btc_dominant_features.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# BTC Dominant Features - Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import libraries\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pathlib import Path\n",
    "import os\n",
    "import requests\n",
    "import datetime\n",
    "\n",
    "from collections import Counter\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read in annual coin metrics from csv\n",
    "metrics_2021 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_1.csv', parse_dates=True)\n",
    "metrics_2020 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_2.csv', parse_dates=True)\n",
    "metrics_2019 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_3.csv', parse_dates=True)\n",
    "metrics_2018 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_4.csv', parse_dates=True)\n",
    "metrics_2017 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_5.csv', parse_dates=True)\n",
    "metrics_2016 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_6.csv', parse_dates=True)\n",
    "metrics_2015 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_7.csv', parse_dates=True)\n",
    "metrics_2014 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_8.csv', parse_dates=True)\n",
    "metrics_2013 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_9.csv', parse_dates=True)\n",
    "metrics_2012 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_10.csv', parse_dates=True)\n",
    "metrics_2011 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_11.csv', parse_dates=True)\n",
    "metrics_2010 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_12.csv', parse_dates=True)\n",
    "metrics_2009 = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_annual_metrics/btc_metrics_13.csv', parse_dates=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BTC_Trends</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-12-27</th>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-28</th>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-29</th>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-30</th>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-31</th>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            BTC_Trends\n",
       "time                  \n",
       "2021-12-27          45\n",
       "2021-12-28          45\n",
       "2021-12-29          46\n",
       "2021-12-30          43\n",
       "2021-12-31          41"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in combined BTC Google trends data\n",
    "btc_google_trends = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/btc_google_trends/btc_trends_combined.csv', parse_dates=True)\n",
    "btc_google_trends.reset_index(inplace=True)\n",
    "btc_google_trends.rename(columns={'date':'time', 'bitcoin':'BTC_Trends'},inplace=True)\n",
    "btc_google_trends = btc_google_trends.drop(columns='index')\n",
    "btc_google_trends['time'] = pd.to_datetime(btc_google_trends['time']).dt.date\n",
    "btc_google_trends = btc_google_trends.set_index('time')\n",
    "btc_google_trends.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>asset</th>\n",
       "      <th>AdrActCnt</th>\n",
       "      <th>CapMVRVCur</th>\n",
       "      <th>CapMrktCurUSD</th>\n",
       "      <th>HashRate</th>\n",
       "      <th>NVTAdj</th>\n",
       "      <th>PriceUSD</th>\n",
       "      <th>SplyCur</th>\n",
       "      <th>TxCnt</th>\n",
       "      <th>TxTfrValAdjUSD</th>\n",
       "      <th>VelCur1yr</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>346</th>\n",
       "      <td>2021-12-13 00:00:00+00:00</td>\n",
       "      <td>btc</td>\n",
       "      <td>996713</td>\n",
       "      <td>1.915300</td>\n",
       "      <td>8.842453e+11</td>\n",
       "      <td>1.695883e+08</td>\n",
       "      <td>69.789128</td>\n",
       "      <td>46785.327841</td>\n",
       "      <td>1.890006e+07</td>\n",
       "      <td>281785</td>\n",
       "      <td>1.267024e+10</td>\n",
       "      <td>14.525969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>347</th>\n",
       "      <td>2021-12-14 00:00:00+00:00</td>\n",
       "      <td>btc</td>\n",
       "      <td>966596</td>\n",
       "      <td>1.974878</td>\n",
       "      <td>9.127791e+11</td>\n",
       "      <td>1.780076e+08</td>\n",
       "      <td>72.469440</td>\n",
       "      <td>48292.682250</td>\n",
       "      <td>1.890098e+07</td>\n",
       "      <td>281921</td>\n",
       "      <td>1.259537e+10</td>\n",
       "      <td>14.533207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>348</th>\n",
       "      <td>2021-12-15 00:00:00+00:00</td>\n",
       "      <td>btc</td>\n",
       "      <td>1058257</td>\n",
       "      <td>1.995927</td>\n",
       "      <td>9.229897e+11</td>\n",
       "      <td>1.695883e+08</td>\n",
       "      <td>62.867727</td>\n",
       "      <td>48830.622386</td>\n",
       "      <td>1.890186e+07</td>\n",
       "      <td>280841</td>\n",
       "      <td>1.468146e+10</td>\n",
       "      <td>14.536127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>349</th>\n",
       "      <td>2021-12-16 00:00:00+00:00</td>\n",
       "      <td>btc</td>\n",
       "      <td>873836</td>\n",
       "      <td>1.948720</td>\n",
       "      <td>9.009626e+11</td>\n",
       "      <td>1.383167e+08</td>\n",
       "      <td>56.242508</td>\n",
       "      <td>47663.469108</td>\n",
       "      <td>1.890258e+07</td>\n",
       "      <td>256019</td>\n",
       "      <td>1.601925e+10</td>\n",
       "      <td>14.533316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>350</th>\n",
       "      <td>2021-12-17 00:00:00+00:00</td>\n",
       "      <td>btc</td>\n",
       "      <td>979005</td>\n",
       "      <td>1.896527</td>\n",
       "      <td>8.758845e+11</td>\n",
       "      <td>1.792104e+08</td>\n",
       "      <td>48.186258</td>\n",
       "      <td>46334.483749</td>\n",
       "      <td>1.890351e+07</td>\n",
       "      <td>260119</td>\n",
       "      <td>1.817706e+10</td>\n",
       "      <td>14.538415</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          time asset  AdrActCnt  CapMVRVCur  CapMrktCurUSD  \\\n",
       "346  2021-12-13 00:00:00+00:00   btc     996713    1.915300   8.842453e+11   \n",
       "347  2021-12-14 00:00:00+00:00   btc     966596    1.974878   9.127791e+11   \n",
       "348  2021-12-15 00:00:00+00:00   btc    1058257    1.995927   9.229897e+11   \n",
       "349  2021-12-16 00:00:00+00:00   btc     873836    1.948720   9.009626e+11   \n",
       "350  2021-12-17 00:00:00+00:00   btc     979005    1.896527   8.758845e+11   \n",
       "\n",
       "         HashRate     NVTAdj      PriceUSD       SplyCur   TxCnt  \\\n",
       "346  1.695883e+08  69.789128  46785.327841  1.890006e+07  281785   \n",
       "347  1.780076e+08  72.469440  48292.682250  1.890098e+07  281921   \n",
       "348  1.695883e+08  62.867727  48830.622386  1.890186e+07  280841   \n",
       "349  1.383167e+08  56.242508  47663.469108  1.890258e+07  256019   \n",
       "350  1.792104e+08  48.186258  46334.483749  1.890351e+07  260119   \n",
       "\n",
       "     TxTfrValAdjUSD  VelCur1yr  \n",
       "346    1.267024e+10  14.525969  \n",
       "347    1.259537e+10  14.533207  \n",
       "348    1.468146e+10  14.536127  \n",
       "349    1.601925e+10  14.533316  \n",
       "350    1.817706e+10  14.538415  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Concat all csv files  and print combined csv\n",
    "combined = pd.concat([metrics_2009,metrics_2010,metrics_2011,metrics_2012,metrics_2013,metrics_2014,metrics_2015,metrics_2016,metrics_2017,metrics_2018,metrics_2019,metrics_2020,metrics_2021], axis=0)\n",
    "combined_df = pd.DataFrame(combined)\n",
    "combined_df = combined_df.drop_duplicates(keep='first')\n",
    "combined_df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time                0\n",
       "asset               0\n",
       "AdrActCnt           0\n",
       "CapMVRVCur        561\n",
       "CapMrktCurUSD     561\n",
       "HashRate            6\n",
       "NVTAdj            260\n",
       "PriceUSD          561\n",
       "SplyCur             0\n",
       "TxCnt               0\n",
       "TxTfrValAdjUSD    561\n",
       "VelCur1yr           6\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combined_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time              0\n",
       "asset             0\n",
       "AdrActCnt         0\n",
       "CapMVRVCur        0\n",
       "CapMrktCurUSD     0\n",
       "HashRate          0\n",
       "NVTAdj            0\n",
       "PriceUSD          0\n",
       "SplyCur           0\n",
       "TxCnt             0\n",
       "TxTfrValAdjUSD    0\n",
       "VelCur1yr         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combined_df.dropna(inplace=True)\n",
    "combined_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove timestamp in time column\n",
    "combined_df['time'] = pd.to_datetime(combined_df['time']).dt.date\n",
    "combined_df = combined_df.set_index('time')\n",
    "combined.to_csv('resources/btc_annual_metrics/combined.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>asset</th>\n",
       "      <th>AdrActCnt</th>\n",
       "      <th>CapMVRVCur</th>\n",
       "      <th>CapMrktCurUSD</th>\n",
       "      <th>HashRate</th>\n",
       "      <th>NVTAdj</th>\n",
       "      <th>PriceUSD</th>\n",
       "      <th>SplyCur</th>\n",
       "      <th>TxCnt</th>\n",
       "      <th>TxTfrValAdjUSD</th>\n",
       "      <th>VelCur1yr</th>\n",
       "      <th>BTC_Trends</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-12-13</th>\n",
       "      <td>btc</td>\n",
       "      <td>996713.0</td>\n",
       "      <td>1.915300</td>\n",
       "      <td>8.842453e+11</td>\n",
       "      <td>1.695883e+08</td>\n",
       "      <td>69.789128</td>\n",
       "      <td>46785.327841</td>\n",
       "      <td>1.890006e+07</td>\n",
       "      <td>281785.0</td>\n",
       "      <td>1.267024e+10</td>\n",
       "      <td>14.525969</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-14</th>\n",
       "      <td>btc</td>\n",
       "      <td>966596.0</td>\n",
       "      <td>1.974878</td>\n",
       "      <td>9.127791e+11</td>\n",
       "      <td>1.780076e+08</td>\n",
       "      <td>72.469440</td>\n",
       "      <td>48292.682250</td>\n",
       "      <td>1.890098e+07</td>\n",
       "      <td>281921.0</td>\n",
       "      <td>1.259537e+10</td>\n",
       "      <td>14.533207</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-15</th>\n",
       "      <td>btc</td>\n",
       "      <td>1058257.0</td>\n",
       "      <td>1.995927</td>\n",
       "      <td>9.229897e+11</td>\n",
       "      <td>1.695883e+08</td>\n",
       "      <td>62.867727</td>\n",
       "      <td>48830.622386</td>\n",
       "      <td>1.890186e+07</td>\n",
       "      <td>280841.0</td>\n",
       "      <td>1.468146e+10</td>\n",
       "      <td>14.536127</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-16</th>\n",
       "      <td>btc</td>\n",
       "      <td>873836.0</td>\n",
       "      <td>1.948720</td>\n",
       "      <td>9.009626e+11</td>\n",
       "      <td>1.383167e+08</td>\n",
       "      <td>56.242508</td>\n",
       "      <td>47663.469108</td>\n",
       "      <td>1.890258e+07</td>\n",
       "      <td>256019.0</td>\n",
       "      <td>1.601925e+10</td>\n",
       "      <td>14.533316</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-17</th>\n",
       "      <td>btc</td>\n",
       "      <td>979005.0</td>\n",
       "      <td>1.896527</td>\n",
       "      <td>8.758845e+11</td>\n",
       "      <td>1.792104e+08</td>\n",
       "      <td>48.186258</td>\n",
       "      <td>46334.483749</td>\n",
       "      <td>1.890351e+07</td>\n",
       "      <td>260119.0</td>\n",
       "      <td>1.817706e+10</td>\n",
       "      <td>14.538415</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           asset  AdrActCnt  CapMVRVCur  CapMrktCurUSD      HashRate  \\\n",
       "time                                                                   \n",
       "2021-12-13   btc   996713.0    1.915300   8.842453e+11  1.695883e+08   \n",
       "2021-12-14   btc   966596.0    1.974878   9.127791e+11  1.780076e+08   \n",
       "2021-12-15   btc  1058257.0    1.995927   9.229897e+11  1.695883e+08   \n",
       "2021-12-16   btc   873836.0    1.948720   9.009626e+11  1.383167e+08   \n",
       "2021-12-17   btc   979005.0    1.896527   8.758845e+11  1.792104e+08   \n",
       "\n",
       "               NVTAdj      PriceUSD       SplyCur     TxCnt  TxTfrValAdjUSD  \\\n",
       "time                                                                          \n",
       "2021-12-13  69.789128  46785.327841  1.890006e+07  281785.0    1.267024e+10   \n",
       "2021-12-14  72.469440  48292.682250  1.890098e+07  281921.0    1.259537e+10   \n",
       "2021-12-15  62.867727  48830.622386  1.890186e+07  280841.0    1.468146e+10   \n",
       "2021-12-16  56.242508  47663.469108  1.890258e+07  256019.0    1.601925e+10   \n",
       "2021-12-17  48.186258  46334.483749  1.890351e+07  260119.0    1.817706e+10   \n",
       "\n",
       "            VelCur1yr  BTC_Trends  \n",
       "time                               \n",
       "2021-12-13  14.525969          51  \n",
       "2021-12-14  14.533207          57  \n",
       "2021-12-15  14.536127          53  \n",
       "2021-12-16  14.533316          48  \n",
       "2021-12-17  14.538415          49  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Concat metrics and google trends data\n",
    "combined_df = pd.concat([combined_df,btc_google_trends],axis=1).dropna()\n",
    "combined_df = combined_df.reset_index()\n",
    "combined_df = combined_df.rename(columns={'index':'time'})\n",
    "combined_df = combined_df.set_index('time')\n",
    "combined_df.tail()"
   ]
  },
  {
   "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>RSI_14</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-12-13</th>\n",
       "      <td>33.164148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-14</th>\n",
       "      <td>37.885222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-15</th>\n",
       "      <td>39.526932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-16</th>\n",
       "      <td>37.227881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-17</th>\n",
       "      <td>34.749431</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               RSI_14\n",
       "time                 \n",
       "2021-12-13  33.164148\n",
       "2021-12-14  37.885222\n",
       "2021-12-15  39.526932\n",
       "2021-12-16  37.227881\n",
       "2021-12-17  34.749431"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in Relative Strength(RSI) data\n",
    "rsi_btc = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/rsi_btc.csv', parse_dates=True, index_col='time')\n",
    "rsi_btc.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>ROC_14</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-12-13</th>\n",
       "      <td>-19.225218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-14</th>\n",
       "      <td>-15.419950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-15</th>\n",
       "      <td>-14.602830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-16</th>\n",
       "      <td>-15.763465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-17</th>\n",
       "      <td>-13.670414</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               ROC_14\n",
       "time                 \n",
       "2021-12-13 -19.225218\n",
       "2021-12-14 -15.419950\n",
       "2021-12-15 -14.602830\n",
       "2021-12-16 -15.763465\n",
       "2021-12-17 -13.670414"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in Rate of Change (ROC) data\n",
    "roc_btc = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/roc_btc.csv', parse_dates=True, index_col='time')\n",
    "roc_btc.tail()"
   ]
  },
  {
   "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>MFI_14</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2022-01-22</th>\n",
       "      <td>26.423542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-01-23</th>\n",
       "      <td>26.598839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-01-24</th>\n",
       "      <td>26.323550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-01-25</th>\n",
       "      <td>25.662763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-01-26</th>\n",
       "      <td>24.314011</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               MFI_14\n",
       "time                 \n",
       "2022-01-22  26.423542\n",
       "2022-01-23  26.598839\n",
       "2022-01-24  26.323550\n",
       "2022-01-25  25.662763\n",
       "2022-01-26  24.314011"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in Money Flow Index (MFI) data\n",
    "mfi_btc = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/mfi_btc.csv', parse_dates=True)\n",
    "mfi_btc = mfi_btc.rename(columns={'Date':'time'})\n",
    "mfi_btc['time'] = pd.to_datetime(mfi_btc['time']).dt.date\n",
    "mfi_btc = mfi_btc.set_index('time').dropna()\n",
    "mfi_btc.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>TSI_13_25_13</th>\n",
       "      <th>TSIs_13_25_13</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-12-13</th>\n",
       "      <td>-28.461637</td>\n",
       "      <td>-22.250541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-14</th>\n",
       "      <td>-27.879649</td>\n",
       "      <td>-23.054700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-15</th>\n",
       "      <td>-26.909529</td>\n",
       "      <td>-23.605390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-16</th>\n",
       "      <td>-26.771374</td>\n",
       "      <td>-24.057673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-17</th>\n",
       "      <td>-27.390082</td>\n",
       "      <td>-24.533731</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            TSI_13_25_13  TSIs_13_25_13\n",
       "time                                   \n",
       "2021-12-13    -28.461637     -22.250541\n",
       "2021-12-14    -27.879649     -23.054700\n",
       "2021-12-15    -26.909529     -23.605390\n",
       "2021-12-16    -26.771374     -24.057673\n",
       "2021-12-17    -27.390082     -24.533731"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in True Strength (TSI) data\n",
    "tsi_btc = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/tsi_btc.csv', parse_dates=True)\n",
    "tsi_btc = tsi_btc.rename(columns={'Date':'time'})\n",
    "tsi_btc['time'] = pd.to_datetime(tsi_btc['time']).dt.date\n",
    "tsi_btc = tsi_btc.set_index('time').dropna()\n",
    "tsi_btc.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>STOCHk_14_3_3</th>\n",
       "      <th>STOCHd_14_3_3</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-10-04</th>\n",
       "      <td>-37.509344</td>\n",
       "      <td>-25.445139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-10-05</th>\n",
       "      <td>-48.047226</td>\n",
       "      <td>-35.980253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-10-06</th>\n",
       "      <td>-50.582789</td>\n",
       "      <td>-45.379786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-10-07</th>\n",
       "      <td>-39.987669</td>\n",
       "      <td>-46.205895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-10-08</th>\n",
       "      <td>-26.924530</td>\n",
       "      <td>-39.164996</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            STOCHk_14_3_3  STOCHd_14_3_3\n",
       "time                                    \n",
       "2014-10-04     -37.509344     -25.445139\n",
       "2014-10-05     -48.047226     -35.980253\n",
       "2014-10-06     -50.582789     -45.379786\n",
       "2014-10-07     -39.987669     -46.205895\n",
       "2014-10-08     -26.924530     -39.164996"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read in Stochastic Oscilator (TSI) data\n",
    "stoch_btc = pd.read_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/stoch_btc.csv', parse_dates=True)\n",
    "stoch_btc = stoch_btc.rename(columns={'Date':'time'})\n",
    "stoch_btc['time'] = pd.to_datetime(stoch_btc['time']).dt.date\n",
    "stoch_btc = stoch_btc.set_index('time').dropna()\n",
    "stoch_btc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>asset</th>\n",
       "      <th>AdrActCnt</th>\n",
       "      <th>CapMVRVCur</th>\n",
       "      <th>CapMrktCurUSD</th>\n",
       "      <th>HashRate</th>\n",
       "      <th>NVTAdj</th>\n",
       "      <th>PriceUSD</th>\n",
       "      <th>SplyCur</th>\n",
       "      <th>TxCnt</th>\n",
       "      <th>TxTfrValAdjUSD</th>\n",
       "      <th>VelCur1yr</th>\n",
       "      <th>BTC_Trends</th>\n",
       "      <th>RSI_14</th>\n",
       "      <th>STOCHk_14_3_3</th>\n",
       "      <th>STOCHd_14_3_3</th>\n",
       "      <th>ROC_14</th>\n",
       "      <th>MFI_14</th>\n",
       "      <th>TSI_13_25_13</th>\n",
       "      <th>TSIs_13_25_13</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-07-18</th>\n",
       "      <td>btc</td>\n",
       "      <td>860.0</td>\n",
       "      <td>146.038332</td>\n",
       "      <td>295959.152000</td>\n",
       "      <td>0.001552</td>\n",
       "      <td>200.680421</td>\n",
       "      <td>0.085840</td>\n",
       "      <td>3447800.0</td>\n",
       "      <td>248.0</td>\n",
       "      <td>1474.778412</td>\n",
       "      <td>1.820565</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-19</th>\n",
       "      <td>btc</td>\n",
       "      <td>929.0</td>\n",
       "      <td>62.565130</td>\n",
       "      <td>279285.200000</td>\n",
       "      <td>0.001570</td>\n",
       "      <td>85.889156</td>\n",
       "      <td>0.080800</td>\n",
       "      <td>3456500.0</td>\n",
       "      <td>334.0</td>\n",
       "      <td>3251.693384</td>\n",
       "      <td>1.828863</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-20</th>\n",
       "      <td>btc</td>\n",
       "      <td>936.0</td>\n",
       "      <td>45.106630</td>\n",
       "      <td>259000.404990</td>\n",
       "      <td>0.001633</td>\n",
       "      <td>215.744281</td>\n",
       "      <td>0.074736</td>\n",
       "      <td>3465550.0</td>\n",
       "      <td>423.0</td>\n",
       "      <td>1200.497201</td>\n",
       "      <td>1.831412</td>\n",
       "      <td>17.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-21</th>\n",
       "      <td>btc</td>\n",
       "      <td>784.0</td>\n",
       "      <td>38.357560</td>\n",
       "      <td>275266.471297</td>\n",
       "      <td>0.001868</td>\n",
       "      <td>166.836660</td>\n",
       "      <td>0.079193</td>\n",
       "      <td>3475900.0</td>\n",
       "      <td>247.0</td>\n",
       "      <td>1649.915979</td>\n",
       "      <td>1.833749</td>\n",
       "      <td>34.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-07-22</th>\n",
       "      <td>btc</td>\n",
       "      <td>594.0</td>\n",
       "      <td>25.132348</td>\n",
       "      <td>203749.573975</td>\n",
       "      <td>0.001588</td>\n",
       "      <td>105.440311</td>\n",
       "      <td>0.058470</td>\n",
       "      <td>3484700.0</td>\n",
       "      <td>221.0</td>\n",
       "      <td>1932.368865</td>\n",
       "      <td>1.840245</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           asset  AdrActCnt  CapMVRVCur  CapMrktCurUSD  HashRate      NVTAdj  \\\n",
       "time                                                                           \n",
       "2010-07-18   btc      860.0  146.038332  295959.152000  0.001552  200.680421   \n",
       "2010-07-19   btc      929.0   62.565130  279285.200000  0.001570   85.889156   \n",
       "2010-07-20   btc      936.0   45.106630  259000.404990  0.001633  215.744281   \n",
       "2010-07-21   btc      784.0   38.357560  275266.471297  0.001868  166.836660   \n",
       "2010-07-22   btc      594.0   25.132348  203749.573975  0.001588  105.440311   \n",
       "\n",
       "            PriceUSD    SplyCur  TxCnt  TxTfrValAdjUSD  VelCur1yr  BTC_Trends  \\\n",
       "time                                                                            \n",
       "2010-07-18  0.085840  3447800.0  248.0     1474.778412   1.820565         0.0   \n",
       "2010-07-19  0.080800  3456500.0  334.0     3251.693384   1.828863         0.0   \n",
       "2010-07-20  0.074736  3465550.0  423.0     1200.497201   1.831412        17.0   \n",
       "2010-07-21  0.079193  3475900.0  247.0     1649.915979   1.833749        34.0   \n",
       "2010-07-22  0.058470  3484700.0  221.0     1932.368865   1.840245         0.0   \n",
       "\n",
       "            RSI_14  STOCHk_14_3_3  STOCHd_14_3_3  ROC_14  MFI_14  \\\n",
       "time                                                               \n",
       "2010-07-18     NaN            NaN            NaN     NaN     NaN   \n",
       "2010-07-19     NaN            NaN            NaN     NaN     NaN   \n",
       "2010-07-20     NaN            NaN            NaN     NaN     NaN   \n",
       "2010-07-21     NaN            NaN            NaN     NaN     NaN   \n",
       "2010-07-22     NaN            NaN            NaN     NaN     NaN   \n",
       "\n",
       "            TSI_13_25_13  TSIs_13_25_13  \n",
       "time                                     \n",
       "2010-07-18           NaN            NaN  \n",
       "2010-07-19           NaN            NaN  \n",
       "2010-07-20           NaN            NaN  \n",
       "2010-07-21           NaN            NaN  \n",
       "2010-07-22           NaN            NaN  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create model features dataframe\n",
    "features_btc = pd.concat([combined_df,rsi_btc, stoch_btc, roc_btc, mfi_btc,tsi_btc], axis=1)\n",
    "features_btc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'EMA BTC'}, xlabel='time', ylabel='BTC Price ($)'>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Set short and long windows for EMA (Exponential Moving Average) to create trading signals features\n",
    "short_window = 20\n",
    "long_window = 50\n",
    "\n",
    "# Construct a `Fast` and `Slow` Exponential Moving Average from short and long windows, respectively\n",
    "features_btc['Fast_EMA'] = features_btc['PriceUSD'].ewm(halflife=short_window).mean()\n",
    "features_btc['Slow_EMA'] = features_btc['PriceUSD'].ewm(halflife=long_window).mean()\n",
    "\n",
    "# Plot the EMA of ETH closing prices\n",
    "features_btc[['PriceUSD', 'Fast_EMA', 'Slow_EMA']].plot(figsize=(12,7), title='EMA BTC', ylabel='BTC Price ($)', rot=45) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Construct a crossover trading signal\n",
    "features_btc['Crossover_Long'] = np.where(features_btc['Fast_EMA'] > features_btc['Slow_EMA'], 1.0, 0.0)\n",
    "features_btc['Crossover_Short'] = np.where(features_btc['Fast_EMA'] < features_btc['Slow_EMA'],-1.0, 0.0)\n",
    "features_btc['Crossover_Signal'] = features_btc['Crossover_Long'] + features_btc['Crossover_Short']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate daily return\n",
    "features_btc['Daily_Return'] = features_btc['PriceUSD'].pct_change()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Construct the dependent variable where if daily return is greater than 0, then 1, else, 0.\n",
    "features_btc['Positive_Return'] = np.where(features_btc['Daily_Return'] > 0, 1.0, 0.0)\n",
    "\n",
    "features_btc = features_btc.dropna()\n",
    "\n",
    "# Convert date to str to set test and train windows \n",
    "features_btc.index = pd.to_datetime(features_btc.index, format = '%Y/%m/%d')\n",
    "features_btc.tail()\n",
    "features_btc.to_csv('/Users/draganbogatic/Documents/Fin_Tech/crypto-analysis/resources/features_btc.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Select model features and timeframe for train data\n",
    "X = features_btc[['AdrActCnt','CapMrktCurUSD','CapMVRVCur','HashRate','NVTAdj','TxCnt','TxTfrValAdjUSD','VelCur1yr','Crossover_Signal','BTC_Trends','RSI_14','ROC_14','MFI_14','TSI_13_25_13','TSIs_13_25_13','STOCHk_14_3_3','STOCHd_14_3_3']]\n",
    "X_train = X[:'2020']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Select prediction variable and timeframe for train data\n",
    "y = features_btc['Positive_Return']\n",
    "y_train = y[:'2020']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define timeframe for testing data\n",
    "X_test = X['2021':]\n",
    "y_test = y['2021':]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the StandardScaler instance\n",
    "scaler = StandardScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fit the Standard Scaler with the training data\n",
    "X_scaler = scaler.fit(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale the training data\n",
    "X_train_scaled = X_scaler.transform(X_train)\n",
    "X_test_scaled = X_scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best: 0.638822 using {'criterion': 'entropy', 'max_depth': 10, 'n_estimators': 500}\n"
     ]
    }
   ],
   "source": [
    "# Perform the grid search for the optimal set of tree parameters\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "n_estimators = [20,500]\n",
    "max_depth= [5,10]\n",
    "num_folds = 10\n",
    "scoring = 'accuracy'\n",
    "\n",
    "criterion = [\"gini\",\"entropy\"]\n",
    "param_grid = dict(n_estimators=n_estimators, max_depth=max_depth, \\\n",
    "  criterion = criterion )\n",
    "\n",
    "model = RandomForestClassifier(n_jobs=-1)\n",
    "\n",
    "kfold = KFold(n_splits=num_folds, random_state=None)\n",
    "grid = GridSearchCV(estimator=model, param_grid=param_grid, \\\n",
    "  scoring=scoring, cv=kfold)\n",
    "grid_result = grid.fit(X_train, y_train)\n",
    "print(\"Best: %f using %s\" % (grid_result.best_score_,\\\n",
    "  grid_result.best_params_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the model using optimized settings (also use if you want to run without optimizer above)\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "model = RandomForestClassifier(n_jobs=-1)\n",
    "model = RandomForestClassifier(criterion='entropy', max_depth=10, n_estimators=500, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fit the model\n",
    "model = model.fit(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Making predictions using the testing data\n",
    "predictions = model.predict(X_test_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculating the confusion matrix\n",
    "cm = confusion_matrix(y_test, predictions)\n",
    "cm_df = pd.DataFrame(cm, index=[\"Actual 0\", \"Actual 1\"], columns=[\"Predicted 0\", \"Predicted 1\"])\n",
    "\n",
    "# Calculating the accuracy score\n",
    "acc_score = accuracy_score(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix\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>Predicted 0</th>\n",
       "      <th>Predicted 1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Actual 0</th>\n",
       "      <td>59</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Actual 1</th>\n",
       "      <td>18</td>\n",
       "      <td>169</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Predicted 0  Predicted 1\n",
       "Actual 0           59          105\n",
       "Actual 1           18          169"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score : 0.6495726495726496\n",
      "Classification Report\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.77      0.36      0.49       164\n",
      "         1.0       0.62      0.90      0.73       187\n",
      "\n",
      "    accuracy                           0.65       351\n",
      "   macro avg       0.69      0.63      0.61       351\n",
      "weighted avg       0.69      0.65      0.62       351\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Displaying results\n",
    "print(\"Confusion Matrix\")\n",
    "display(cm_df)\n",
    "print(f\"Accuracy Score : {acc_score}\")\n",
    "print(\"Classification Report\")\n",
    "print(classification_report(y_test, predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>Positive_Return</th>\n",
       "      <th>Predicted Value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-11-28</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-11-29</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-11-30</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-01</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-02</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-03</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-04</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-05</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-06</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-07</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-08</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-09</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-10</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-11</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-12</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-13</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-14</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-15</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-16</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-12-17</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Positive_Return  Predicted Value\n",
       "time                                        \n",
       "2021-11-28              1.0              1.0\n",
       "2021-11-29              1.0              1.0\n",
       "2021-11-30              0.0              1.0\n",
       "2021-12-01              1.0              0.0\n",
       "2021-12-02              0.0              0.0\n",
       "2021-12-03              0.0              0.0\n",
       "2021-12-04              0.0              0.0\n",
       "2021-12-05              1.0              0.0\n",
       "2021-12-06              1.0              1.0\n",
       "2021-12-07              1.0              1.0\n",
       "2021-12-08              0.0              1.0\n",
       "2021-12-09              0.0              1.0\n",
       "2021-12-10              0.0              1.0\n",
       "2021-12-11              1.0              1.0\n",
       "2021-12-12              1.0              1.0\n",
       "2021-12-13              0.0              1.0\n",
       "2021-12-14              1.0              1.0\n",
       "2021-12-15              1.0              1.0\n",
       "2021-12-16              0.0              1.0\n",
       "2021-12-17              0.0              1.0"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Assemble actual y data (Y_test) with predicted y data (from just above) into two columns in a dataframe\n",
    "Results = y_test.to_frame()\n",
    "Results[\"Predicted Value\"] = predictions\n",
    "Results[-20:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0.18459320199153859, 'RSI_14'),\n",
       " (0.08836494626696598, 'STOCHd_14_3_3'),\n",
       " (0.08124538759910134, 'TSI_13_25_13'),\n",
       " (0.08039414833154292, 'STOCHk_14_3_3'),\n",
       " (0.06870978938128426, 'ROC_14'),\n",
       " (0.060278791586542355, 'MFI_14'),\n",
       " (0.05915355902935296, 'TSIs_13_25_13'),\n",
       " (0.04935631318043112, 'NVTAdj'),\n",
       " (0.04515476070536063, 'CapMVRVCur'),\n",
       " (0.0443575313095147, 'TxCnt'),\n",
       " (0.042776858150995925, 'AdrActCnt'),\n",
       " (0.03985229100903861, 'TxTfrValAdjUSD'),\n",
       " (0.03978078969312018, 'BTC_Trends'),\n",
       " (0.03943856408239912, 'VelCur1yr'),\n",
       " (0.03721873124951833, 'CapMrktCurUSD'),\n",
       " (0.03687335045447467, 'HashRate'),\n",
       " (0.0024509859788183447, 'Crossover_Signal')]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# List the features sorted in descending order by feature importance\n",
    "importances = model.feature_importances_\n",
    "sorted(zip(model.feature_importances_, X.columns), reverse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the features by importance\n",
    "importances_df = pd.DataFrame(sorted(zip(model.feature_importances_,    X.columns), reverse=True))\n",
    "importances_df.set_index(importances_df[1], inplace=True)\n",
    "importances_df.drop(columns=1, inplace=True)\n",
    "importances_df.rename(columns={0: 'Feature Importances'}, inplace=True)\n",
    "\n",
    "importances_sorted = importances_df.sort_values(by=\n",
    "'Feature Importances')\n",
    "\n",
    "importances_plot = importances_sorted.plot(kind='barh', color='lightgreen', title='Features Importances', legend=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# RSI is the most dominant feature which shows that BTC price is primarily determined by momentum."
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "1d1365bc7bea9ca53bbaaff2a498c5cc15b63a6fcc47bade387e2edfe7e7b068"
  },
  "kernelspec": {
   "display_name": "Python 3.7.9 64-bit ('algotrading': conda)",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}