1.18481    1.18423      1.18447\n",
       "31218     1.18453     1.18510    1.18453      1.18502\n",
       "31219     1.18506     1.18506    1.18461      1.18463\n",
       "31220     1.18454     1.18472    1.18410      1.18460\n",
       "31221     1.18453     1.18487    1.18440      1.18471\n",
       "31222     1.18476     1.18529    1.18460      1.18496\n",
       "31223     1.18476     1.18476    1.18421      1.18429\n",
       "31224     1.18432     1.18478    1.18430      1.18474\n",
       "31225     1.18484     1.18484    1.18449      1.18449\n",
       "31226     1.18442     1.18464    1.18430      1.18430\n",
       "31227     1.18435     1.18489    1.18430      1.18489\n",
       "31228     1.18492     1.18503    1.18461      1.18477\n",
       "31229     1.18465     1.18519    1.18465      1.18484\n",
       "31230     1.18482     1.18506    1.18473      1.18476\n",
       "31231     1.18484     1.18492    1.18460      1.18466\n",
       "31232     1.18457     1.18491    1.18457      1.18464\n",
       "31233     1.18467     1.18486    1.18457      1.18457\n",
       "31234     1.18460     1.18531    1.18451      1.18531\n",
       "31235     1.18527     1.18534    1.18460      1.18460\n",
       "31236     1.18430     1.18439    1.18402      1.18437\n",
       "31237     1.18442     1.18470    1.18432      1.18470\n",
       "31238     1.18464     1.18488    1.18461      1.18483\n",
       "31239     1.18486     1.18497    1.18463      1.18496\n",
       "31240     1.18506     1.18526    1.18506      1.18507\n",
       "31241     1.18498     1.18506    1.18474      1.18505\n",
       "31242     1.18518     1.18539    1.18514      1.18514\n",
       "31243     1.18511     1.18537    1.18490      1.18490\n",
       "\n",
       "[31244 rows x 4 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#X_train as dataframe (optional, only for printing. See note in the beginning)\n",
    "X_Train = pd.DataFrame(data=X_train)\n",
    "X_Train.columns = X_raw.columns\n",
    "print(\"X_train\")\n",
    "X_Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_test\n"
     ]
    },
    {
     "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>EURUSDopen</th>\n",
       "      <th>EURUSDhigh</th>\n",
       "      <th>EURUSDlow</th>\n",
       "      <th>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.18488</td>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18488</td>\n",
       "      <td>1.18506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.18509</td>\n",
       "      <td>1.18531</td>\n",
       "      <td>1.18509</td>\n",
       "      <td>1.18522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.18519</td>\n",
       "      <td>1.18522</td>\n",
       "      <td>1.18510</td>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.18510</td>\n",
       "      <td>1.18513</td>\n",
       "      <td>1.18483</td>\n",
       "      <td>1.18488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.18504</td>\n",
       "      <td>1.18504</td>\n",
       "      <td>1.18485</td>\n",
       "      <td>1.18485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.18491</td>\n",
       "      <td>1.18491</td>\n",
       "      <td>1.18462</td>\n",
       "      <td>1.18469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.18467</td>\n",
       "      <td>1.18506</td>\n",
       "      <td>1.18467</td>\n",
       "      <td>1.18497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.18502</td>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.18494</td>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.18512</td>\n",
       "      <td>1.18526</td>\n",
       "      <td>1.18512</td>\n",
       "      <td>1.18524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.18523</td>\n",
       "      <td>1.18527</td>\n",
       "      <td>1.18519</td>\n",
       "      <td>1.18521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.18522</td>\n",
       "      <td>1.18522</td>\n",
       "      <td>1.18494</td>\n",
       "      <td>1.18498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.18501</td>\n",
       "      <td>1.18516</td>\n",
       "      <td>1.18493</td>\n",
       "      <td>1.18502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.18491</td>\n",
       "      <td>1.18499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.18502</td>\n",
       "      <td>1.18511</td>\n",
       "      <td>1.18486</td>\n",
       "      <td>1.18497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.18485</td>\n",
       "      <td>1.18503</td>\n",
       "      <td>1.18463</td>\n",
       "      <td>1.18476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.18478</td>\n",
       "      <td>1.18492</td>\n",
       "      <td>1.18473</td>\n",
       "      <td>1.18490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.18478</td>\n",
       "      <td>1.18507</td>\n",
       "      <td>1.18452</td>\n",
       "      <td>1.18471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.18473</td>\n",
       "      <td>1.18486</td>\n",
       "      <td>1.18468</td>\n",
       "      <td>1.18474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.18497</td>\n",
       "      <td>1.18497</td>\n",
       "      <td>1.18474</td>\n",
       "      <td>1.18474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.18489</td>\n",
       "      <td>1.18498</td>\n",
       "      <td>1.18476</td>\n",
       "      <td>1.18496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.18498</td>\n",
       "      <td>1.18513</td>\n",
       "      <td>1.18477</td>\n",
       "      <td>1.18513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.18513</td>\n",
       "      <td>1.18513</td>\n",
       "      <td>1.18481</td>\n",
       "      <td>1.18484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1.18481</td>\n",
       "      <td>1.18527</td>\n",
       "      <td>1.18481</td>\n",
       "      <td>1.18519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.18520</td>\n",
       "      <td>1.18500</td>\n",
       "      <td>1.18503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1.18501</td>\n",
       "      <td>1.18544</td>\n",
       "      <td>1.18500</td>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.18493</td>\n",
       "      <td>1.18493</td>\n",
       "      <td>1.18446</td>\n",
       "      <td>1.18466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.18469</td>\n",
       "      <td>1.18469</td>\n",
       "      <td>1.18404</td>\n",
       "      <td>1.18414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1.18412</td>\n",
       "      <td>1.18427</td>\n",
       "      <td>1.18398</td>\n",
       "      <td>1.18427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1.18429</td>\n",
       "      <td>1.18429</td>\n",
       "      <td>1.18371</td>\n",
       "      <td>1.18378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1.18396</td>\n",
       "      <td>1.18471</td>\n",
       "      <td>1.18396</td>\n",
       "      <td>1.18471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3442</th>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19911</td>\n",
       "      <td>1.19882</td>\n",
       "      <td>1.19907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3443</th>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19946</td>\n",
       "      <td>1.19894</td>\n",
       "      <td>1.19926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3444</th>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19962</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3445</th>\n",
       "      <td>1.19951</td>\n",
       "      <td>1.19955</td>\n",
       "      <td>1.19880</td>\n",
       "      <td>1.19911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3446</th>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19905</td>\n",
       "      <td>1.19981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3447</th>\n",
       "      <td>1.19977</td>\n",
       "      <td>1.20069</td>\n",
       "      <td>1.19973</td>\n",
       "      <td>1.20069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3448</th>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20097</td>\n",
       "      <td>1.20023</td>\n",
       "      <td>1.20082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3449</th>\n",
       "      <td>1.20100</td>\n",
       "      <td>1.20128</td>\n",
       "      <td>1.20058</td>\n",
       "      <td>1.20070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3450</th>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3451</th>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "      <td>1.20142</td>\n",
       "      <td>1.20215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3452</th>\n",
       "      <td>1.20220</td>\n",
       "      <td>1.20247</td>\n",
       "      <td>1.20193</td>\n",
       "      <td>1.20208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3453</th>\n",
       "      <td>1.20197</td>\n",
       "      <td>1.20238</td>\n",
       "      <td>1.20195</td>\n",
       "      <td>1.20229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3454</th>\n",
       "      <td>1.20243</td>\n",
       "      <td>1.20255</td>\n",
       "      <td>1.20201</td>\n",
       "      <td>1.20210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3455</th>\n",
       "      <td>1.20222</td>\n",
       "      <td>1.20236</td>\n",
       "      <td>1.20196</td>\n",
       "      <td>1.20197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3456</th>\n",
       "      <td>1.20214</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20200</td>\n",
       "      <td>1.20225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3457</th>\n",
       "      <td>1.20225</td>\n",
       "      <td>1.20231</td>\n",
       "      <td>1.20189</td>\n",
       "      <td>1.20189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3458</th>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20181</td>\n",
       "      <td>1.20150</td>\n",
       "      <td>1.20155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3459</th>\n",
       "      <td>1.20164</td>\n",
       "      <td>1.20167</td>\n",
       "      <td>1.20124</td>\n",
       "      <td>1.20133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3460</th>\n",
       "      <td>1.20134</td>\n",
       "      <td>1.20138</td>\n",
       "      <td>1.20094</td>\n",
       "      <td>1.20094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3461</th>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20103</td>\n",
       "      <td>1.20087</td>\n",
       "      <td>1.20100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3462</th>\n",
       "      <td>1.20098</td>\n",
       "      <td>1.20099</td>\n",
       "      <td>1.20078</td>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3463</th>\n",
       "      <td>1.20082</td>\n",
       "      <td>1.20114</td>\n",
       "      <td>1.20071</td>\n",
       "      <td>1.20106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3464</th>\n",
       "      <td>1.20092</td>\n",
       "      <td>1.20104</td>\n",
       "      <td>1.20072</td>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3465</th>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20080</td>\n",
       "      <td>1.20049</td>\n",
       "      <td>1.20053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3466</th>\n",
       "      <td>1.20051</td>\n",
       "      <td>1.20052</td>\n",
       "      <td>1.20019</td>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3467</th>\n",
       "      <td>1.20025</td>\n",
       "      <td>1.20028</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.19983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3468</th>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20035</td>\n",
       "      <td>1.19978</td>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3469</th>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.20033</td>\n",
       "      <td>1.19958</td>\n",
       "      <td>1.19963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3470</th>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.19988</td>\n",
       "      <td>1.19963</td>\n",
       "      <td>1.19966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3471</th>\n",
       "      <td>1.19966</td>\n",
       "      <td>1.20009</td>\n",
       "      <td>1.19927</td>\n",
       "      <td>1.19982</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3472 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      EURUSDopen  EURUSDhigh  EURUSDlow  EURUSDclose\n",
       "0        1.18488     1.18506    1.18488      1.18506\n",
       "1        1.18509     1.18531    1.18509      1.18522\n",
       "2        1.18519     1.18522    1.18510      1.18514\n",
       "3        1.18510     1.18513    1.18483      1.18488\n",
       "4        1.18504     1.18504    1.18485      1.18485\n",
       "5        1.18491     1.18491    1.18462      1.18469\n",
       "6        1.18467     1.18506    1.18467      1.18497\n",
       "7        1.18502     1.18514    1.18494      1.18514\n",
       "8        1.18512     1.18526    1.18512      1.18524\n",
       "9        1.18523     1.18527    1.18519      1.18521\n",
       "10       1.18522     1.18522    1.18494      1.18498\n",
       "11       1.18501     1.18516    1.18493      1.18502\n",
       "12       1.18514     1.18514    1.18491      1.18499\n",
       "13       1.18502     1.18511    1.18486      1.18497\n",
       "14       1.18485     1.18503    1.18463      1.18476\n",
       "15       1.18478     1.18492    1.18473      1.18490\n",
       "16       1.18478     1.18507    1.18452      1.18471\n",
       "17       1.18473     1.18486    1.18468      1.18474\n",
       "18       1.18497     1.18497    1.18474      1.18474\n",
       "19       1.18489     1.18498    1.18476      1.18496\n",
       "20       1.18498     1.18513    1.18477      1.18513\n",
       "21       1.18513     1.18513    1.18481      1.18484\n",
       "22       1.18481     1.18527    1.18481      1.18519\n",
       "23       1.18514     1.18520    1.18500      1.18503\n",
       "24       1.18501     1.18544    1.18500      1.18514\n",
       "25       1.18493     1.18493    1.18446      1.18466\n",
       "26       1.18469     1.18469    1.18404      1.18414\n",
       "27       1.18412     1.18427    1.18398      1.18427\n",
       "28       1.18429     1.18429    1.18371      1.18378\n",
       "29       1.18396     1.18471    1.18396      1.18471\n",
       "...          ...         ...        ...          ...\n",
       "3442     1.19882     1.19911    1.19882      1.19907\n",
       "3443     1.19894     1.19946    1.19894      1.19926\n",
       "3444     1.19927     1.19962    1.19927      1.19952\n",
       "3445     1.19951     1.19955    1.19880      1.19911\n",
       "3446     1.19905     1.19981    1.19905      1.19981\n",
       "3447     1.19977     1.20069    1.19973      1.20069\n",
       "3448     1.20097     1.20097    1.20023      1.20082\n",
       "3449     1.20100     1.20128    1.20058      1.20070\n",
       "3450     1.20072     1.20141    1.20072      1.20141\n",
       "3451     1.20142     1.20215    1.20142      1.20215\n",
       "3452     1.20220     1.20247    1.20193      1.20208\n",
       "3453     1.20197     1.20238    1.20195      1.20229\n",
       "3454     1.20243     1.20255    1.20201      1.20210\n",
       "3455     1.20222     1.20236    1.20196      1.20197\n",
       "3456     1.20214     1.20231    1.20200      1.20225\n",
       "3457     1.20225     1.20231    1.20189      1.20189\n",
       "3458     1.20181     1.20181    1.20150      1.20155\n",
       "3459     1.20164     1.20167    1.20124      1.20133\n",
       "3460     1.20134     1.20138    1.20094      1.20094\n",
       "3461     1.20087     1.20103    1.20087      1.20100\n",
       "3462     1.20098     1.20099    1.20078      1.20081\n",
       "3463     1.20082     1.20114    1.20071      1.20106\n",
       "3464     1.20092     1.20104    1.20072      1.20081\n",
       "3465     1.20080     1.20080    1.20049      1.20053\n",
       "3466     1.20051     1.20052    1.20019      1.20024\n",
       "3467     1.20025     1.20028    1.19978      1.19983\n",
       "3468     1.19978     1.20035    1.19978      1.20024\n",
       "3469     1.20033     1.20033    1.19958      1.19963\n",
       "3470     1.19981     1.19988    1.19963      1.19966\n",
       "3471     1.19966     1.20009    1.19927      1.19982\n",
       "\n",
       "[3472 rows x 4 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#X_test as dataframe (optional, only for printing. See note in the beginning)\n",
    "X_Test = pd.DataFrame(data=X_test)\n",
    "X_Test.columns = X_raw.columns\n",
    "print(\"X_test\")\n",
    "X_Test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Transform features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**\n",
    "\n",
    "Becareful not to write `X_test_std = sc.fit_transform(X_test)` instead of `X_test_std = sc.transform(X_test)`. In this case, it wouldn't make a great difference since the mean and standard deviation of the test set should be (quite) similar to the training set. However, this is not always the case in Forex market data, as has been well stablished in literature. The correct way is to re-use parameters from the training set if we are doing any kind of transformation. So, the test set should basically stand for \"new, unseen\" data.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale data\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "sc = StandardScaler()\n",
    "X_train_std = sc.fit_transform(X_train)\n",
    "X_test_std = sc.transform(X_test)\n",
    "\n",
    "y_train_std = sc.fit_transform(y_train.reshape(-1, 1))\n",
    "y_test_std = sc.transform(y_test.reshape(-1, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean: [1.12382284]\n",
      "Variance [0.00252333]\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean:\",sc.mean_)\n",
    "print(\"Variance\",sc.var_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train_std\n"
     ]
    },
    {
     "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>EURUSDopen</th>\n",
       "      <th>EURUSDhigh</th>\n",
       "      <th>EURUSDlow</th>\n",
       "      <th>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.437944</td>\n",
       "      <td>-1.432007</td>\n",
       "      <td>-1.433057</td>\n",
       "      <td>-1.428509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.426407</td>\n",
       "      <td>-1.432007</td>\n",
       "      <td>-1.429674</td>\n",
       "      <td>-1.434083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.431976</td>\n",
       "      <td>-1.432803</td>\n",
       "      <td>-1.430470</td>\n",
       "      <td>-1.435875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.431379</td>\n",
       "      <td>-1.436982</td>\n",
       "      <td>-1.438032</td>\n",
       "      <td>-1.439657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.438540</td>\n",
       "      <td>-1.439569</td>\n",
       "      <td>-1.434251</td>\n",
       "      <td>-1.438462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-1.439137</td>\n",
       "      <td>-1.444146</td>\n",
       "      <td>-1.447385</td>\n",
       "      <td>-1.445231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>-1.445303</td>\n",
       "      <td>-1.445738</td>\n",
       "      <td>-1.442609</td>\n",
       "      <td>-1.448217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-1.448685</td>\n",
       "      <td>-1.454294</td>\n",
       "      <td>-1.478032</td>\n",
       "      <td>-1.483652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-1.490855</td>\n",
       "      <td>-1.493496</td>\n",
       "      <td>-1.510669</td>\n",
       "      <td>-1.508934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-1.509951</td>\n",
       "      <td>-1.509017</td>\n",
       "      <td>-1.505097</td>\n",
       "      <td>-1.506147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>-1.504978</td>\n",
       "      <td>-1.496282</td>\n",
       "      <td>-1.500122</td>\n",
       "      <td>-1.494601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>-1.493441</td>\n",
       "      <td>-1.497476</td>\n",
       "      <td>-1.492161</td>\n",
       "      <td>-1.495795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>-1.495430</td>\n",
       "      <td>-1.497277</td>\n",
       "      <td>-1.493355</td>\n",
       "      <td>-1.493605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-1.491850</td>\n",
       "      <td>-1.497476</td>\n",
       "      <td>-1.503107</td>\n",
       "      <td>-1.508735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-1.508161</td>\n",
       "      <td>-1.508022</td>\n",
       "      <td>-1.503306</td>\n",
       "      <td>-1.503161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-1.504182</td>\n",
       "      <td>-1.507226</td>\n",
       "      <td>-1.505495</td>\n",
       "      <td>-1.508735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>-1.507564</td>\n",
       "      <td>-1.511803</td>\n",
       "      <td>-1.503505</td>\n",
       "      <td>-1.509133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-1.507365</td>\n",
       "      <td>-1.512997</td>\n",
       "      <td>-1.506888</td>\n",
       "      <td>-1.509730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-1.508558</td>\n",
       "      <td>-1.514191</td>\n",
       "      <td>-1.508480</td>\n",
       "      <td>-1.510526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>-1.509553</td>\n",
       "      <td>-1.514589</td>\n",
       "      <td>-1.510072</td>\n",
       "      <td>-1.512915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>-1.512139</td>\n",
       "      <td>-1.517773</td>\n",
       "      <td>-1.518430</td>\n",
       "      <td>-1.524063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>-1.524074</td>\n",
       "      <td>-1.524936</td>\n",
       "      <td>-1.521216</td>\n",
       "      <td>-1.526452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>-1.523676</td>\n",
       "      <td>-1.524339</td>\n",
       "      <td>-1.523803</td>\n",
       "      <td>-1.527049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>-1.525068</td>\n",
       "      <td>-1.530110</td>\n",
       "      <td>-1.532957</td>\n",
       "      <td>-1.537401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>-1.537401</td>\n",
       "      <td>-1.543045</td>\n",
       "      <td>-1.540520</td>\n",
       "      <td>-1.539392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>-1.539390</td>\n",
       "      <td>-1.538070</td>\n",
       "      <td>-1.534549</td>\n",
       "      <td>-1.537799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>-1.537600</td>\n",
       "      <td>-1.541254</td>\n",
       "      <td>-1.534748</td>\n",
       "      <td>-1.538197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>-1.536804</td>\n",
       "      <td>-1.539463</td>\n",
       "      <td>-1.532161</td>\n",
       "      <td>-1.535609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>-1.537401</td>\n",
       "      <td>-1.538269</td>\n",
       "      <td>-1.533753</td>\n",
       "      <td>-1.533818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>-1.533224</td>\n",
       "      <td>-1.535085</td>\n",
       "      <td>-1.531365</td>\n",
       "      <td>-1.530234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31214</th>\n",
       "      <td>1.204443</td>\n",
       "      <td>1.199858</td>\n",
       "      <td>1.200579</td>\n",
       "      <td>1.198837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31215</th>\n",
       "      <td>1.196885</td>\n",
       "      <td>1.195878</td>\n",
       "      <td>1.202967</td>\n",
       "      <td>1.198837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31216</th>\n",
       "      <td>1.209615</td>\n",
       "      <td>1.211797</td>\n",
       "      <td>1.215306</td>\n",
       "      <td>1.213568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31217</th>\n",
       "      <td>1.212798</td>\n",
       "      <td>1.208215</td>\n",
       "      <td>1.207345</td>\n",
       "      <td>1.207397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31218</th>\n",
       "      <td>1.207228</td>\n",
       "      <td>1.213986</td>\n",
       "      <td>1.213316</td>\n",
       "      <td>1.218346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31219</th>\n",
       "      <td>1.217771</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.214908</td>\n",
       "      <td>1.210582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31220</th>\n",
       "      <td>1.207427</td>\n",
       "      <td>1.206424</td>\n",
       "      <td>1.204758</td>\n",
       "      <td>1.209985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31221</th>\n",
       "      <td>1.207228</td>\n",
       "      <td>1.209409</td>\n",
       "      <td>1.210728</td>\n",
       "      <td>1.212175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31222</th>\n",
       "      <td>1.211803</td>\n",
       "      <td>1.217767</td>\n",
       "      <td>1.214709</td>\n",
       "      <td>1.217151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31223</th>\n",
       "      <td>1.211803</td>\n",
       "      <td>1.207220</td>\n",
       "      <td>1.206947</td>\n",
       "      <td>1.203814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31224</th>\n",
       "      <td>1.203051</td>\n",
       "      <td>1.207618</td>\n",
       "      <td>1.208738</td>\n",
       "      <td>1.212772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31225</th>\n",
       "      <td>1.213395</td>\n",
       "      <td>1.208812</td>\n",
       "      <td>1.212519</td>\n",
       "      <td>1.207795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31226</th>\n",
       "      <td>1.205040</td>\n",
       "      <td>1.204832</td>\n",
       "      <td>1.208738</td>\n",
       "      <td>1.204013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31227</th>\n",
       "      <td>1.203648</td>\n",
       "      <td>1.209807</td>\n",
       "      <td>1.208738</td>\n",
       "      <td>1.215758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31228</th>\n",
       "      <td>1.214986</td>\n",
       "      <td>1.212593</td>\n",
       "      <td>1.214908</td>\n",
       "      <td>1.213369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31229</th>\n",
       "      <td>1.209615</td>\n",
       "      <td>1.215777</td>\n",
       "      <td>1.215704</td>\n",
       "      <td>1.214763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31230</th>\n",
       "      <td>1.212997</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.217296</td>\n",
       "      <td>1.213170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31231</th>\n",
       "      <td>1.213395</td>\n",
       "      <td>1.210404</td>\n",
       "      <td>1.214709</td>\n",
       "      <td>1.211179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31232</th>\n",
       "      <td>1.208024</td>\n",
       "      <td>1.210205</td>\n",
       "      <td>1.214112</td>\n",
       "      <td>1.210781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31233</th>\n",
       "      <td>1.210013</td>\n",
       "      <td>1.209210</td>\n",
       "      <td>1.214112</td>\n",
       "      <td>1.209388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31234</th>\n",
       "      <td>1.208621</td>\n",
       "      <td>1.218165</td>\n",
       "      <td>1.212917</td>\n",
       "      <td>1.224119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31235</th>\n",
       "      <td>1.221948</td>\n",
       "      <td>1.218762</td>\n",
       "      <td>1.214709</td>\n",
       "      <td>1.209985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31236</th>\n",
       "      <td>1.202653</td>\n",
       "      <td>1.199858</td>\n",
       "      <td>1.203166</td>\n",
       "      <td>1.205406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31237</th>\n",
       "      <td>1.205040</td>\n",
       "      <td>1.206026</td>\n",
       "      <td>1.209136</td>\n",
       "      <td>1.211976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31238</th>\n",
       "      <td>1.209416</td>\n",
       "      <td>1.209608</td>\n",
       "      <td>1.214908</td>\n",
       "      <td>1.214564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31239</th>\n",
       "      <td>1.213792</td>\n",
       "      <td>1.211399</td>\n",
       "      <td>1.215306</td>\n",
       "      <td>1.217151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31240</th>\n",
       "      <td>1.217771</td>\n",
       "      <td>1.217170</td>\n",
       "      <td>1.223863</td>\n",
       "      <td>1.219341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31241</th>\n",
       "      <td>1.216179</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.217495</td>\n",
       "      <td>1.218943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31242</th>\n",
       "      <td>1.220158</td>\n",
       "      <td>1.219757</td>\n",
       "      <td>1.225455</td>\n",
       "      <td>1.220735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31243</th>\n",
       "      <td>1.218765</td>\n",
       "      <td>1.219359</td>\n",
       "      <td>1.220679</td>\n",
       "      <td>1.215957</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31244 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       EURUSDopen  EURUSDhigh  EURUSDlow  EURUSDclose\n",
       "0       -1.437944   -1.432007  -1.433057    -1.428509\n",
       "1       -1.426407   -1.432007  -1.429674    -1.434083\n",
       "2       -1.431976   -1.432803  -1.430470    -1.435875\n",
       "3       -1.431379   -1.436982  -1.438032    -1.439657\n",
       "4       -1.438540   -1.439569  -1.434251    -1.438462\n",
       "5       -1.439137   -1.444146  -1.447385    -1.445231\n",
       "6       -1.445303   -1.445738  -1.442609    -1.448217\n",
       "7       -1.448685   -1.454294  -1.478032    -1.483652\n",
       "8       -1.490855   -1.493496  -1.510669    -1.508934\n",
       "9       -1.509951   -1.509017  -1.505097    -1.506147\n",
       "10      -1.504978   -1.496282  -1.500122    -1.494601\n",
       "11      -1.493441   -1.497476  -1.492161    -1.495795\n",
       "12      -1.495430   -1.497277  -1.493355    -1.493605\n",
       "13      -1.491850   -1.497476  -1.503107    -1.508735\n",
       "14      -1.508161   -1.508022  -1.503306    -1.503161\n",
       "15      -1.504182   -1.507226  -1.505495    -1.508735\n",
       "16      -1.507564   -1.511803  -1.503505    -1.509133\n",
       "17      -1.507365   -1.512997  -1.506888    -1.509730\n",
       "18      -1.508558   -1.514191  -1.508480    -1.510526\n",
       "19      -1.509553   -1.514589  -1.510072    -1.512915\n",
       "20      -1.512139   -1.517773  -1.518430    -1.524063\n",
       "21      -1.524074   -1.524936  -1.521216    -1.526452\n",
       "22      -1.523676   -1.524339  -1.523803    -1.527049\n",
       "23      -1.525068   -1.530110  -1.532957    -1.537401\n",
       "24      -1.537401   -1.543045  -1.540520    -1.539392\n",
       "25      -1.539390   -1.538070  -1.534549    -1.537799\n",
       "26      -1.537600   -1.541254  -1.534748    -1.538197\n",
       "27      -1.536804   -1.539463  -1.532161    -1.535609\n",
       "28      -1.537401   -1.538269  -1.533753    -1.533818\n",
       "29      -1.533224   -1.535085  -1.531365    -1.530234\n",
       "...           ...         ...        ...          ...\n",
       "31214    1.204443    1.199858   1.200579     1.198837\n",
       "31215    1.196885    1.195878   1.202967     1.198837\n",
       "31216    1.209615    1.211797   1.215306     1.213568\n",
       "31217    1.212798    1.208215   1.207345     1.207397\n",
       "31218    1.207228    1.213986   1.213316     1.218346\n",
       "31219    1.217771    1.213190   1.214908     1.210582\n",
       "31220    1.207427    1.206424   1.204758     1.209985\n",
       "31221    1.207228    1.209409   1.210728     1.212175\n",
       "31222    1.211803    1.217767   1.214709     1.217151\n",
       "31223    1.211803    1.207220   1.206947     1.203814\n",
       "31224    1.203051    1.207618   1.208738     1.212772\n",
       "31225    1.213395    1.208812   1.212519     1.207795\n",
       "31226    1.205040    1.204832   1.208738     1.204013\n",
       "31227    1.203648    1.209807   1.208738     1.215758\n",
       "31228    1.214986    1.212593   1.214908     1.213369\n",
       "31229    1.209615    1.215777   1.215704     1.214763\n",
       "31230    1.212997    1.213190   1.217296     1.213170\n",
       "31231    1.213395    1.210404   1.214709     1.211179\n",
       "31232    1.208024    1.210205   1.214112     1.210781\n",
       "31233    1.210013    1.209210   1.214112     1.209388\n",
       "31234    1.208621    1.218165   1.212917     1.224119\n",
       "31235    1.221948    1.218762   1.214709     1.209985\n",
       "31236    1.202653    1.199858   1.203166     1.205406\n",
       "31237    1.205040    1.206026   1.209136     1.211976\n",
       "31238    1.209416    1.209608   1.214908     1.214564\n",
       "31239    1.213792    1.211399   1.215306     1.217151\n",
       "31240    1.217771    1.217170   1.223863     1.219341\n",
       "31241    1.216179    1.213190   1.217495     1.218943\n",
       "31242    1.220158    1.219757   1.225455     1.220735\n",
       "31243    1.218765    1.219359   1.220679     1.215957\n",
       "\n",
       "[31244 rows x 4 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#X_train_std as dataframe (optional, only for printing. See note in the beginning)\n",
    "X_Train_std = pd.DataFrame(data=X_train_std)\n",
    "X_Train_std.columns = X_Train.columns\n",
    "print(\"X_train_std\")\n",
    "X_Train_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_test_std\n"
     ]
    },
    {
     "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>EURUSDopen</th>\n",
       "      <th>EURUSDhigh</th>\n",
       "      <th>EURUSDlow</th>\n",
       "      <th>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.214190</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.220281</td>\n",
       "      <td>1.219142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.218367</td>\n",
       "      <td>1.218165</td>\n",
       "      <td>1.224460</td>\n",
       "      <td>1.222327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.220357</td>\n",
       "      <td>1.216374</td>\n",
       "      <td>1.224659</td>\n",
       "      <td>1.220735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.218566</td>\n",
       "      <td>1.214583</td>\n",
       "      <td>1.219286</td>\n",
       "      <td>1.215559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.217373</td>\n",
       "      <td>1.212792</td>\n",
       "      <td>1.219684</td>\n",
       "      <td>1.214962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.214787</td>\n",
       "      <td>1.210205</td>\n",
       "      <td>1.215107</td>\n",
       "      <td>1.211777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.210013</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.216102</td>\n",
       "      <td>1.217351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.216975</td>\n",
       "      <td>1.214782</td>\n",
       "      <td>1.221475</td>\n",
       "      <td>1.220735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.218964</td>\n",
       "      <td>1.217170</td>\n",
       "      <td>1.225057</td>\n",
       "      <td>1.222725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.221152</td>\n",
       "      <td>1.217369</td>\n",
       "      <td>1.226450</td>\n",
       "      <td>1.222128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.220953</td>\n",
       "      <td>1.216374</td>\n",
       "      <td>1.221475</td>\n",
       "      <td>1.217550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.216776</td>\n",
       "      <td>1.215180</td>\n",
       "      <td>1.221276</td>\n",
       "      <td>1.218346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.219362</td>\n",
       "      <td>1.214782</td>\n",
       "      <td>1.220878</td>\n",
       "      <td>1.217749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.216975</td>\n",
       "      <td>1.214185</td>\n",
       "      <td>1.219883</td>\n",
       "      <td>1.217351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.213593</td>\n",
       "      <td>1.212593</td>\n",
       "      <td>1.215306</td>\n",
       "      <td>1.213170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.212201</td>\n",
       "      <td>1.210404</td>\n",
       "      <td>1.217296</td>\n",
       "      <td>1.215957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.212201</td>\n",
       "      <td>1.213389</td>\n",
       "      <td>1.213116</td>\n",
       "      <td>1.212175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.211206</td>\n",
       "      <td>1.209210</td>\n",
       "      <td>1.216301</td>\n",
       "      <td>1.212772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.215980</td>\n",
       "      <td>1.211399</td>\n",
       "      <td>1.217495</td>\n",
       "      <td>1.212772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.214389</td>\n",
       "      <td>1.211598</td>\n",
       "      <td>1.217893</td>\n",
       "      <td>1.217151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.216179</td>\n",
       "      <td>1.214583</td>\n",
       "      <td>1.218092</td>\n",
       "      <td>1.220536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.219163</td>\n",
       "      <td>1.214583</td>\n",
       "      <td>1.218888</td>\n",
       "      <td>1.214763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1.212798</td>\n",
       "      <td>1.217369</td>\n",
       "      <td>1.218888</td>\n",
       "      <td>1.221730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.219362</td>\n",
       "      <td>1.215976</td>\n",
       "      <td>1.222669</td>\n",
       "      <td>1.218545</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1.216776</td>\n",
       "      <td>1.220752</td>\n",
       "      <td>1.222669</td>\n",
       "      <td>1.220735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.215185</td>\n",
       "      <td>1.210603</td>\n",
       "      <td>1.211922</td>\n",
       "      <td>1.211179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.210411</td>\n",
       "      <td>1.205827</td>\n",
       "      <td>1.203564</td>\n",
       "      <td>1.200828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1.199073</td>\n",
       "      <td>1.197470</td>\n",
       "      <td>1.202370</td>\n",
       "      <td>1.203416</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1.202454</td>\n",
       "      <td>1.197868</td>\n",
       "      <td>1.196997</td>\n",
       "      <td>1.193661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1.195890</td>\n",
       "      <td>1.206225</td>\n",
       "      <td>1.201972</td>\n",
       "      <td>1.212175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3442</th>\n",
       "      <td>1.491478</td>\n",
       "      <td>1.492773</td>\n",
       "      <td>1.497694</td>\n",
       "      <td>1.498041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3443</th>\n",
       "      <td>1.493865</td>\n",
       "      <td>1.499738</td>\n",
       "      <td>1.500082</td>\n",
       "      <td>1.501824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3444</th>\n",
       "      <td>1.500429</td>\n",
       "      <td>1.502922</td>\n",
       "      <td>1.506649</td>\n",
       "      <td>1.507000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3445</th>\n",
       "      <td>1.505203</td>\n",
       "      <td>1.501529</td>\n",
       "      <td>1.497296</td>\n",
       "      <td>1.498838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3446</th>\n",
       "      <td>1.496053</td>\n",
       "      <td>1.506703</td>\n",
       "      <td>1.502271</td>\n",
       "      <td>1.512773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3447</th>\n",
       "      <td>1.510375</td>\n",
       "      <td>1.524214</td>\n",
       "      <td>1.515804</td>\n",
       "      <td>1.530291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3448</th>\n",
       "      <td>1.534244</td>\n",
       "      <td>1.529786</td>\n",
       "      <td>1.525754</td>\n",
       "      <td>1.532879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3449</th>\n",
       "      <td>1.534841</td>\n",
       "      <td>1.535955</td>\n",
       "      <td>1.532719</td>\n",
       "      <td>1.530490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3450</th>\n",
       "      <td>1.529272</td>\n",
       "      <td>1.538542</td>\n",
       "      <td>1.535505</td>\n",
       "      <td>1.544624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3451</th>\n",
       "      <td>1.543196</td>\n",
       "      <td>1.553267</td>\n",
       "      <td>1.549435</td>\n",
       "      <td>1.559355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3452</th>\n",
       "      <td>1.558711</td>\n",
       "      <td>1.559635</td>\n",
       "      <td>1.559585</td>\n",
       "      <td>1.557962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3453</th>\n",
       "      <td>1.554136</td>\n",
       "      <td>1.557844</td>\n",
       "      <td>1.559983</td>\n",
       "      <td>1.562142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3454</th>\n",
       "      <td>1.563286</td>\n",
       "      <td>1.561227</td>\n",
       "      <td>1.561177</td>\n",
       "      <td>1.558360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3455</th>\n",
       "      <td>1.559109</td>\n",
       "      <td>1.557446</td>\n",
       "      <td>1.560182</td>\n",
       "      <td>1.555772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3456</th>\n",
       "      <td>1.557517</td>\n",
       "      <td>1.556451</td>\n",
       "      <td>1.560978</td>\n",
       "      <td>1.561346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3457</th>\n",
       "      <td>1.559706</td>\n",
       "      <td>1.556451</td>\n",
       "      <td>1.558789</td>\n",
       "      <td>1.554180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3458</th>\n",
       "      <td>1.550953</td>\n",
       "      <td>1.546501</td>\n",
       "      <td>1.551027</td>\n",
       "      <td>1.547411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3459</th>\n",
       "      <td>1.547572</td>\n",
       "      <td>1.543715</td>\n",
       "      <td>1.545853</td>\n",
       "      <td>1.543032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3460</th>\n",
       "      <td>1.541604</td>\n",
       "      <td>1.537945</td>\n",
       "      <td>1.539883</td>\n",
       "      <td>1.535268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3461</th>\n",
       "      <td>1.532255</td>\n",
       "      <td>1.530980</td>\n",
       "      <td>1.538490</td>\n",
       "      <td>1.536462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3462</th>\n",
       "      <td>1.534443</td>\n",
       "      <td>1.530184</td>\n",
       "      <td>1.536699</td>\n",
       "      <td>1.532680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3463</th>\n",
       "      <td>1.531261</td>\n",
       "      <td>1.533169</td>\n",
       "      <td>1.535306</td>\n",
       "      <td>1.537657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3464</th>\n",
       "      <td>1.533250</td>\n",
       "      <td>1.531179</td>\n",
       "      <td>1.535505</td>\n",
       "      <td>1.532680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3465</th>\n",
       "      <td>1.530863</td>\n",
       "      <td>1.526403</td>\n",
       "      <td>1.530928</td>\n",
       "      <td>1.527106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3466</th>\n",
       "      <td>1.525094</td>\n",
       "      <td>1.520831</td>\n",
       "      <td>1.524958</td>\n",
       "      <td>1.521333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3467</th>\n",
       "      <td>1.519923</td>\n",
       "      <td>1.516055</td>\n",
       "      <td>1.516799</td>\n",
       "      <td>1.513171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3468</th>\n",
       "      <td>1.510574</td>\n",
       "      <td>1.517448</td>\n",
       "      <td>1.516799</td>\n",
       "      <td>1.521333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3469</th>\n",
       "      <td>1.521514</td>\n",
       "      <td>1.517050</td>\n",
       "      <td>1.512818</td>\n",
       "      <td>1.509189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3470</th>\n",
       "      <td>1.511170</td>\n",
       "      <td>1.508096</td>\n",
       "      <td>1.513813</td>\n",
       "      <td>1.509787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3471</th>\n",
       "      <td>1.508187</td>\n",
       "      <td>1.512275</td>\n",
       "      <td>1.506649</td>\n",
       "      <td>1.512972</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3472 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      EURUSDopen  EURUSDhigh  EURUSDlow  EURUSDclose\n",
       "0       1.214190    1.213190   1.220281     1.219142\n",
       "1       1.218367    1.218165   1.224460     1.222327\n",
       "2       1.220357    1.216374   1.224659     1.220735\n",
       "3       1.218566    1.214583   1.219286     1.215559\n",
       "4       1.217373    1.212792   1.219684     1.214962\n",
       "5       1.214787    1.210205   1.215107     1.211777\n",
       "6       1.210013    1.213190   1.216102     1.217351\n",
       "7       1.216975    1.214782   1.221475     1.220735\n",
       "8       1.218964    1.217170   1.225057     1.222725\n",
       "9       1.221152    1.217369   1.226450     1.222128\n",
       "10      1.220953    1.216374   1.221475     1.217550\n",
       "11      1.216776    1.215180   1.221276     1.218346\n",
       "12      1.219362    1.214782   1.220878     1.217749\n",
       "13      1.216975    1.214185   1.219883     1.217351\n",
       "14      1.213593    1.212593   1.215306     1.213170\n",
       "15      1.212201    1.210404   1.217296     1.215957\n",
       "16      1.212201    1.213389   1.213116     1.212175\n",
       "17      1.211206    1.209210   1.216301     1.212772\n",
       "18      1.215980    1.211399   1.217495     1.212772\n",
       "19      1.214389    1.211598   1.217893     1.217151\n",
       "20      1.216179    1.214583   1.218092     1.220536\n",
       "21      1.219163    1.214583   1.218888     1.214763\n",
       "22      1.212798    1.217369   1.218888     1.221730\n",
       "23      1.219362    1.215976   1.222669     1.218545\n",
       "24      1.216776    1.220752   1.222669     1.220735\n",
       "25      1.215185    1.210603   1.211922     1.211179\n",
       "26      1.210411    1.205827   1.203564     1.200828\n",
       "27      1.199073    1.197470   1.202370     1.203416\n",
       "28      1.202454    1.197868   1.196997     1.193661\n",
       "29      1.195890    1.206225   1.201972     1.212175\n",
       "...          ...         ...        ...          ...\n",
       "3442    1.491478    1.492773   1.497694     1.498041\n",
       "3443    1.493865    1.499738   1.500082     1.501824\n",
       "3444    1.500429    1.502922   1.506649     1.507000\n",
       "3445    1.505203    1.501529   1.497296     1.498838\n",
       "3446    1.496053    1.506703   1.502271     1.512773\n",
       "3447    1.510375    1.524214   1.515804     1.530291\n",
       "3448    1.534244    1.529786   1.525754     1.532879\n",
       "3449    1.534841    1.535955   1.532719     1.530490\n",
       "3450    1.529272    1.538542   1.535505     1.544624\n",
       "3451    1.543196    1.553267   1.549435     1.559355\n",
       "3452    1.558711    1.559635   1.559585     1.557962\n",
       "3453    1.554136    1.557844   1.559983     1.562142\n",
       "3454    1.563286    1.561227   1.561177     1.558360\n",
       "3455    1.559109    1.557446   1.560182     1.555772\n",
       "3456    1.557517    1.556451   1.560978     1.561346\n",
       "3457    1.559706    1.556451   1.558789     1.554180\n",
       "3458    1.550953    1.546501   1.551027     1.547411\n",
       "3459    1.547572    1.543715   1.545853     1.543032\n",
       "3460    1.541604    1.537945   1.539883     1.535268\n",
       "3461    1.532255    1.530980   1.538490     1.536462\n",
       "3462    1.534443    1.530184   1.536699     1.532680\n",
       "3463    1.531261    1.533169   1.535306     1.537657\n",
       "3464    1.533250    1.531179   1.535505     1.532680\n",
       "3465    1.530863    1.526403   1.530928     1.527106\n",
       "3466    1.525094    1.520831   1.524958     1.521333\n",
       "3467    1.519923    1.516055   1.516799     1.513171\n",
       "3468    1.510574    1.517448   1.516799     1.521333\n",
       "3469    1.521514    1.517050   1.512818     1.509189\n",
       "3470    1.511170    1.508096   1.513813     1.509787\n",
       "3471    1.508187    1.512275   1.506649     1.512972\n",
       "\n",
       "[3472 rows x 4 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#X_train_std as dataframe (optional, only for printing. See note in the beginning)\n",
    "X_Test_std = pd.DataFrame(data=X_test_std)\n",
    "X_Test_std.columns = X_Test.columns\n",
    "\n",
    "print(\"X_test_std\")\n",
    "X_Test_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.05178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.05169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.05150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.05156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.05122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.05107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.04929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.04802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.04816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.04874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.04868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.04879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.04831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.04803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.04801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.04798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.04794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.04782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.04726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.04714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.04711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1.04659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.04649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1.04657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.04655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.04668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1.04677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1.04695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1.04733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31214</th>\n",
       "      <td>1.18404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31215</th>\n",
       "      <td>1.18478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31216</th>\n",
       "      <td>1.18447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31217</th>\n",
       "      <td>1.18502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31218</th>\n",
       "      <td>1.18463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31219</th>\n",
       "      <td>1.18460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31220</th>\n",
       "      <td>1.18471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31221</th>\n",
       "      <td>1.18496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31222</th>\n",
       "      <td>1.18429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31223</th>\n",
       "      <td>1.18474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31224</th>\n",
       "      <td>1.18449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31225</th>\n",
       "      <td>1.18430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31226</th>\n",
       "      <td>1.18489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31227</th>\n",
       "      <td>1.18477</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31228</th>\n",
       "      <td>1.18484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31229</th>\n",
       "      <td>1.18476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31230</th>\n",
       "      <td>1.18466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31231</th>\n",
       "      <td>1.18464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31232</th>\n",
       "      <td>1.18457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31233</th>\n",
       "      <td>1.18531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31234</th>\n",
       "      <td>1.18460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31235</th>\n",
       "      <td>1.18437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31236</th>\n",
       "      <td>1.18470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31237</th>\n",
       "      <td>1.18483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31238</th>\n",
       "      <td>1.18496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31239</th>\n",
       "      <td>1.18507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31240</th>\n",
       "      <td>1.18505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31241</th>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31242</th>\n",
       "      <td>1.18490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31243</th>\n",
       "      <td>1.18506</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31244 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       EURUSDclose\n",
       "0          1.05178\n",
       "1          1.05169\n",
       "2          1.05150\n",
       "3          1.05156\n",
       "4          1.05122\n",
       "5          1.05107\n",
       "6          1.04929\n",
       "7          1.04802\n",
       "8          1.04816\n",
       "9          1.04874\n",
       "10         1.04868\n",
       "11         1.04879\n",
       "12         1.04803\n",
       "13         1.04831\n",
       "14         1.04803\n",
       "15         1.04801\n",
       "16         1.04798\n",
       "17         1.04794\n",
       "18         1.04782\n",
       "19         1.04726\n",
       "20         1.04714\n",
       "21         1.04711\n",
       "22         1.04659\n",
       "23         1.04649\n",
       "24         1.04657\n",
       "25         1.04655\n",
       "26         1.04668\n",
       "27         1.04677\n",
       "28         1.04695\n",
       "29         1.04733\n",
       "...            ...\n",
       "31214      1.18404\n",
       "31215      1.18478\n",
       "31216      1.18447\n",
       "31217      1.18502\n",
       "31218      1.18463\n",
       "31219      1.18460\n",
       "31220      1.18471\n",
       "31221      1.18496\n",
       "31222      1.18429\n",
       "31223      1.18474\n",
       "31224      1.18449\n",
       "31225      1.18430\n",
       "31226      1.18489\n",
       "31227      1.18477\n",
       "31228      1.18484\n",
       "31229      1.18476\n",
       "31230      1.18466\n",
       "31231      1.18464\n",
       "31232      1.18457\n",
       "31233      1.18531\n",
       "31234      1.18460\n",
       "31235      1.18437\n",
       "31236      1.18470\n",
       "31237      1.18483\n",
       "31238      1.18496\n",
       "31239      1.18507\n",
       "31240      1.18505\n",
       "31241      1.18514\n",
       "31242      1.18490\n",
       "31243      1.18506\n",
       "\n",
       "[31244 rows x 1 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#y_train as panda dataframe (optional, only for printing. See note in the beginning)\n",
    "y_Train = pd.DataFrame(data=y_train)\n",
    "y_Train.columns=[\"EURUSDclose\"]\n",
    "y_Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.18522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.18488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.18485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.18469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.18497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.18524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.18521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.18498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.18502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.18499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.18497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.18476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.18490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.18471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.18474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.18474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.18496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.18513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.18484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.18519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1.18503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.18514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1.18466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.18414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.18427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1.18378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1.18471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1.18479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3442</th>\n",
       "      <td>1.19926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3443</th>\n",
       "      <td>1.19952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3444</th>\n",
       "      <td>1.19911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3445</th>\n",
       "      <td>1.19981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3446</th>\n",
       "      <td>1.20069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3447</th>\n",
       "      <td>1.20082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3448</th>\n",
       "      <td>1.20070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3449</th>\n",
       "      <td>1.20141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3450</th>\n",
       "      <td>1.20215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3451</th>\n",
       "      <td>1.20208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3452</th>\n",
       "      <td>1.20229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3453</th>\n",
       "      <td>1.20210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3454</th>\n",
       "      <td>1.20197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3455</th>\n",
       "      <td>1.20225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3456</th>\n",
       "      <td>1.20189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3457</th>\n",
       "      <td>1.20155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3458</th>\n",
       "      <td>1.20133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3459</th>\n",
       "      <td>1.20094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3460</th>\n",
       "      <td>1.20100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3461</th>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3462</th>\n",
       "      <td>1.20106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3463</th>\n",
       "      <td>1.20081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3464</th>\n",
       "      <td>1.20053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3465</th>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3466</th>\n",
       "      <td>1.19983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3467</th>\n",
       "      <td>1.20024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3468</th>\n",
       "      <td>1.19963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3469</th>\n",
       "      <td>1.19966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3470</th>\n",
       "      <td>1.19982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3471</th>\n",
       "      <td>1.19982</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3472 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      EURUSDclose\n",
       "0         1.18522\n",
       "1         1.18514\n",
       "2         1.18488\n",
       "3         1.18485\n",
       "4         1.18469\n",
       "5         1.18497\n",
       "6         1.18514\n",
       "7         1.18524\n",
       "8         1.18521\n",
       "9         1.18498\n",
       "10        1.18502\n",
       "11        1.18499\n",
       "12        1.18497\n",
       "13        1.18476\n",
       "14        1.18490\n",
       "15        1.18471\n",
       "16        1.18474\n",
       "17        1.18474\n",
       "18        1.18496\n",
       "19        1.18513\n",
       "20        1.18484\n",
       "21        1.18519\n",
       "22        1.18503\n",
       "23        1.18514\n",
       "24        1.18466\n",
       "25        1.18414\n",
       "26        1.18427\n",
       "27        1.18378\n",
       "28        1.18471\n",
       "29        1.18479\n",
       "...           ...\n",
       "3442      1.19926\n",
       "3443      1.19952\n",
       "3444      1.19911\n",
       "3445      1.19981\n",
       "3446      1.20069\n",
       "3447      1.20082\n",
       "3448      1.20070\n",
       "3449      1.20141\n",
       "3450      1.20215\n",
       "3451      1.20208\n",
       "3452      1.20229\n",
       "3453      1.20210\n",
       "3454      1.20197\n",
       "3455      1.20225\n",
       "3456      1.20189\n",
       "3457      1.20155\n",
       "3458      1.20133\n",
       "3459      1.20094\n",
       "3460      1.20100\n",
       "3461      1.20081\n",
       "3462      1.20106\n",
       "3463      1.20081\n",
       "3464      1.20053\n",
       "3465      1.20024\n",
       "3466      1.19983\n",
       "3467      1.20024\n",
       "3468      1.19963\n",
       "3469      1.19966\n",
       "3470      1.19982\n",
       "3471      1.19982\n",
       "\n",
       "[3472 rows x 1 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#y_train as panda dataframe (optional, only for printing. See note in the beginning)\n",
    "y_Test = pd.DataFrame(data=y_test)\n",
    "y_Test.columns=[\"EURUSDclose\"]\n",
    "y_Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.434180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.435972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.439754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.438560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.445328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-1.448315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>-1.483750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-1.509032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-1.506245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-1.494699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>-1.495893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>-1.493703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>-1.508833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-1.503259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-1.508833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-1.509231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>-1.509828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-1.510625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-1.513013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>-1.524161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>-1.526550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>-1.527148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>-1.537499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>-1.539490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>-1.537898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>-1.538296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>-1.535708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>-1.533916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>-1.530333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>-1.522768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31214</th>\n",
       "      <td>1.198763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31215</th>\n",
       "      <td>1.213494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31216</th>\n",
       "      <td>1.207323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31217</th>\n",
       "      <td>1.218272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31218</th>\n",
       "      <td>1.210508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31219</th>\n",
       "      <td>1.209911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31220</th>\n",
       "      <td>1.212101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31221</th>\n",
       "      <td>1.217078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31222</th>\n",
       "      <td>1.203740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31223</th>\n",
       "      <td>1.212698</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31224</th>\n",
       "      <td>1.207721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31225</th>\n",
       "      <td>1.203939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31226</th>\n",
       "      <td>1.215684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31227</th>\n",
       "      <td>1.213295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31228</th>\n",
       "      <td>1.214689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31229</th>\n",
       "      <td>1.213096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31230</th>\n",
       "      <td>1.211105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31231</th>\n",
       "      <td>1.210707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31232</th>\n",
       "      <td>1.209314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31233</th>\n",
       "      <td>1.224045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31234</th>\n",
       "      <td>1.209911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31235</th>\n",
       "      <td>1.205332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31236</th>\n",
       "      <td>1.211902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31237</th>\n",
       "      <td>1.214490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31238</th>\n",
       "      <td>1.217078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31239</th>\n",
       "      <td>1.219267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31240</th>\n",
       "      <td>1.218869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31241</th>\n",
       "      <td>1.220661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31242</th>\n",
       "      <td>1.215883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31243</th>\n",
       "      <td>1.219068</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31244 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       EURUSDclose\n",
       "0        -1.434180\n",
       "1        -1.435972\n",
       "2        -1.439754\n",
       "3        -1.438560\n",
       "4        -1.445328\n",
       "5        -1.448315\n",
       "6        -1.483750\n",
       "7        -1.509032\n",
       "8        -1.506245\n",
       "9        -1.494699\n",
       "10       -1.495893\n",
       "11       -1.493703\n",
       "12       -1.508833\n",
       "13       -1.503259\n",
       "14       -1.508833\n",
       "15       -1.509231\n",
       "16       -1.509828\n",
       "17       -1.510625\n",
       "18       -1.513013\n",
       "19       -1.524161\n",
       "20       -1.526550\n",
       "21       -1.527148\n",
       "22       -1.537499\n",
       "23       -1.539490\n",
       "24       -1.537898\n",
       "25       -1.538296\n",
       "26       -1.535708\n",
       "27       -1.533916\n",
       "28       -1.530333\n",
       "29       -1.522768\n",
       "...            ...\n",
       "31214     1.198763\n",
       "31215     1.213494\n",
       "31216     1.207323\n",
       "31217     1.218272\n",
       "31218     1.210508\n",
       "31219     1.209911\n",
       "31220     1.212101\n",
       "31221     1.217078\n",
       "31222     1.203740\n",
       "31223     1.212698\n",
       "31224     1.207721\n",
       "31225     1.203939\n",
       "31226     1.215684\n",
       "31227     1.213295\n",
       "31228     1.214689\n",
       "31229     1.213096\n",
       "31230     1.211105\n",
       "31231     1.210707\n",
       "31232     1.209314\n",
       "31233     1.224045\n",
       "31234     1.209911\n",
       "31235     1.205332\n",
       "31236     1.211902\n",
       "31237     1.214490\n",
       "31238     1.217078\n",
       "31239     1.219267\n",
       "31240     1.218869\n",
       "31241     1.220661\n",
       "31242     1.215883\n",
       "31243     1.219068\n",
       "\n",
       "[31244 rows x 1 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#y_train_std as panda dataframe (optional, only for printing. See note in the beginning)\n",
    "y_Train_std = pd.DataFrame(data=y_train_std)\n",
    "y_Train_std.columns=[\"EURUSDclose\"]\n",
    "y_Train_std"
   ]
  },
  {
   "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>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.434180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.435972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.439754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.438560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.445328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-1.448315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>-1.483750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-1.509032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-1.506245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-1.494699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>-1.495893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>-1.493703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>-1.508833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-1.503259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-1.508833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-1.509231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>-1.509828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-1.510625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-1.513013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>-1.524161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>-1.526550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>-1.527148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>-1.537499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>-1.539490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>-1.537898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>-1.538296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>-1.535708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>-1.533916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>-1.530333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>-1.522768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31214</th>\n",
       "      <td>1.198763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31215</th>\n",
       "      <td>1.213494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31216</th>\n",
       "      <td>1.207323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31217</th>\n",
       "      <td>1.218272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31218</th>\n",
       "      <td>1.210508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31219</th>\n",
       "      <td>1.209911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31220</th>\n",
       "      <td>1.212101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31221</th>\n",
       "      <td>1.217078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31222</th>\n",
       "      <td>1.203740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31223</th>\n",
       "      <td>1.212698</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31224</th>\n",
       "      <td>1.207721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31225</th>\n",
       "      <td>1.203939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31226</th>\n",
       "      <td>1.215684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31227</th>\n",
       "      <td>1.213295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31228</th>\n",
       "      <td>1.214689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31229</th>\n",
       "      <td>1.213096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31230</th>\n",
       "      <td>1.211105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31231</th>\n",
       "      <td>1.210707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31232</th>\n",
       "      <td>1.209314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31233</th>\n",
       "      <td>1.224045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31234</th>\n",
       "      <td>1.209911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31235</th>\n",
       "      <td>1.205332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31236</th>\n",
       "      <td>1.211902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31237</th>\n",
       "      <td>1.214490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31238</th>\n",
       "      <td>1.217078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31239</th>\n",
       "      <td>1.219267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31240</th>\n",
       "      <td>1.218869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31241</th>\n",
       "      <td>1.220661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31242</th>\n",
       "      <td>1.215883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31243</th>\n",
       "      <td>1.219068</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31244 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       EURUSDclose\n",
       "0        -1.434180\n",
       "1        -1.435972\n",
       "2        -1.439754\n",
       "3        -1.438560\n",
       "4        -1.445328\n",
       "5        -1.448315\n",
       "6        -1.483750\n",
       "7        -1.509032\n",
       "8        -1.506245\n",
       "9        -1.494699\n",
       "10       -1.495893\n",
       "11       -1.493703\n",
       "12       -1.508833\n",
       "13       -1.503259\n",
       "14       -1.508833\n",
       "15       -1.509231\n",
       "16       -1.509828\n",
       "17       -1.510625\n",
       "18       -1.513013\n",
       "19       -1.524161\n",
       "20       -1.526550\n",
       "21       -1.527148\n",
       "22       -1.537499\n",
       "23       -1.539490\n",
       "24       -1.537898\n",
       "25       -1.538296\n",
       "26       -1.535708\n",
       "27       -1.533916\n",
       "28       -1.530333\n",
       "29       -1.522768\n",
       "...            ...\n",
       "31214     1.198763\n",
       "31215     1.213494\n",
       "31216     1.207323\n",
       "31217     1.218272\n",
       "31218     1.210508\n",
       "31219     1.209911\n",
       "31220     1.212101\n",
       "31221     1.217078\n",
       "31222     1.203740\n",
       "31223     1.212698\n",
       "31224     1.207721\n",
       "31225     1.203939\n",
       "31226     1.215684\n",
       "31227     1.213295\n",
       "31228     1.214689\n",
       "31229     1.213096\n",
       "31230     1.211105\n",
       "31231     1.210707\n",
       "31232     1.209314\n",
       "31233     1.224045\n",
       "31234     1.209911\n",
       "31235     1.205332\n",
       "31236     1.211902\n",
       "31237     1.214490\n",
       "31238     1.217078\n",
       "31239     1.219267\n",
       "31240     1.218869\n",
       "31241     1.220661\n",
       "31242     1.215883\n",
       "31243     1.219068\n",
       "\n",
       "[31244 rows x 1 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#y_train_std as panda dataframe (optional, only for printing. See note in the beginning)\n",
    "y_Test_std = pd.DataFrame(data=y_train_std)\n",
    "y_Test_std.columns=[\"EURUSDclose\"]\n",
    "y_Test_std"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Implement the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Clears the default graph stack and resets the global default graph\n",
    "ops.reset_default_graph()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Placeholders\n",
      "Initializers\n",
      "Cost function\n",
      "Optimizer\n"
     ]
    }
   ],
   "source": [
    "# make results reproducible\n",
    "seed = 2\n",
    "tf.set_random_seed(seed)\n",
    "np.random.seed(seed)  \n",
    "\n",
    "\n",
    "# Parameters\n",
    "learning_rate = 0.005\n",
    "batch_size = 256\n",
    "n_features = X_train.shape[1]#  Number of features in training data\n",
    "epochs = 1000\n",
    "display_step = 100\n",
    "model_path = model_dir+\"07_First_Forex_Prediction\"\n",
    "production_model_path = production_dir+\"models/\"+\"07_First_Forex_Prediction\"\n",
    "n_classes = 1\n",
    "\n",
    "# Network Parameters\n",
    "# See figure of the model\n",
    "d0 = D = n_features # Layer 0 (Input layer number of features)\n",
    "d1 = 1024 # Layer 1 (1024 hidden nodes)\n",
    "d2 = 512 # Layer 2 (512 hidden nodes) \n",
    "d3 = 256 # Layer 3 (256 hidden nodes)\n",
    "d4 = 128 # Layer 4 (128 hidden nodes)\n",
    "d5 = C = 1 # Layer 5 (Output layer)\n",
    "\n",
    "# tf Graph input\n",
    "print(\"Placeholders\")\n",
    "X = tf.placeholder(dtype=tf.float32, shape=[None, n_features], name=\"X\")\n",
    "y = tf.placeholder(dtype=tf.float32, shape=[None,n_classes], name=\"y\")\n",
    "\n",
    "\n",
    "# Initializers\n",
    "print(\"Initializers\")\n",
    "sigma = 1\n",
    "weight_initializer = tf.variance_scaling_initializer(mode=\"fan_avg\", distribution=\"uniform\", scale=sigma)\n",
    "bias_initializer = tf.zeros_initializer()\n",
    "\n",
    "# Create model\n",
    "def multilayer_perceptron(X, variables):\n",
    "    # Hidden layer with ReLU activation\n",
    "    layer_1 = tf.nn.relu(tf.add(tf.matmul(X, variables['W1']), variables['bias1']), name=\"layer_1\")\n",
    "    # Hidden layer with ReLU activation\n",
    "    layer_2 = tf.nn.relu(tf.add(tf.matmul(layer_1, variables['W2']), variables['bias2']), name=\"layer_2\")\n",
    "    # Hidden layer with ReLU activation\n",
    "    layer_3 = tf.nn.relu(tf.add(tf.matmul(layer_2, variables['W3']), variables['bias3']), name=\"layer_3\")\n",
    "    # Hidden layer with ReLU activation\n",
    "    layer_4 = tf.nn.relu(tf.add(tf.matmul(layer_3, variables['W4']), variables['bias4']), name=\"layer_4\")\n",
    "    # Output layer with ReLU activation\n",
    "    out_layer = tf.nn.relu(tf.add(tf.matmul(layer_4, variables['W5']), variables['bias5']), name=\"out_layer\")\n",
    "    return out_layer\n",
    "\n",
    "# Store layers weight & bias\n",
    "variables = {\n",
    "    'W1': tf.Variable(weight_initializer([n_features, d1]), name=\"W1\"), # inputs -> d1 hidden neurons\n",
    "    'bias1': tf.Variable(bias_initializer([d1]), name=\"bias1\"), # one biases for each d1 hidden neurons\n",
    "    'W2': tf.Variable(weight_initializer([d1, d2]), name=\"W2\"), # d1 hidden inputs -> d2 hidden neurons\n",
    "    'bias2': tf.Variable(bias_initializer([d2]), name=\"bias2\"), # one biases for each d2 hidden neurons\n",
    "    'W3': tf.Variable(weight_initializer([d2, d3]), name=\"W3\"), ## d2 hidden inputs -> d3 hidden neurons\n",
    "    'bias3': tf.Variable(bias_initializer([d3]), name=\"bias3\"), # one biases for each d3 hidden neurons\n",
    "    'W4': tf.Variable(weight_initializer([d3, d4]), name=\"W4\"), ## d3 hidden inputs -> d4 hidden neurons\n",
    "    'bias4': tf.Variable(bias_initializer([d4]), name=\"bias4\"), # one biases for each d4 hidden neurons\n",
    "    'W5': tf.Variable(weight_initializer([d4, d5]), name=\"W5\"), # d4 hidden inputs -> 1 output\n",
    "    'bias5': tf.Variable(bias_initializer([d5]), name=\"bias5\") # 1 bias for the output\n",
    "}\n",
    "\n",
    "# Construct model\n",
    "y_hat = multilayer_perceptron(X, variables)\n",
    "\n",
    "# Cost function\n",
    "print(\"Cost function\")\n",
    "mse = tf.reduce_mean(tf.squared_difference(y_hat, y))\n",
    "\n",
    "# Optimizer\n",
    "print(\"Optimizer\")\n",
    "optimizer = tf.train.AdamOptimizer().minimize(mse)\n",
    "\n",
    "# Initialize the variables (i.e. assign their default value)\n",
    "init = tf.global_variables_initializer()\n",
    "\n",
    "# 'Saver' op to save and restore all the variables\n",
    "saver = tf.train.Saver()\n",
    "\n",
    "def generate_one_epoch(X_train, y_train, batch_size):\n",
    "    num_batches = int(len(X_train)) // batch_size\n",
    "    if batch_size * num_batches < len(X_train):\n",
    "        num_batches += 1\n",
    "\n",
    "    batch_indices = range(num_batches)\n",
    "    batch_n = np.random.permutation(batch_indices)\n",
    "    for j in batch_n:\n",
    "        batch_X = X_train[j * batch_size: (j + 1) * batch_size]\n",
    "        batch_y = y_train[j * batch_size: (j + 1) * batch_size]\n",
    "\n",
    "    yield batch_X, batch_y\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the model  and Evaluate the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train_std.shape =  (31244, 4) y_train.shape = (31244,) \n",
      "X_train_std.shape =   (31244, 4) y_test.shape =  (3472,)\n"
     ]
    }
   ],
   "source": [
    "# Shape of tensors\n",
    "print(\"X_train_std.shape = \", X_train_std.shape, \"y_train.shape =\", y_train.shape,\n",
    "      \"\\nX_train_std.shape =  \", X_train_std.shape, \"y_test.shape = \", y_test.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fit neural net\n",
      "Run\n",
      "Epoch:  100\n",
      "MSE Train:  0.04068575\n",
      "MSE Test:  0.0024781413\n",
      "Epoch:  200\n",
      "MSE Train:  0.009442378\n",
      "MSE Test:  0.010252378\n",
      "Epoch:  300\n",
      "MSE Train:  0.0006010969\n",
      "MSE Test:  0.0002887453\n",
      "Epoch:  400\n",
      "MSE Train:  0.00069611386\n",
      "MSE Test:  0.0014286637\n",
      "Epoch:  500\n",
      "MSE Train:  6.761727e-05\n",
      "MSE Test:  6.0221642e-05\n",
      "Epoch:  600\n",
      "MSE Train:  9.946661e-05\n",
      "MSE Test:  0.00022536414\n",
      "Epoch:  700\n",
      "MSE Train:  9.6010235e-06\n",
      "MSE Test:  3.9607035e-06\n",
      "Epoch:  800\n",
      "MSE Train:  1.2783626e-05\n",
      "MSE Test:  4.552653e-06\n",
      "Epoch:  900\n",
      "MSE Train:  0.00059785775\n",
      "MSE Test:  0.00030184354\n",
      "0.00027231404\n",
      "Model saved in file: ../models/07_First_Forex_Prediction\n",
      "First Optimization Finished!\n"
     ]
    }
   ],
   "source": [
    "# Fit neural net\n",
    "print(\"Fit neural net\")\n",
    "\n",
    "with tf.Session() as sess:\n",
    "\n",
    "    # Writer to record image, scalar, histogram and graph for display in tensorboard\n",
    "    writer = tf.summary.FileWriter(\"/tmp/tensorflow_logs\", sess.graph)  # create writer\n",
    "    writer.add_graph(sess.graph)\n",
    "\n",
    "    # Run the initializer\n",
    "    sess.run(init)\n",
    "    \n",
    "    # Restore model weights from previously saved model\n",
    "    #Ram saver.restore(sess, model_path)\n",
    "    #Ram print(\"Model restored from file: %s\" % model_path)\n",
    "    '''\n",
    "    # Try to restore a model if any.\n",
    "    try:\n",
    "        saver.restore(sess, model_path)\n",
    "        print(\"Model restored from file: %s\" % model_path)\n",
    "    except Exception:\n",
    "        print(\"No model file to restore\")\n",
    "        pass\n",
    "    '''\n",
    "    # Training cycle\n",
    "    mse_train = []\n",
    "    mse_test = []\n",
    "\n",
    "    \n",
    "    # Run\n",
    "    print(\"Run\")\n",
    "    printcounter = 0\n",
    "    for e in range(epochs):\n",
    "        # Minibatch training\n",
    "        \n",
    "        for batch_X, batch_y in generate_one_epoch(X_train_std, y_train, batch_size):\n",
    "            # Run optimizer with batch\n",
    "            sess.run(optimizer, feed_dict={X: batch_X, y: np.transpose([batch_y])})\n",
    "            #Ram print(\"batch_X\",batch_X)\n",
    "            #Ram print(\"batch_y\",batch_y)\n",
    "         \n",
    "        # Show progress\n",
    "        if (printcounter == display_step):\n",
    "            \n",
    "            printcounter = 0\n",
    "            print(\"Epoch: \", e)\n",
    "            # MSE train and test\n",
    "            mse_train.append(sess.run(mse, feed_dict={X: X_train_std, y: np.transpose([y_train])}))\n",
    "            mse_test.append(sess.run(mse, feed_dict={X: X_test_std, y: np.transpose([y_test])}))\n",
    "            print('MSE Train: ', mse_train[-1])\n",
    "            print('MSE Test: ', mse_test[-1])\n",
    "        \n",
    "        printcounter += 1\n",
    "        \n",
    "    # Print final MSE after Training\n",
    "    mse_final = sess.run(mse, feed_dict={X: X_test, y: np.transpose([y_test])})\n",
    "    print(mse_final)\n",
    "    \n",
    "    # Close writer\n",
    "    writer.flush()\n",
    "    writer.close()\n",
    "        \n",
    "    # Save model weights to disk\n",
    "    save_path = saver.save(sess, model_path)\n",
    "    print(\"Model saved in file: %s\" % save_path)\n",
    "    print(\"First Optimization Finished!\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(256,)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch_y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "# Plot loss (MSE) over time\n",
    "plt.plot(mse_train, 'k-', label='mse_train')\n",
    "plt.plot(mse_test, 'r--', label='mse_test')\n",
    "plt.title('Loss (MSE) vs epoch')\n",
    "plt.legend(loc='upper right')\n",
    "plt.xlabel('Generation x'+str(display_step))\n",
    "plt.ylabel('Loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tensorboard Graph\n",
    "\n",
    "\n",
    "What follows is the graph we have executed and all data about it. Note the \"save\" label and the several layers.\n",
    "\n",
    "\n",
    "![graph_4](../images/graph_4.png)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Saving a Tensorflow model\n",
    "\n",
    "So, now we have our model saved.\n",
    "\n",
    "Tensorflow model has four main files:\n",
    "* a) Meta graph:\n",
    "This is a protocol buffer which saves the complete Tensorflow graph; i.e. all variables, operations, collections etc. This file has .meta extension.\n",
    "\n",
    "\n",
    "* b) y c) Checkpoint files:\n",
    "It is a binary file which contains all the values of the weights, biases, gradients and all the other variables saved. Tensorflow has changed from version 0.11. Instead of a single .ckpt file, we have now two files: .index and .data file that contains our training variables. \n",
    "\n",
    "\n",
    "* d) Along with this, Tensorflow also has a file named checkpoint which simply keeps a record of latest checkpoint files saved.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we can use the model to make some predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting prediction session...\n",
      "INFO:tensorflow:Restoring parameters from ../models/07_First_Forex_Prediction\n",
      "Model restored from file: ../models/07_First_Forex_Prediction\n",
      "[[1.201417 ]\n",
      " [1.2017072]\n",
      " [1.2017524]\n",
      " ...\n",
      " [1.2206069]\n",
      " [1.2204254]\n",
      " [1.2196295]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:No assets to save.\n",
      "INFO:tensorflow:No assets to write.\n",
      "INFO:tensorflow:SavedModel written to: b'/home/parrondo/PRODUCTION/models/07_First_Forex_Prediction/45/saved_model.pb'\n"
     ]
    }
   ],
   "source": [
    "# Running a new session for predictions and export model to production\n",
    "print(\"Starting prediction session...\")\n",
    "with tf.Session() as sess:\n",
    "    # Initialize variables\n",
    "    sess.run(init)\n",
    "\n",
    "    # Try to restore a model if any.\n",
    "    try:\n",
    "        saver.restore(sess, model_path)\n",
    "        print(\"Model restored from file: %s\" % model_path)\n",
    "        # We try to predict the close price of test samples\n",
    "        feed_dict = {X: X_test_std}\n",
    "        \n",
    "        prediction = sess.run(y_hat, feed_dict)\n",
    "        print(prediction) \n",
    "\n",
    "        %matplotlib inline\n",
    "        # Plot Prices over time\n",
    "        plt.plot(y_test, 'k-', label='y_test')\n",
    "        plt.plot(prediction, 'r--', label='prediction')\n",
    "        plt.title('Price over time')\n",
    "        plt.legend(loc='upper right')\n",
    "        plt.xlabel('Time')\n",
    "        plt.ylabel('Price')\n",
    "        plt.show()\n",
    "        \n",
    "        if to_production:\n",
    "            # Pick out the model input and output\n",
    "            X_tensor = sess.graph.get_tensor_by_name(\"X\"+ ':0')\n",
    "            y_tensor = sess.graph.get_tensor_by_name(\"out_layer\" + ':0')\n",
    "\n",
    "            model_input = build_tensor_info(X_tensor)\n",
    "            model_output = build_tensor_info(y_tensor)\n",
    "            \n",
    "            # Create a signature definition for tfserving\n",
    "            signature_definition = signature_def_utils.build_signature_def(\n",
    "                inputs={\"X\": model_input},\n",
    "                outputs={\"out_layer\": model_output},\n",
    "                method_name=signature_constants.PREDICT_METHOD_NAME)\n",
    "\n",
    "            model_version = 1\n",
    "            export_model_dir = production_model_path+\"/\"+str(model_version)\n",
    "            while os.path.exists(export_model_dir):\n",
    "                model_version += 1\n",
    "                export_model_dir = production_model_path+\"/\"+str(model_version)       \n",
    "    \n",
    "            builder = saved_model_builder.SavedModelBuilder(export_model_dir)  \n",
    "            builder.add_meta_graph_and_variables(sess,\n",
    "                                                 [tag_constants.SERVING],\n",
    "                                                 signature_def_map={\n",
    "                                                     signature_constants.DEFAULT_SERVING_SIGNATURE_DEF_KEY:\n",
    "                                                     signature_definition})\n",
    "\n",
    "            # Save the model so we can serve it with a model server :)\n",
    "            builder.save()\n",
    "\n",
    "    except Exception:\n",
    "        print(\"Unexpected error:\", sys.exc_info()[0])\n",
    "        pass\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>y_test</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Abs.error</th>\n",
       "      <th>Relat.error</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.18522</td>\n",
       "      <td>1.201417</td>\n",
       "      <td>-0.016197</td>\n",
       "      <td>1.366579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.201707</td>\n",
       "      <td>-0.016567</td>\n",
       "      <td>1.397915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.18488</td>\n",
       "      <td>1.201752</td>\n",
       "      <td>-0.016872</td>\n",
       "      <td>1.423977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.18485</td>\n",
       "      <td>1.201519</td>\n",
       "      <td>-0.016669</td>\n",
       "      <td>1.406886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.18469</td>\n",
       "      <td>1.201575</td>\n",
       "      <td>-0.016885</td>\n",
       "      <td>1.425301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.18497</td>\n",
       "      <td>1.201281</td>\n",
       "      <td>-0.016311</td>\n",
       "      <td>1.376456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.201171</td>\n",
       "      <td>-0.016031</td>\n",
       "      <td>1.352630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.18524</td>\n",
       "      <td>1.201427</td>\n",
       "      <td>-0.016187</td>\n",
       "      <td>1.365744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.18521</td>\n",
       "      <td>1.201704</td>\n",
       "      <td>-0.016494</td>\n",
       "      <td>1.391664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.18498</td>\n",
       "      <td>1.201861</td>\n",
       "      <td>-0.016881</td>\n",
       "      <td>1.424573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.18502</td>\n",
       "      <td>1.201622</td>\n",
       "      <td>-0.016602</td>\n",
       "      <td>1.401000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.18499</td>\n",
       "      <td>1.201588</td>\n",
       "      <td>-0.016598</td>\n",
       "      <td>1.400720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.18497</td>\n",
       "      <td>1.201528</td>\n",
       "      <td>-0.016558</td>\n",
       "      <td>1.397341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.18476</td>\n",
       "      <td>1.201476</td>\n",
       "      <td>-0.016716</td>\n",
       "      <td>1.410947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.18490</td>\n",
       "      <td>1.201295</td>\n",
       "      <td>-0.016395</td>\n",
       "      <td>1.383632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.18471</td>\n",
       "      <td>1.201277</td>\n",
       "      <td>-0.016567</td>\n",
       "      <td>1.398383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.18474</td>\n",
       "      <td>1.201177</td>\n",
       "      <td>-0.016437</td>\n",
       "      <td>1.387433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.18474</td>\n",
       "      <td>1.201377</td>\n",
       "      <td>-0.016637</td>\n",
       "      <td>1.404307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.18496</td>\n",
       "      <td>1.201480</td>\n",
       "      <td>-0.016520</td>\n",
       "      <td>1.394112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.18513</td>\n",
       "      <td>1.201250</td>\n",
       "      <td>-0.016120</td>\n",
       "      <td>1.360225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.18484</td>\n",
       "      <td>1.201086</td>\n",
       "      <td>-0.016246</td>\n",
       "      <td>1.371190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.18519</td>\n",
       "      <td>1.201521</td>\n",
       "      <td>-0.016331</td>\n",
       "      <td>1.377956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1.18503</td>\n",
       "      <td>1.201230</td>\n",
       "      <td>-0.016200</td>\n",
       "      <td>1.367088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.18514</td>\n",
       "      <td>1.201699</td>\n",
       "      <td>-0.016559</td>\n",
       "      <td>1.397220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1.18466</td>\n",
       "      <td>1.201730</td>\n",
       "      <td>-0.017070</td>\n",
       "      <td>1.440881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.18414</td>\n",
       "      <td>1.200984</td>\n",
       "      <td>-0.016844</td>\n",
       "      <td>1.422467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.18427</td>\n",
       "      <td>1.200763</td>\n",
       "      <td>-0.016493</td>\n",
       "      <td>1.392711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1.18378</td>\n",
       "      <td>1.200426</td>\n",
       "      <td>-0.016646</td>\n",
       "      <td>1.406172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1.18471</td>\n",
       "      <td>1.200484</td>\n",
       "      <td>-0.015774</td>\n",
       "      <td>1.331478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1.18479</td>\n",
       "      <td>1.200080</td>\n",
       "      <td>-0.015290</td>\n",
       "      <td>1.290537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3442</th>\n",
       "      <td>1.19926</td>\n",
       "      <td>1.218956</td>\n",
       "      <td>-0.019696</td>\n",
       "      <td>1.642385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3443</th>\n",
       "      <td>1.19952</td>\n",
       "      <td>1.219204</td>\n",
       "      <td>-0.019684</td>\n",
       "      <td>1.640985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3444</th>\n",
       "      <td>1.19911</td>\n",
       "      <td>1.219689</td>\n",
       "      <td>-0.020579</td>\n",
       "      <td>1.716220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3445</th>\n",
       "      <td>1.19981</td>\n",
       "      <td>1.219091</td>\n",
       "      <td>-0.019281</td>\n",
       "      <td>1.606979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3446</th>\n",
       "      <td>1.20069</td>\n",
       "      <td>1.219039</td>\n",
       "      <td>-0.018349</td>\n",
       "      <td>1.528182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3447</th>\n",
       "      <td>1.20082</td>\n",
       "      <td>1.220027</td>\n",
       "      <td>-0.019207</td>\n",
       "      <td>1.599508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3448</th>\n",
       "      <td>1.20070</td>\n",
       "      <td>1.221001</td>\n",
       "      <td>-0.020301</td>\n",
       "      <td>1.690736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3449</th>\n",
       "      <td>1.20141</td>\n",
       "      <td>1.222078</td>\n",
       "      <td>-0.020668</td>\n",
       "      <td>1.720279</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3450</th>\n",
       "      <td>1.20215</td>\n",
       "      <td>1.221685</td>\n",
       "      <td>-0.019535</td>\n",
       "      <td>1.625039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3451</th>\n",
       "      <td>1.20208</td>\n",
       "      <td>1.222760</td>\n",
       "      <td>-0.020680</td>\n",
       "      <td>1.720328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3452</th>\n",
       "      <td>1.20229</td>\n",
       "      <td>1.224053</td>\n",
       "      <td>-0.021763</td>\n",
       "      <td>1.810121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3453</th>\n",
       "      <td>1.20210</td>\n",
       "      <td>1.223820</td>\n",
       "      <td>-0.021720</td>\n",
       "      <td>1.806816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3454</th>\n",
       "      <td>1.20197</td>\n",
       "      <td>1.224236</td>\n",
       "      <td>-0.022266</td>\n",
       "      <td>1.852500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3455</th>\n",
       "      <td>1.20225</td>\n",
       "      <td>1.224174</td>\n",
       "      <td>-0.021924</td>\n",
       "      <td>1.823622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3456</th>\n",
       "      <td>1.20189</td>\n",
       "      <td>1.223920</td>\n",
       "      <td>-0.022030</td>\n",
       "      <td>1.832955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3457</th>\n",
       "      <td>1.20155</td>\n",
       "      <td>1.224083</td>\n",
       "      <td>-0.022533</td>\n",
       "      <td>1.875343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3458</th>\n",
       "      <td>1.20133</td>\n",
       "      <td>1.223351</td>\n",
       "      <td>-0.022021</td>\n",
       "      <td>1.833032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3459</th>\n",
       "      <td>1.20094</td>\n",
       "      <td>1.222963</td>\n",
       "      <td>-0.022023</td>\n",
       "      <td>1.833811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3460</th>\n",
       "      <td>1.20100</td>\n",
       "      <td>1.222607</td>\n",
       "      <td>-0.021607</td>\n",
       "      <td>1.799076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3461</th>\n",
       "      <td>1.20081</td>\n",
       "      <td>1.222203</td>\n",
       "      <td>-0.021393</td>\n",
       "      <td>1.781509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3462</th>\n",
       "      <td>1.20106</td>\n",
       "      <td>1.222192</td>\n",
       "      <td>-0.021132</td>\n",
       "      <td>1.759430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3463</th>\n",
       "      <td>1.20081</td>\n",
       "      <td>1.221870</td>\n",
       "      <td>-0.021060</td>\n",
       "      <td>1.753831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3464</th>\n",
       "      <td>1.20053</td>\n",
       "      <td>1.222101</td>\n",
       "      <td>-0.021571</td>\n",
       "      <td>1.796827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3465</th>\n",
       "      <td>1.20024</td>\n",
       "      <td>1.221787</td>\n",
       "      <td>-0.021547</td>\n",
       "      <td>1.795242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3466</th>\n",
       "      <td>1.19983</td>\n",
       "      <td>1.221326</td>\n",
       "      <td>-0.021496</td>\n",
       "      <td>1.791547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3467</th>\n",
       "      <td>1.20024</td>\n",
       "      <td>1.220780</td>\n",
       "      <td>-0.020540</td>\n",
       "      <td>1.711336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3468</th>\n",
       "      <td>1.19963</td>\n",
       "      <td>1.220399</td>\n",
       "      <td>-0.020769</td>\n",
       "      <td>1.731276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3469</th>\n",
       "      <td>1.19966</td>\n",
       "      <td>1.220607</td>\n",
       "      <td>-0.020947</td>\n",
       "      <td>1.746072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3470</th>\n",
       "      <td>1.19982</td>\n",
       "      <td>1.220425</td>\n",
       "      <td>-0.020605</td>\n",
       "      <td>1.717372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3471</th>\n",
       "      <td>1.19982</td>\n",
       "      <td>1.219630</td>\n",
       "      <td>-0.019810</td>\n",
       "      <td>1.651042</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3472 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       y_test  prediction  Abs.error  Relat.error\n",
       "0     1.18522    1.201417  -0.016197     1.366579\n",
       "1     1.18514    1.201707  -0.016567     1.397915\n",
       "2     1.18488    1.201752  -0.016872     1.423977\n",
       "3     1.18485    1.201519  -0.016669     1.406886\n",
       "4     1.18469    1.201575  -0.016885     1.425301\n",
       "5     1.18497    1.201281  -0.016311     1.376456\n",
       "6     1.18514    1.201171  -0.016031     1.352630\n",
       "7     1.18524    1.201427  -0.016187     1.365744\n",
       "8     1.18521    1.201704  -0.016494     1.391664\n",
       "9     1.18498    1.201861  -0.016881     1.424573\n",
       "10    1.18502    1.201622  -0.016602     1.401000\n",
       "11    1.18499    1.201588  -0.016598     1.400720\n",
       "12    1.18497    1.201528  -0.016558     1.397341\n",
       "13    1.18476    1.201476  -0.016716     1.410947\n",
       "14    1.18490    1.201295  -0.016395     1.383632\n",
       "15    1.18471    1.201277  -0.016567     1.398383\n",
       "16    1.18474    1.201177  -0.016437     1.387433\n",
       "17    1.18474    1.201377  -0.016637     1.404307\n",
       "18    1.18496    1.201480  -0.016520     1.394112\n",
       "19    1.18513    1.201250  -0.016120     1.360225\n",
       "20    1.18484    1.201086  -0.016246     1.371190\n",
       "21    1.18519    1.201521  -0.016331     1.377956\n",
       "22    1.18503    1.201230  -0.016200     1.367088\n",
       "23    1.18514    1.201699  -0.016559     1.397220\n",
       "24    1.18466    1.201730  -0.017070     1.440881\n",
       "25    1.18414    1.200984  -0.016844     1.422467\n",
       "26    1.18427    1.200763  -0.016493     1.392711\n",
       "27    1.18378    1.200426  -0.016646     1.406172\n",
       "28    1.18471    1.200484  -0.015774     1.331478\n",
       "29    1.18479    1.200080  -0.015290     1.290537\n",
       "...       ...         ...        ...          ...\n",
       "3442  1.19926    1.218956  -0.019696     1.642385\n",
       "3443  1.19952    1.219204  -0.019684     1.640985\n",
       "3444  1.19911    1.219689  -0.020579     1.716220\n",
       "3445  1.19981    1.219091  -0.019281     1.606979\n",
       "3446  1.20069    1.219039  -0.018349     1.528182\n",
       "3447  1.20082    1.220027  -0.019207     1.599508\n",
       "3448  1.20070    1.221001  -0.020301     1.690736\n",
       "3449  1.20141    1.222078  -0.020668     1.720279\n",
       "3450  1.20215    1.221685  -0.019535     1.625039\n",
       "3451  1.20208    1.222760  -0.020680     1.720328\n",
       "3452  1.20229    1.224053  -0.021763     1.810121\n",
       "3453  1.20210    1.223820  -0.021720     1.806816\n",
       "3454  1.20197    1.224236  -0.022266     1.852500\n",
       "3455  1.20225    1.224174  -0.021924     1.823622\n",
       "3456  1.20189    1.223920  -0.022030     1.832955\n",
       "3457  1.20155    1.224083  -0.022533     1.875343\n",
       "3458  1.20133    1.223351  -0.022021     1.833032\n",
       "3459  1.20094    1.222963  -0.022023     1.833811\n",
       "3460  1.20100    1.222607  -0.021607     1.799076\n",
       "3461  1.20081    1.222203  -0.021393     1.781509\n",
       "3462  1.20106    1.222192  -0.021132     1.759430\n",
       "3463  1.20081    1.221870  -0.021060     1.753831\n",
       "3464  1.20053    1.222101  -0.021571     1.796827\n",
       "3465  1.20024    1.221787  -0.021547     1.795242\n",
       "3466  1.19983    1.221326  -0.021496     1.791547\n",
       "3467  1.20024    1.220780  -0.020540     1.711336\n",
       "3468  1.19963    1.220399  -0.020769     1.731276\n",
       "3469  1.19966    1.220607  -0.020947     1.746072\n",
       "3470  1.19982    1.220425  -0.020605     1.717372\n",
       "3471  1.19982    1.219630  -0.019810     1.651042\n",
       "\n",
       "[3472 rows x 4 columns]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfvisual = pd.DataFrame()\n",
    "dfvisual[\"y_test\"] = y_test\n",
    "dfvisual[\"prediction\"]= prediction\n",
    "dfvisual[\"Abs.error\"]=dfvisual[\"y_test\"]-dfvisual[\"prediction\"]\n",
    "dfvisual[\"Relat.error\"]=abs(dfvisual[\"Abs.error\"]/dfvisual[\"y_test\"])*100\n",
    "dfvisual"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x7f52328e8cc0>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Seaborn visualization library\n",
    "import seaborn as sns\n",
    "\n",
    "# Create the default pairplot\n",
    "sns.pairplot(dfvisual)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "OK, better results, but still not very good results. We could try to improve them with a deeper network (more layers) or retouching the net parameters and number of neurons. That is another story."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting prediction session...\n",
      "INFO:tensorflow:Restoring parameters from ../models/07_First_Forex_Prediction\n",
      "Model restored from file: ../models/07_First_Forex_Prediction\n",
      "[[1.2739148]]\n"
     ]
    }
   ],
   "source": [
    "# Running a new session for predictions and export model to production\n",
    "print(\"Starting prediction session...\")\n",
    "with tf.Session() as sess:\n",
    "    # Initialize variables\n",
    "    sess.run(init)\n",
    "\n",
    "    # Try to restore a model if any.\n",
    "    try:\n",
    "        saver.restore(sess, model_path)\n",
    "        print(\"Model restored from file: %s\" % model_path)\n",
    "        # We try to predict the close price of test samples\n",
    "        X_test_pif = [[1.87374825, 1.87106024, 1.87083053, 1.86800846]]\n",
    "        feed_dict = {X: X_test_pif}\n",
    "        \n",
    "        prediction = sess.run(y_hat, feed_dict)\n",
    "        print(prediction)\n",
    "\n",
    "    except Exception:\n",
    "        print(\"Unexpected error:\", sys.exc_info()[0])\n",
    "        pass\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.21419018, 1.21319001, 1.22028069, 1.21914221],\n",
       "       [1.2183674 , 1.2181648 , 1.2244598 , 1.22232735],\n",
       "       [1.22035655, 1.21637387, 1.22465881, 1.22073478],\n",
       "       ...,\n",
       "       [1.52151387, 1.51705039, 1.51281846, 1.5091894 ],\n",
       "       [1.51117029, 1.50809577, 1.51381349, 1.50978661],\n",
       "       [1.50818657, 1.51227459, 1.5066493 , 1.51297176]])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting prediction session...\n",
      "INFO:tensorflow:Restoring parameters from ../models/07_First_Forex_Prediction\n",
      "Model restored from file: ../models/07_First_Forex_Prediction\n",
      "[[1.194145 ]\n",
      " [1.1944305]\n",
      " [1.1943581]\n",
      " ...\n",
      " [1.217765 ]\n",
      " [1.2181967]\n",
      " [1.2174956]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:No assets to save.\n",
      "INFO:tensorflow:No assets to write.\n",
      "INFO:tensorflow:SavedModel written to: b'/home/parrondo/PRODUCTION/models/07_First_Forex_Prediction/47/saved_model.pb'\n"
     ]
    }
   ],
   "source": [
    "# Running a new session for predictions and export model to production\n",
    "print(\"Starting prediction session...\")\n",
    "with tf.Session() as sess:\n",
    "    # Initialize variables\n",
    "    sess.run(init)\n",
    "\n",
    "    saver.restore(sess, model_path)\n",
    "    print(\"Model restored from file: %s\" % model_path)\n",
    "    # We try to predict the close price of test samples\n",
    "    feed_dict = {X: X_test_std}\n",
    "    \n",
    "    prediction = sess.run(y_hat, feed_dict)\n",
    "    print(prediction) \n",
    "\n",
    "    %matplotlib inline\n",
    "    # Plot Prices over time\n",
    "    plt.plot(y_test, 'k-', label='y_test')\n",
    "    plt.plot(prediction, 'r--', label='prediction')\n",
    "    plt.title('Price over time')\n",
    "    plt.legend(loc='upper right')\n",
    "    plt.xlabel('Time')\n",
    "    plt.ylabel('Price')\n",
    "    plt.show()\n",
    "        \n",
    "\n",
    "    # Pick out the model input and output\n",
    "    X_tensor = sess.graph.get_tensor_by_name(\"X\"+ ':0')\n",
    "    y_tensor = sess.graph.get_tensor_by_name(\"out_layer\" + ':0')\n",
    "\n",
    "    model_input = build_tensor_info(X_tensor)\n",
    "    model_output = build_tensor_info(y_tensor)\n",
    "         \n",
    "    # Create a signature definition for tfserving\n",
    "    signature_definition = signature_def_utils.build_signature_def(\n",
    "         inputs={\"X\": model_input},\n",
    "         outputs={\"out_layer\": model_output},\n",
    "         method_name=signature_constants.PREDICT_METHOD_NAME)\n",
    "    \n",
    "    model_version = 1\n",
    "    export_model_dir = production_model_path+\"/\"+str(model_version)\n",
    "    while os.path.exists(export_model_dir):\n",
    "        model_version += 1\n",
    "        export_model_dir = production_model_path+\"/\"+str(model_version)       \n",
    "    \n",
    "    builder = saved_model_builder.SavedModelBuilder(export_model_dir)\n",
    "    builder.add_meta_graph_and_variables(sess,\n",
    "                                         [tag_constants.SERVING],\n",
    "                                         signature_def_map={\n",
    "                                             signature_constants.DEFAULT_SERVING_SIGNATURE_DEF_KEY:\n",
    "                                             signature_definition})\n",
    "\n",
    "    # Save the model so we can serve it with a model server :)\n",
    "    builder.save()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "31244"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>EURUSDopen</th>\n",
       "      <th>EURUSDhigh</th>\n",
       "      <th>EURUSDlow</th>\n",
       "      <th>EURUSDclose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.437944</td>\n",
       "      <td>-1.432007</td>\n",
       "      <td>-1.433057</td>\n",
       "      <td>-1.428509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.426407</td>\n",
       "      <td>-1.432007</td>\n",
       "      <td>-1.429674</td>\n",
       "      <td>-1.434083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.431976</td>\n",
       "      <td>-1.432803</td>\n",
       "      <td>-1.430470</td>\n",
       "      <td>-1.435875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.431379</td>\n",
       "      <td>-1.436982</td>\n",
       "      <td>-1.438032</td>\n",
       "      <td>-1.439657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.438540</td>\n",
       "      <td>-1.439569</td>\n",
       "      <td>-1.434251</td>\n",
       "      <td>-1.438462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-1.439137</td>\n",
       "      <td>-1.444146</td>\n",
       "      <td>-1.447385</td>\n",
       "      <td>-1.445231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>-1.445303</td>\n",
       "      <td>-1.445738</td>\n",
       "      <td>-1.442609</td>\n",
       "      <td>-1.448217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-1.448685</td>\n",
       "      <td>-1.454294</td>\n",
       "      <td>-1.478032</td>\n",
       "      <td>-1.483652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-1.490855</td>\n",
       "      <td>-1.493496</td>\n",
       "      <td>-1.510669</td>\n",
       "      <td>-1.508934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-1.509951</td>\n",
       "      <td>-1.509017</td>\n",
       "      <td>-1.505097</td>\n",
       "      <td>-1.506147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>-1.504978</td>\n",
       "      <td>-1.496282</td>\n",
       "      <td>-1.500122</td>\n",
       "      <td>-1.494601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>-1.493441</td>\n",
       "      <td>-1.497476</td>\n",
       "      <td>-1.492161</td>\n",
       "      <td>-1.495795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>-1.495430</td>\n",
       "      <td>-1.497277</td>\n",
       "      <td>-1.493355</td>\n",
       "      <td>-1.493605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-1.491850</td>\n",
       "      <td>-1.497476</td>\n",
       "      <td>-1.503107</td>\n",
       "      <td>-1.508735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-1.508161</td>\n",
       "      <td>-1.508022</td>\n",
       "      <td>-1.503306</td>\n",
       "      <td>-1.503161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-1.504182</td>\n",
       "      <td>-1.507226</td>\n",
       "      <td>-1.505495</td>\n",
       "      <td>-1.508735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>-1.507564</td>\n",
       "      <td>-1.511803</td>\n",
       "      <td>-1.503505</td>\n",
       "      <td>-1.509133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-1.507365</td>\n",
       "      <td>-1.512997</td>\n",
       "      <td>-1.506888</td>\n",
       "      <td>-1.509730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-1.508558</td>\n",
       "      <td>-1.514191</td>\n",
       "      <td>-1.508480</td>\n",
       "      <td>-1.510526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>-1.509553</td>\n",
       "      <td>-1.514589</td>\n",
       "      <td>-1.510072</td>\n",
       "      <td>-1.512915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>-1.512139</td>\n",
       "      <td>-1.517773</td>\n",
       "      <td>-1.518430</td>\n",
       "      <td>-1.524063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>-1.524074</td>\n",
       "      <td>-1.524936</td>\n",
       "      <td>-1.521216</td>\n",
       "      <td>-1.526452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>-1.523676</td>\n",
       "      <td>-1.524339</td>\n",
       "      <td>-1.523803</td>\n",
       "      <td>-1.527049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>-1.525068</td>\n",
       "      <td>-1.530110</td>\n",
       "      <td>-1.532957</td>\n",
       "      <td>-1.537401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>-1.537401</td>\n",
       "      <td>-1.543045</td>\n",
       "      <td>-1.540520</td>\n",
       "      <td>-1.539392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>-1.539390</td>\n",
       "      <td>-1.538070</td>\n",
       "      <td>-1.534549</td>\n",
       "      <td>-1.537799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>-1.537600</td>\n",
       "      <td>-1.541254</td>\n",
       "      <td>-1.534748</td>\n",
       "      <td>-1.538197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>-1.536804</td>\n",
       "      <td>-1.539463</td>\n",
       "      <td>-1.532161</td>\n",
       "      <td>-1.535609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>-1.537401</td>\n",
       "      <td>-1.538269</td>\n",
       "      <td>-1.533753</td>\n",
       "      <td>-1.533818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>-1.533224</td>\n",
       "      <td>-1.535085</td>\n",
       "      <td>-1.531365</td>\n",
       "      <td>-1.530234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31214</th>\n",
       "      <td>1.204443</td>\n",
       "      <td>1.199858</td>\n",
       "      <td>1.200579</td>\n",
       "      <td>1.198837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31215</th>\n",
       "      <td>1.196885</td>\n",
       "      <td>1.195878</td>\n",
       "      <td>1.202967</td>\n",
       "      <td>1.198837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31216</th>\n",
       "      <td>1.209615</td>\n",
       "      <td>1.211797</td>\n",
       "      <td>1.215306</td>\n",
       "      <td>1.213568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31217</th>\n",
       "      <td>1.212798</td>\n",
       "      <td>1.208215</td>\n",
       "      <td>1.207345</td>\n",
       "      <td>1.207397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31218</th>\n",
       "      <td>1.207228</td>\n",
       "      <td>1.213986</td>\n",
       "      <td>1.213316</td>\n",
       "      <td>1.218346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31219</th>\n",
       "      <td>1.217771</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.214908</td>\n",
       "      <td>1.210582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31220</th>\n",
       "      <td>1.207427</td>\n",
       "      <td>1.206424</td>\n",
       "      <td>1.204758</td>\n",
       "      <td>1.209985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31221</th>\n",
       "      <td>1.207228</td>\n",
       "      <td>1.209409</td>\n",
       "      <td>1.210728</td>\n",
       "      <td>1.212175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31222</th>\n",
       "      <td>1.211803</td>\n",
       "      <td>1.217767</td>\n",
       "      <td>1.214709</td>\n",
       "      <td>1.217151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31223</th>\n",
       "      <td>1.211803</td>\n",
       "      <td>1.207220</td>\n",
       "      <td>1.206947</td>\n",
       "      <td>1.203814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31224</th>\n",
       "      <td>1.203051</td>\n",
       "      <td>1.207618</td>\n",
       "      <td>1.208738</td>\n",
       "      <td>1.212772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31225</th>\n",
       "      <td>1.213395</td>\n",
       "      <td>1.208812</td>\n",
       "      <td>1.212519</td>\n",
       "      <td>1.207795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31226</th>\n",
       "      <td>1.205040</td>\n",
       "      <td>1.204832</td>\n",
       "      <td>1.208738</td>\n",
       "      <td>1.204013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31227</th>\n",
       "      <td>1.203648</td>\n",
       "      <td>1.209807</td>\n",
       "      <td>1.208738</td>\n",
       "      <td>1.215758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31228</th>\n",
       "      <td>1.214986</td>\n",
       "      <td>1.212593</td>\n",
       "      <td>1.214908</td>\n",
       "      <td>1.213369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31229</th>\n",
       "      <td>1.209615</td>\n",
       "      <td>1.215777</td>\n",
       "      <td>1.215704</td>\n",
       "      <td>1.214763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31230</th>\n",
       "      <td>1.212997</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.217296</td>\n",
       "      <td>1.213170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31231</th>\n",
       "      <td>1.213395</td>\n",
       "      <td>1.210404</td>\n",
       "      <td>1.214709</td>\n",
       "      <td>1.211179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31232</th>\n",
       "      <td>1.208024</td>\n",
       "      <td>1.210205</td>\n",
       "      <td>1.214112</td>\n",
       "      <td>1.210781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31233</th>\n",
       "      <td>1.210013</td>\n",
       "      <td>1.209210</td>\n",
       "      <td>1.214112</td>\n",
       "      <td>1.209388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31234</th>\n",
       "      <td>1.208621</td>\n",
       "      <td>1.218165</td>\n",
       "      <td>1.212917</td>\n",
       "      <td>1.224119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31235</th>\n",
       "      <td>1.221948</td>\n",
       "      <td>1.218762</td>\n",
       "      <td>1.214709</td>\n",
       "      <td>1.209985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31236</th>\n",
       "      <td>1.202653</td>\n",
       "      <td>1.199858</td>\n",
       "      <td>1.203166</td>\n",
       "      <td>1.205406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31237</th>\n",
       "      <td>1.205040</td>\n",
       "      <td>1.206026</td>\n",
       "      <td>1.209136</td>\n",
       "      <td>1.211976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31238</th>\n",
       "      <td>1.209416</td>\n",
       "      <td>1.209608</td>\n",
       "      <td>1.214908</td>\n",
       "      <td>1.214564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31239</th>\n",
       "      <td>1.213792</td>\n",
       "      <td>1.211399</td>\n",
       "      <td>1.215306</td>\n",
       "      <td>1.217151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31240</th>\n",
       "      <td>1.217771</td>\n",
       "      <td>1.217170</td>\n",
       "      <td>1.223863</td>\n",
       "      <td>1.219341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31241</th>\n",
       "      <td>1.216179</td>\n",
       "      <td>1.213190</td>\n",
       "      <td>1.217495</td>\n",
       "      <td>1.218943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31242</th>\n",
       "      <td>1.220158</td>\n",
       "      <td>1.219757</td>\n",
       "      <td>1.225455</td>\n",
       "      <td>1.220735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31243</th>\n",
       "      <td>1.218765</td>\n",
       "      <td>1.219359</td>\n",
       "      <td>1.220679</td>\n",
       "      <td>1.215957</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31244 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       EURUSDopen  EURUSDhigh  EURUSDlow  EURUSDclose\n",
       "0       -1.437944   -1.432007  -1.433057    -1.428509\n",
       "1       -1.426407   -1.432007  -1.429674    -1.434083\n",
       "2       -1.431976   -1.432803  -1.430470    -1.435875\n",
       "3       -1.431379   -1.436982  -1.438032    -1.439657\n",
       "4       -1.438540   -1.439569  -1.434251    -1.438462\n",
       "5       -1.439137   -1.444146  -1.447385    -1.445231\n",
       "6       -1.445303   -1.445738  -1.442609    -1.448217\n",
       "7       -1.448685   -1.454294  -1.478032    -1.483652\n",
       "8       -1.490855   -1.493496  -1.510669    -1.508934\n",
       "9       -1.509951   -1.509017  -1.505097    -1.506147\n",
       "10      -1.504978   -1.496282  -1.500122    -1.494601\n",
       "11      -1.493441   -1.497476  -1.492161    -1.495795\n",
       "12      -1.495430   -1.497277  -1.493355    -1.493605\n",
       "13      -1.491850   -1.497476  -1.503107    -1.508735\n",
       "14      -1.508161   -1.508022  -1.503306    -1.503161\n",
       "15      -1.504182   -1.507226  -1.505495    -1.508735\n",
       "16      -1.507564   -1.511803  -1.503505    -1.509133\n",
       "17      -1.507365   -1.512997  -1.506888    -1.509730\n",
       "18      -1.508558   -1.514191  -1.508480    -1.510526\n",
       "19      -1.509553   -1.514589  -1.510072    -1.512915\n",
       "20      -1.512139   -1.517773  -1.518430    -1.524063\n",
       "21      -1.524074   -1.524936  -1.521216    -1.526452\n",
       "22      -1.523676   -1.524339  -1.523803    -1.527049\n",
       "23      -1.525068   -1.530110  -1.532957    -1.537401\n",
       "24      -1.537401   -1.543045  -1.540520    -1.539392\n",
       "25      -1.539390   -1.538070  -1.534549    -1.537799\n",
       "26      -1.537600   -1.541254  -1.534748    -1.538197\n",
       "27      -1.536804   -1.539463  -1.532161    -1.535609\n",
       "28      -1.537401   -1.538269  -1.533753    -1.533818\n",
       "29      -1.533224   -1.535085  -1.531365    -1.530234\n",
       "...           ...         ...        ...          ...\n",
       "31214    1.204443    1.199858   1.200579     1.198837\n",
       "31215    1.196885    1.195878   1.202967     1.198837\n",
       "31216    1.209615    1.211797   1.215306     1.213568\n",
       "31217    1.212798    1.208215   1.207345     1.207397\n",
       "31218    1.207228    1.213986   1.213316     1.218346\n",
       "31219    1.217771    1.213190   1.214908     1.210582\n",
       "31220    1.207427    1.206424   1.204758     1.209985\n",
       "31221    1.207228    1.209409   1.210728     1.212175\n",
       "31222    1.211803    1.217767   1.214709     1.217151\n",
       "31223    1.211803    1.207220   1.206947     1.203814\n",
       "31224    1.203051    1.207618   1.208738     1.212772\n",
       "31225    1.213395    1.208812   1.212519     1.207795\n",
       "31226    1.205040    1.204832   1.208738     1.204013\n",
       "31227    1.203648    1.209807   1.208738     1.215758\n",
       "31228    1.214986    1.212593   1.214908     1.213369\n",
       "31229    1.209615    1.215777   1.215704     1.214763\n",
       "31230    1.212997    1.213190   1.217296     1.213170\n",
       "31231    1.213395    1.210404   1.214709     1.211179\n",
       "31232    1.208024    1.210205   1.214112     1.210781\n",
       "31233    1.210013    1.209210   1.214112     1.209388\n",
       "31234    1.208621    1.218165   1.212917     1.224119\n",
       "31235    1.221948    1.218762   1.214709     1.209985\n",
       "31236    1.202653    1.199858   1.203166     1.205406\n",
       "31237    1.205040    1.206026   1.209136     1.211976\n",
       "31238    1.209416    1.209608   1.214908     1.214564\n",
       "31239    1.213792    1.211399   1.215306     1.217151\n",
       "31240    1.217771    1.217170   1.223863     1.219341\n",
       "31241    1.216179    1.213190   1.217495     1.218943\n",
       "31242    1.220158    1.219757   1.225455     1.220735\n",
       "31243    1.218765    1.219359   1.220679     1.215957\n",
       "\n",
       "[31244 rows x 4 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_Train_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(31244,)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.41867760978729"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean relative error\n",
    "dfvisual[\"Relat.error\"].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.016801861367884887"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean Absoulte error\n",
    "dfvisual[\"Abs.error\"].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "deeptrading",
   "language": "python",
   "name": "deeptrading"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}