AFML 5.4.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Fractional differentiate\n",
    "\n",
    "This notebook will cover exercise answer.\n",
    "\n",
    "* Exercise 5.6\n",
    "\n",
    "As we go along, there will be some explanations.\n",
    "\n",
    "Stationarity is a key concept in time-series, by now the idea itself has been demostrated in previous notebooks (Feat Importance).\n",
    "\n",
    "Most of the functions below can be found under research/Features and /Ensemble\n",
    "\n",
    "Contact: boyboi86@gmail.com"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num of CPU core:  4\n",
      "Machine info:  Windows-10-10.0.18362-SP0\n",
      "Python 3.7.4 (default, Aug  9 2019, 18:34:13) [MSC v.1915 64 bit (AMD64)]\n",
      "Numpy 1.17.3\n",
      "Pandas 1.0.3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Wei_X\\Anaconda3\\lib\\site-packages\\statsmodels\\tools\\_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  import pandas.util.testing as tm\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 1500x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import research as rs\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from statsmodels.tsa.stattools import coint\n",
    "\n",
    "p = print\n",
    "\n",
    "#pls take note of version\n",
    "#numba 0.49.1 #https://github.com/numba/numba/issues/4255\n",
    "#numpy 1.17.3\n",
    "#pandas 1.0.3\n",
    "#sklearn 0.21.3\n",
    "\n",
    "dollar = pd.read_csv('./research/Sample_data/dollar_bars.txt', \n",
    "                 sep=',', \n",
    "                 header=0, \n",
    "                 parse_dates = True, \n",
    "                 index_col=['date_time'])\n",
    "\n",
    "# For most part of the func we only use 'close'\n",
    "\n",
    "close = dollar['close'].to_frame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "ADF pVal: 0.04165 with critical value: 5%\n"
     ]
    }
   ],
   "source": [
    "ffd_series = close.apply(np.log).cumsum()\n",
    "ffd_series = rs.fracDiff_FFD(ffd_series, \n",
    "                    d = 1.99999889, \n",
    "                    thres=1e-5\n",
    "                   ).dropna()\n",
    "\n",
    "adf_pval = adfuller(ffd_series.squeeze(), \n",
    "                    maxlag = 1,\n",
    "                    regression='c', \n",
    "                    autolag=None)[1]\n",
    "\n",
    "p(\"\\nADF pVal: {0:.5f} with critical value: {1}%\".format(adf_pval, 5)) #confirm stationarity"
   ]
  },
  {
   "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>close</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date_time</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-02 09:35:57.204</th>\n",
       "      <td>-0.001693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02 12:59:42.176</th>\n",
       "      <td>0.000243</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02 14:19:33.847</th>\n",
       "      <td>0.001329</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02 14:33:39.311</th>\n",
       "      <td>0.000355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-02 14:42:28.315</th>\n",
       "      <td>0.000714</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            close\n",
       "date_time                        \n",
       "2015-01-02 09:35:57.204 -0.001693\n",
       "2015-01-02 12:59:42.176  0.000243\n",
       "2015-01-02 14:19:33.847  0.001329\n",
       "2015-01-02 14:33:39.311  0.000355\n",
       "2015-01-02 14:42:28.315  0.000714"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ffd_series.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>close</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date_time</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016-12-30 20:59:58.848</th>\n",
       "      <td>-0.101709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-30 21:00:00.352</th>\n",
       "      <td>-0.102272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-30 21:00:24.294</th>\n",
       "      <td>-0.101718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-30 21:03:03.027</th>\n",
       "      <td>-0.102393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-30 21:13:31.990</th>\n",
       "      <td>-0.102062</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            close\n",
       "date_time                        \n",
       "2016-12-30 20:59:58.848 -0.101709\n",
       "2016-12-30 21:00:00.352 -0.102272\n",
       "2016-12-30 21:00:24.294 -0.101718\n",
       "2016-12-30 21:03:03.027 -0.102393\n",
       "2016-12-30 21:13:31.990 -0.102062"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ffd_series.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "839\n"
     ]
    }
   ],
   "source": [
    "cs_event = rs.cs_filter(data = ffd_series, limit=(ffd_series.std() * 0.2))\n",
    "\n",
    "p(len(cs_event))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**\n",
    "\n",
    "This is the real reason why cs_filter was created, it was meant to be applied to stationary series.\n",
    "\n",
    "Our stationary series can act as our anchor for our mean-reversion strategy, that is why we are using cs_filter on this series to sample feature matrix.\n",
    "\n",
    "However, due to the lack of data sample/ resource, I cannot use 2 x std. (In fact, we would have been able to increase our F1 score in previous chapters, if we had enough data samples.)\n",
    "\n",
    "Based on my previous trials, when 1 x std is applied, the series itself will reduce heteroscedasticity even if it was not stationary.\n",
    "\n",
    "If you have the means to sample 5 - 10 years data, feel free to change to 2 x std."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>close</th>\n",
       "      <th>ffd_series</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-04 23:20:12.567</th>\n",
       "      <td>2040.75</td>\n",
       "      <td>-0.003825</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-07 10:55:23.194</th>\n",
       "      <td>2008.00</td>\n",
       "      <td>0.004166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-07 15:10:50.900</th>\n",
       "      <td>2014.50</td>\n",
       "      <td>-0.002360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 01:48:57.964</th>\n",
       "      <td>2037.25</td>\n",
       "      <td>0.006857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 05:47:32.006</th>\n",
       "      <td>2032.75</td>\n",
       "      <td>-0.003119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-11-28 01:31:48.252</th>\n",
       "      <td>2205.75</td>\n",
       "      <td>-0.099969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-04 23:32:49.403</th>\n",
       "      <td>2183.50</td>\n",
       "      <td>-0.102717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-05 02:06:52.025</th>\n",
       "      <td>2189.00</td>\n",
       "      <td>-0.096321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-08 12:46:05.346</th>\n",
       "      <td>2233.00</td>\n",
       "      <td>-0.101778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-14 20:09:50.768</th>\n",
       "      <td>2248.25</td>\n",
       "      <td>-0.103682</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>839 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           close  ffd_series\n",
       "2015-01-04 23:20:12.567  2040.75   -0.003825\n",
       "2015-01-07 10:55:23.194  2008.00    0.004166\n",
       "2015-01-07 15:10:50.900  2014.50   -0.002360\n",
       "2015-01-08 01:48:57.964  2037.25    0.006857\n",
       "2015-01-08 05:47:32.006  2032.75   -0.003119\n",
       "...                          ...         ...\n",
       "2016-11-28 01:31:48.252  2205.75   -0.099969\n",
       "2016-12-04 23:32:49.403  2183.50   -0.102717\n",
       "2016-12-05 02:06:52.025  2189.00   -0.096321\n",
       "2016-12-08 12:46:05.346  2233.00   -0.101778\n",
       "2016-12-14 20:09:50.768  2248.25   -0.103682\n",
       "\n",
       "[839 rows x 2 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#This portion requires combining events based on ffd_series with original dollar close using filtered index\n",
    "# since filter is based on ffd, therefore index matrix must also be based on filtered ffd\n",
    "df_mtx = pd.DataFrame(index = cs_event).assign(close = close,\n",
    "                                                ffd_series = ffd_series).drop_duplicates().dropna()\n",
    "df_mtx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_mtx['volatility'] = rs.vol(df_mtx.close, span0 = 50) #one of our features, since we do not have a side\n",
    "\n",
    "df_mtx.dropna(inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "vb = rs.vert_barrier(data = df_mtx.close, events = cs_event, period = 'days', freq = 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Wei_X\\Desktop\\Python\\research\\Labels\\triple_barrier_method.py:75: UserWarning: Data and events index shape must be same, reindex data to fit events\n",
      "  warnings.warn('Data and events index shape must be same, reindex data to fit events')\n",
      "C:\\Users\\Wei_X\\Desktop\\Python\\research\\Labels\\triple_barrier_method.py:112: UserWarning: Not Recommended: No side prediction provided\n",
      "  warnings.warn('Not Recommended: No side prediction provided')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[                                             t1                      sl  \\\n",
      "2015-01-07 15:10:50.900 2015-01-12 16:02:08.112                     NaT   \n",
      "2015-01-08 01:48:57.964 2015-01-13 09:38:58.103                     NaT   \n",
      "2015-01-08 05:47:32.006 2015-01-13 09:38:58.103                     NaT   \n",
      "2015-01-09 14:48:46.704 2015-01-14 19:14:20.771 2015-01-14 04:31:40.468   \n",
      "2015-01-12 14:36:34.243 2015-01-19 09:36:49.301 2015-01-14 19:14:20.771   \n",
      "...                                         ...                     ...   \n",
      "2015-09-01 19:46:17.742 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-01 22:05:09.069 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-02 00:20:04.277 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-02 02:09:17.333 2015-09-07 04:40:25.376                     NaT   \n",
      "2015-09-02 08:08:50.931 2015-09-08 01:27:51.915                     NaT   \n",
      "\n",
      "                                             pt  \n",
      "2015-01-07 15:10:50.900 2015-01-08 01:48:57.964  \n",
      "2015-01-08 01:48:57.964                     NaT  \n",
      "2015-01-08 05:47:32.006                     NaT  \n",
      "2015-01-09 14:48:46.704                     NaT  \n",
      "2015-01-12 14:36:34.243                     NaT  \n",
      "...                                         ...  \n",
      "2015-09-01 19:46:17.742                     NaT  \n",
      "2015-09-01 22:05:09.069                     NaT  \n",
      "2015-09-02 00:20:04.277                     NaT  \n",
      "2015-09-02 02:09:17.333                     NaT  \n",
      "2015-09-02 08:08:50.931                     NaT  \n",
      "\n",
      "[279 rows x 3 columns]] this out\n",
      "[                                             t1                      sl  \\\n",
      "2015-01-07 15:10:50.900 2015-01-12 16:02:08.112                     NaT   \n",
      "2015-01-08 01:48:57.964 2015-01-13 09:38:58.103                     NaT   \n",
      "2015-01-08 05:47:32.006 2015-01-13 09:38:58.103                     NaT   \n",
      "2015-01-09 14:48:46.704 2015-01-14 19:14:20.771 2015-01-14 04:31:40.468   \n",
      "2015-01-12 14:36:34.243 2015-01-19 09:36:49.301 2015-01-14 19:14:20.771   \n",
      "...                                         ...                     ...   \n",
      "2015-09-01 19:46:17.742 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-01 22:05:09.069 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-02 00:20:04.277 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-02 02:09:17.333 2015-09-07 04:40:25.376                     NaT   \n",
      "2015-09-02 08:08:50.931 2015-09-08 01:27:51.915                     NaT   \n",
      "\n",
      "                                             pt  \n",
      "2015-01-07 15:10:50.900 2015-01-08 01:48:57.964  \n",
      "2015-01-08 01:48:57.964                     NaT  \n",
      "2015-01-08 05:47:32.006                     NaT  \n",
      "2015-01-09 14:48:46.704                     NaT  \n",
      "2015-01-12 14:36:34.243                     NaT  \n",
      "...                                         ...  \n",
      "2015-09-01 19:46:17.742                     NaT  \n",
      "2015-09-01 22:05:09.069                     NaT  \n",
      "2015-09-02 00:20:04.277                     NaT  \n",
      "2015-09-02 02:09:17.333                     NaT  \n",
      "2015-09-02 08:08:50.931                     NaT  \n",
      "\n",
      "[279 rows x 3 columns],                                              t1  sl  pt\n",
      "2016-01-21 20:15:15.382 2016-01-27 00:04:03.812 NaT NaT\n",
      "2016-01-22 01:19:34.937 2016-01-27 13:41:54.306 NaT NaT\n",
      "2016-01-22 04:21:15.807 2016-01-27 13:41:54.306 NaT NaT\n",
      "2016-01-22 06:19:21.200 2016-01-27 13:41:54.306 NaT NaT\n",
      "2016-01-22 08:48:09.889 2016-01-27 13:41:54.306 NaT NaT\n",
      "...                                         ...  ..  ..\n",
      "2016-11-28 01:31:48.252 2016-12-04 23:32:49.403 NaT NaT\n",
      "2016-12-04 23:32:49.403 2016-12-14 20:09:50.768 NaT NaT\n",
      "2016-12-05 02:06:52.025 2016-12-14 20:09:50.768 NaT NaT\n",
      "2016-12-08 12:46:05.346 2016-12-14 20:09:50.768 NaT NaT\n",
      "2016-12-14 20:09:50.768                     NaT NaT NaT\n",
      "\n",
      "[279 rows x 3 columns]] this out\n",
      "[                                             t1                      sl  \\\n",
      "2015-01-07 15:10:50.900 2015-01-12 16:02:08.112                     NaT   \n",
      "2015-01-08 01:48:57.964 2015-01-13 09:38:58.103                     NaT   \n",
      "2015-01-08 05:47:32.006 2015-01-13 09:38:58.103                     NaT   \n",
      "2015-01-09 14:48:46.704 2015-01-14 19:14:20.771 2015-01-14 04:31:40.468   \n",
      "2015-01-12 14:36:34.243 2015-01-19 09:36:49.301 2015-01-14 19:14:20.771   \n",
      "...                                         ...                     ...   \n",
      "2015-09-01 19:46:17.742 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-01 22:05:09.069 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-02 00:20:04.277 2015-09-07 01:34:00.944                     NaT   \n",
      "2015-09-02 02:09:17.333 2015-09-07 04:40:25.376                     NaT   \n",
      "2015-09-02 08:08:50.931 2015-09-08 01:27:51.915                     NaT   \n",
      "\n",
      "                                             pt  \n",
      "2015-01-07 15:10:50.900 2015-01-08 01:48:57.964  \n",
      "2015-01-08 01:48:57.964                     NaT  \n",
      "2015-01-08 05:47:32.006                     NaT  \n",
      "2015-01-09 14:48:46.704                     NaT  \n",
      "2015-01-12 14:36:34.243                     NaT  \n",
      "...                                         ...  \n",
      "2015-09-01 19:46:17.742                     NaT  \n",
      "2015-09-01 22:05:09.069                     NaT  \n",
      "2015-09-02 00:20:04.277                     NaT  \n",
      "2015-09-02 02:09:17.333                     NaT  \n",
      "2015-09-02 08:08:50.931                     NaT  \n",
      "\n",
      "[279 rows x 3 columns],                                              t1  sl  pt\n",
      "2016-01-21 20:15:15.382 2016-01-27 00:04:03.812 NaT NaT\n",
      "2016-01-22 01:19:34.937 2016-01-27 13:41:54.306 NaT NaT\n",
      "2016-01-22 04:21:15.807 2016-01-27 13:41:54.306 NaT NaT\n",
      "2016-01-22 06:19:21.200 2016-01-27 13:41:54.306 NaT NaT\n",
      "2016-01-22 08:48:09.889 2016-01-27 13:41:54.306 NaT NaT\n",
      "...                                         ...  ..  ..\n",
      "2016-11-28 01:31:48.252 2016-12-04 23:32:49.403 NaT NaT\n",
      "2016-12-04 23:32:49.403 2016-12-14 20:09:50.768 NaT NaT\n",
      "2016-12-05 02:06:52.025 2016-12-14 20:09:50.768 NaT NaT\n",
      "2016-12-08 12:46:05.346 2016-12-14 20:09:50.768 NaT NaT\n",
      "2016-12-14 20:09:50.768                     NaT NaT NaT\n",
      "\n",
      "[279 rows x 3 columns],                                              t1  sl  pt\n",
      "2015-09-03 00:05:02.346 2015-09-08 01:27:51.915 NaT NaT\n",
      "2015-09-03 06:00:48.819 2015-09-08 12:08:06.840 NaT NaT\n",
      "2015-09-03 07:52:14.021 2015-09-08 12:08:06.840 NaT NaT\n",
      "2015-09-03 09:57:43.822 2015-09-08 12:08:06.840 NaT NaT\n",
      "2015-09-03 12:33:59.011 2015-09-09 03:06:59.837 NaT NaT\n",
      "...                                         ...  ..  ..\n",
      "2016-01-21 16:00:11.522 2016-01-27 00:04:03.812 NaT NaT\n",
      "2016-01-21 18:20:55.061 2016-01-27 00:04:03.812 NaT NaT\n",
      "2016-01-21 19:14:12.535 2016-01-27 00:04:03.812 NaT NaT\n",
      "2016-01-21 19:36:37.429 2016-01-27 00:04:03.812 NaT NaT\n",
      "2016-01-21 20:07:23.306 2016-01-27 00:04:03.812 NaT NaT\n",
      "\n",
      "[279 rows x 3 columns]] this out\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2020-05-29 16:48:39.895072 33.33% _pt_sl_t1 done after 0.11 minutes. Remaining 0.23 minutes.\r",
      "2020-05-29 16:48:39.922997 66.67% _pt_sl_t1 done after 0.12 minutes. Remaining 0.06 minutes.\r",
      "2020-05-29 16:48:39.968388 100.0% _pt_sl_t1 done after 0.12 minutes. Remaining 0.0 minutes.\n"
     ]
    }
   ],
   "source": [
    "# triple barrier events based on filter while data is also based on filtered index\n",
    "tb = rs.tri_barrier(data = df_mtx.close, \n",
    "                    events = cs_event, \n",
    "                    trgt = df_mtx['volatility'], \n",
    "                    min_req = 0.0002, \n",
    "                    num_threads = 3, \n",
    "                    ptSl= [2,2], #2x barriers\n",
    "                    t1 = vb, \n",
    "                    side = None)\n",
    "    \n"
   ]
  },
  {
   "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>t1</th>\n",
       "      <th>trgt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-07 15:10:50.900</th>\n",
       "      <td>2015-01-08 01:48:57.964</td>\n",
       "      <td>0.002252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 01:48:57.964</th>\n",
       "      <td>2015-01-13 09:38:58.103</td>\n",
       "      <td>0.007576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 05:47:32.006</th>\n",
       "      <td>2015-01-13 09:38:58.103</td>\n",
       "      <td>0.006867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-09 14:48:46.704</th>\n",
       "      <td>2015-01-14 04:31:40.468</td>\n",
       "      <td>0.009708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-12 14:36:34.243</th>\n",
       "      <td>2015-01-14 19:14:20.771</td>\n",
       "      <td>0.008815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-11-14 02:12:45.857</th>\n",
       "      <td>2016-11-28 01:31:48.252</td>\n",
       "      <td>0.022494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-11-28 01:31:48.252</th>\n",
       "      <td>2016-12-04 23:32:49.403</td>\n",
       "      <td>0.022526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-04 23:32:49.403</th>\n",
       "      <td>2016-12-14 20:09:50.768</td>\n",
       "      <td>0.022102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-05 02:06:52.025</th>\n",
       "      <td>2016-12-14 20:09:50.768</td>\n",
       "      <td>0.021669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-08 12:46:05.346</th>\n",
       "      <td>2016-12-14 20:09:50.768</td>\n",
       "      <td>0.021819</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>836 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             t1      trgt\n",
       "2015-01-07 15:10:50.900 2015-01-08 01:48:57.964  0.002252\n",
       "2015-01-08 01:48:57.964 2015-01-13 09:38:58.103  0.007576\n",
       "2015-01-08 05:47:32.006 2015-01-13 09:38:58.103  0.006867\n",
       "2015-01-09 14:48:46.704 2015-01-14 04:31:40.468  0.009708\n",
       "2015-01-12 14:36:34.243 2015-01-14 19:14:20.771  0.008815\n",
       "...                                         ...       ...\n",
       "2016-11-14 02:12:45.857 2016-11-28 01:31:48.252  0.022494\n",
       "2016-11-28 01:31:48.252 2016-12-04 23:32:49.403  0.022526\n",
       "2016-12-04 23:32:49.403 2016-12-14 20:09:50.768  0.022102\n",
       "2016-12-05 02:06:52.025 2016-12-14 20:09:50.768  0.021669\n",
       "2016-12-08 12:46:05.346 2016-12-14 20:09:50.768  0.021819\n",
       "\n",
       "[836 rows x 2 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "mlabel = rs.meta_label(data = df_mtx.close, \n",
    "                       events = tb, \n",
    "                       drop = 0.05) # because we do not have side, we need to drop rare labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 1.0    489\n",
       "-1.0    345\n",
       "Name: bin, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mlabel['bin'].value_counts() #834"
   ]
  },
  {
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>826</th>\n",
       "      <th>827</th>\n",
       "      <th>828</th>\n",
       "      <th>829</th>\n",
       "      <th>830</th>\n",
       "      <th>831</th>\n",
       "      <th>832</th>\n",
       "      <th>833</th>\n",
       "      <th>834</th>\n",
       "      <th>835</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-07 15:10:50.900</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 01:48:57.964</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 05:47:32.006</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-09 14:48:46.704</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-12 14:36:34.243</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <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>2016-11-28 01:31:48.252</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-04 23:32:49.403</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-05 02:06:52.025</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-08 12:46:05.346</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-14 20:09:50.768</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>837 rows × 836 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                         0    1    2    3    4    5    6    7    8    9    \\\n",
       "2015-01-07 15:10:50.900  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2015-01-08 01:48:57.964  1.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2015-01-08 05:47:32.006  0.0  1.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2015-01-09 14:48:46.704  0.0  1.0  1.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2015-01-12 14:36:34.243  0.0  1.0  1.0  1.0  1.0  0.0  0.0  0.0  0.0  0.0   \n",
       "...                      ...  ...  ...  ...  ...  ...  ...  ...  ...  ...   \n",
       "2016-11-28 01:31:48.252  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2016-12-04 23:32:49.403  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2016-12-05 02:06:52.025  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2016-12-08 12:46:05.346  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2016-12-14 20:09:50.768  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "                         ...  826  827  828  829  830  831  832  833  834  835  \n",
       "2015-01-07 15:10:50.900  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "2015-01-08 01:48:57.964  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "2015-01-08 05:47:32.006  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "2015-01-09 14:48:46.704  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "2015-01-12 14:36:34.243  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "...                      ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  \n",
       "2016-11-28 01:31:48.252  ...  1.0  1.0  1.0  1.0  1.0  1.0  1.0  0.0  0.0  0.0  \n",
       "2016-12-04 23:32:49.403  ...  0.0  0.0  0.0  0.0  0.0  0.0  1.0  1.0  0.0  0.0  \n",
       "2016-12-05 02:06:52.025  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  1.0  1.0  0.0  \n",
       "2016-12-08 12:46:05.346  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  1.0  1.0  1.0  \n",
       "2016-12-14 20:09:50.768  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0  1.0  1.0  1.0  \n",
       "\n",
       "[837 rows x 836 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "troo = rs.mp_idx_matrix(data = df_mtx.close, events = tb)\n",
    "#np.arange(troo.shape[1])\n",
    "troo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Quick Reference Guide/ Summary to func usage**\n",
    "\n",
    "1. Apply cs_filter using stationary series (FFD)\n",
    "\n",
    "2. Apply bband_as_side using close price series\n",
    "\n",
    "3. Create a new dataframe, using both outputs with index = cs_filter output.\n",
    "\n",
    "4. Apply func tri_barrier, Vert_barrier using new dataframe, before applying meta_label func.\n",
    "\n",
    "5. Feed these meta lables and features to random forest.\n",
    "\n",
    "Always include indicators, stationary series, volatility as part of your feature matrix if possible, since these are the features on how you derive meta-labels. \n",
    "\n",
    "As seen in previous exercises, ML models are smart enough to know what was the key parameters which meta-labels was labelled."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Wei_X\\Desktop\\Python\\research\\Sampling\\sample_unique.py:491: RuntimeWarning: invalid value encountered in log\n",
      "  rtn = np.log(data).diff().fillna(0) # Log-returns, so that they are additive\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>close</th>\n",
       "      <th>ffd_series</th>\n",
       "      <th>volatility</th>\n",
       "      <th>wr</th>\n",
       "      <th>wt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-07 15:10:50.900</th>\n",
       "      <td>2014.50</td>\n",
       "      <td>-0.002360</td>\n",
       "      <td>0.002252</td>\n",
       "      <td>3.004931</td>\n",
       "      <td>0.503742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 01:48:57.964</th>\n",
       "      <td>2037.25</td>\n",
       "      <td>0.006857</td>\n",
       "      <td>0.007576</td>\n",
       "      <td>2.548574</td>\n",
       "      <td>0.505364</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08 05:47:32.006</th>\n",
       "      <td>2032.75</td>\n",
       "      <td>-0.003119</td>\n",
       "      <td>0.006867</td>\n",
       "      <td>0.456357</td>\n",
       "      <td>0.506811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-09 14:48:46.704</th>\n",
       "      <td>2049.75</td>\n",
       "      <td>-0.003951</td>\n",
       "      <td>0.009708</td>\n",
       "      <td>0.651985</td>\n",
       "      <td>0.507852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-12 14:36:34.243</th>\n",
       "      <td>2030.50</td>\n",
       "      <td>-0.004798</td>\n",
       "      <td>0.008815</td>\n",
       "      <td>2.571428</td>\n",
       "      <td>0.508686</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>2016-11-14 02:12:45.857</th>\n",
       "      <td>2167.75</td>\n",
       "      <td>-0.097722</td>\n",
       "      <td>0.022494</td>\n",
       "      <td>1.194957</td>\n",
       "      <td>0.992436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-11-28 01:31:48.252</th>\n",
       "      <td>2205.75</td>\n",
       "      <td>-0.099969</td>\n",
       "      <td>0.022526</td>\n",
       "      <td>1.384323</td>\n",
       "      <td>0.994040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-04 23:32:49.403</th>\n",
       "      <td>2183.50</td>\n",
       "      <td>-0.102717</td>\n",
       "      <td>0.022102</td>\n",
       "      <td>1.510426</td>\n",
       "      <td>0.996258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-05 02:06:52.025</th>\n",
       "      <td>2189.00</td>\n",
       "      <td>-0.096321</td>\n",
       "      <td>0.021669</td>\n",
       "      <td>4.223331</td>\n",
       "      <td>0.998337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-08 12:46:05.346</th>\n",
       "      <td>2233.00</td>\n",
       "      <td>-0.101778</td>\n",
       "      <td>0.021819</td>\n",
       "      <td>3.550162</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>834 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           close  ffd_series  volatility        wr        wt\n",
       "2015-01-07 15:10:50.900  2014.50   -0.002360    0.002252  3.004931  0.503742\n",
       "2015-01-08 01:48:57.964  2037.25    0.006857    0.007576  2.548574  0.505364\n",
       "2015-01-08 05:47:32.006  2032.75   -0.003119    0.006867  0.456357  0.506811\n",
       "2015-01-09 14:48:46.704  2049.75   -0.003951    0.009708  0.651985  0.507852\n",
       "2015-01-12 14:36:34.243  2030.50   -0.004798    0.008815  2.571428  0.508686\n",
       "...                          ...         ...         ...       ...       ...\n",
       "2016-11-14 02:12:45.857  2167.75   -0.097722    0.022494  1.194957  0.992436\n",
       "2016-11-28 01:31:48.252  2205.75   -0.099969    0.022526  1.384323  0.994040\n",
       "2016-12-04 23:32:49.403  2183.50   -0.102717    0.022102  1.510426  0.996258\n",
       "2016-12-05 02:06:52.025  2189.00   -0.096321    0.021669  4.223331  0.998337\n",
       "2016-12-08 12:46:05.346  2233.00   -0.101778    0.021819  3.550162  1.000000\n",
       "\n",
       "[834 rows x 5 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X = df_mtx.reindex(mlabel.index)\n",
    "Z = tb.reindex(mlabel.index)\n",
    "y = mlabel['bin']\n",
    "\n",
    "X['wr'] = rs.wght_by_rtn(data = X, events = Z, num_threads = 1)\n",
    "X['wt'] = rs.wght_by_td(data = X, events = Z, num_threads = 1, td = 0.5)\n",
    "\n",
    "X #show example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2015-08-26 10:34:40.144    0.394145\n",
       "2015-08-25 11:37:16.214    0.488536\n",
       "2016-01-08 10:53:07.196    0.817534\n",
       "2015-09-23 13:11:44.559    0.736790\n",
       "2015-07-01 10:25:32.469    0.581751\n",
       "                             ...   \n",
       "2015-03-23 08:14:01.389    1.331041\n",
       "2015-05-29 10:32:41.319    0.509290\n",
       "2015-09-01 13:48:58.967    0.310314\n",
       "2015-12-22 09:01:07.048    0.770506\n",
       "2015-05-12 14:41:49.319    2.245505\n",
       "Name: wr, Length: 583, dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Split data into train and test sets, earlier on we combine sample weights with X because we need to match during fit\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42, shuffle=True, stratify=None)\n",
    "\n",
    "wght_td = X_train['wt']\n",
    "wght_rtn = X_train['wr']\n",
    "\n",
    "#weights are not feature so have to remove\n",
    "X_train = X_train.drop(['wr','wt'], axis = 1)\n",
    "X_test = X_test.drop(['wr','wt'], axis = 1)\n",
    "\n",
    "wght_rtn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_estimator = DecisionTreeClassifier()\n",
    "\n",
    "# common parameters to use, this ensemble method mimics monte-carlos as it randomly selects samples to form \"forest\"\n",
    "n_estimate = 10\n",
    "random_state = 42\n",
    "max_samples = 35"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ave uniqueness for grp sample 1: 0.647674\n",
      "Ave uniqueness for grp sample 2: 0.854561\n",
      "Ave uniqueness for grp sample 3: 0.849644\n",
      "Ave uniqueness for grp sample 4: 0.856805\n",
      "Ave uniqueness for grp sample 5: 0.867614\n",
      "Ave uniqueness for grp sample 6: 0.857027\n",
      "Ave uniqueness for grp sample 7: 0.804856\n",
      "Ave uniqueness for grp sample 8: 0.812000\n",
      "Ave uniqueness for grp sample 9: 0.904695\n",
      "Ave uniqueness for grp sample 10: 0.800973\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "BaggingClassifier(base_estimator=DecisionTreeClassifier(),\n",
       "                  data=2015-01-07 15:10:50.900    2014.50\n",
       "2015-01-08 01:48:57.964    2037.25\n",
       "2015-01-08 05:47:32.006    2032.75\n",
       "2015-01-09 14:48:46.704    2049.75\n",
       "2015-01-12 14:36:34.243    2030.50\n",
       "                            ...   \n",
       "2016-11-28 01:31:48.252    2205.75\n",
       "2016-12-04 23:32:49.403    2183.50\n",
       "2016-12-05 02:06:52.025    2189.00\n",
       "2016-12-08 12:46:05.346    2233.00\n",
       "2016-12-14 20:09:50.768    2...\n",
       "2015-01-12 14:36:34.243 2015-01-14 19:14:20.771  0.008815\n",
       "...                                         ...       ...\n",
       "2016-11-14 02:12:45.857 2016-11-28 01:31:48.252  0.022494\n",
       "2016-11-28 01:31:48.252 2016-12-04 23:32:49.403  0.022526\n",
       "2016-12-04 23:32:49.403 2016-12-14 20:09:50.768  0.022102\n",
       "2016-12-05 02:06:52.025 2016-12-14 20:09:50.768  0.021669\n",
       "2016-12-08 12:46:05.346 2016-12-14 20:09:50.768  0.021819\n",
       "\n",
       "[836 rows x 2 columns],\n",
       "                  max_samples=35, n_jobs=1, oob_score=True, random_state=42)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#warning, this is a heavy algo, will take up to 15 mins to complete with current parameter\n",
    "\n",
    "rtn_bag_cls=rs.BaggingClassifier(data = df_mtx.close,\n",
    "                                events = tb,\n",
    "                                base_estimator = base_estimator,\n",
    "                                n_estimators = n_estimate, # the number of trees as a estimate\n",
    "                                max_samples = max_samples, #if you put max_sample = 1.0, pls mentally get ready to wait\n",
    "                                max_features = 1.0,\n",
    "                                bootstrap = True, #must always be True\n",
    "                                bootstrap_features=False,\n",
    "                                oob_score = True, # You will need to have bootstrap = True\n",
    "                                warm_start = False,\n",
    "                                n_jobs = 1, # depending your machine. might become slower if you use all processors\n",
    "                                random_state = random_state, #seed figure\n",
    "                                verbose = 0)\n",
    "\n",
    "\n",
    "\n",
    "# Fit with absolute return weights\n",
    "rtn_bag_cls.fit(X_train, y_train, sample_weight = wght_rtn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Out-of-bag Score: 0.668954\n"
     ]
    }
   ],
   "source": [
    "p(\"Out-of-bag Score: {0:.6f}\".format(rtn_bag_cls.oob_score_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Classification Report\n",
      "=======================================================\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "        -1.0       0.58      0.70      0.63       102\n",
      "         1.0       0.76      0.66      0.71       149\n",
      "\n",
      "    accuracy                           0.67       251\n",
      "   macro avg       0.67      0.68      0.67       251\n",
      "weighted avg       0.69      0.67      0.68       251\n",
      "\n",
      "\n",
      " Confusion Matrix\n",
      "=======================================================\n",
      "[[TN, FP]\n",
      " [FN, TP]]\n",
      "\n",
      "[[71 31]\n",
      " [51 98]]\n",
      "\n",
      " Accuracy Score\n",
      "=======================================================\n",
      "0.6733067729083665\n",
      "\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"
    }
   ],
   "source": [
    "y_pred = rtn_bag_cls.predict(X_test)\n",
    "y_prob = rtn_bag_cls.predict_proba(X_test)[:,1] #we only want true positive probability\n",
    "\n",
    "rs.report_matrix(y_test, y_pred, y_prob)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ave uniqueness for grp sample 1: 0.647674\n",
      "Ave uniqueness for grp sample 2: 0.854561\n",
      "Ave uniqueness for grp sample 3: 0.849644\n",
      "Ave uniqueness for grp sample 4: 0.856805\n",
      "Ave uniqueness for grp sample 5: 0.867614\n",
      "Ave uniqueness for grp sample 6: 0.857027\n",
      "Ave uniqueness for grp sample 7: 0.804856\n",
      "Ave uniqueness for grp sample 8: 0.812000\n",
      "Ave uniqueness for grp sample 9: 0.904695\n",
      "Ave uniqueness for grp sample 10: 0.800973\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "BaggingClassifier(base_estimator=DecisionTreeClassifier(),\n",
       "                  data=2015-01-07 15:10:50.900    2014.50\n",
       "2015-01-08 01:48:57.964    2037.25\n",
       "2015-01-08 05:47:32.006    2032.75\n",
       "2015-01-09 14:48:46.704    2049.75\n",
       "2015-01-12 14:36:34.243    2030.50\n",
       "                            ...   \n",
       "2016-11-28 01:31:48.252    2205.75\n",
       "2016-12-04 23:32:49.403    2183.50\n",
       "2016-12-05 02:06:52.025    2189.00\n",
       "2016-12-08 12:46:05.346    2233.00\n",
       "2016-12-14 20:09:50.768    2...\n",
       "2015-01-12 14:36:34.243 2015-01-14 19:14:20.771  0.008815\n",
       "...                                         ...       ...\n",
       "2016-11-14 02:12:45.857 2016-11-28 01:31:48.252  0.022494\n",
       "2016-11-28 01:31:48.252 2016-12-04 23:32:49.403  0.022526\n",
       "2016-12-04 23:32:49.403 2016-12-14 20:09:50.768  0.022102\n",
       "2016-12-05 02:06:52.025 2016-12-14 20:09:50.768  0.021669\n",
       "2016-12-08 12:46:05.346 2016-12-14 20:09:50.768  0.021819\n",
       "\n",
       "[836 rows x 2 columns],\n",
       "                  max_samples=35, n_jobs=1, oob_score=True, random_state=42)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "td_bag_cls=rs.BaggingClassifier(data = df_mtx.close,\n",
    "                                events = tb,\n",
    "                                base_estimator = base_estimator,\n",
    "                                n_estimators = n_estimate, # the number of trees as a estimate\n",
    "                                max_samples = max_samples, #if you put max_sample = 1.0, pls mentally get ready to wait\n",
    "                                max_features = 1.0,\n",
    "                                bootstrap = True, #must always be True\n",
    "                                bootstrap_features=False,\n",
    "                                oob_score = True, # You will need to have bootstrap = True\n",
    "                                warm_start = False,\n",
    "                                n_jobs = 1, # depending your machine. might become slower if you use all processors\n",
    "                                random_state = random_state, #seed figure\n",
    "                                verbose = 0)\n",
    "\n",
    "\n",
    "\n",
    "# Fit sample weight by time-decay\n",
    "td_bag_cls.fit(X_train, y_train, sample_weight = wght_td)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Out-of-bag Score: 0.662093\n"
     ]
    }
   ],
   "source": [
    "p(\"Out-of-bag Score: {0:.6f}\".format(td_bag_cls.oob_score_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Classification Report\n",
      "=======================================================\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "        -1.0       0.57      0.61      0.59       102\n",
      "         1.0       0.72      0.68      0.70       149\n",
      "\n",
      "    accuracy                           0.65       251\n",
      "   macro avg       0.64      0.65      0.64       251\n",
      "weighted avg       0.66      0.65      0.65       251\n",
      "\n",
      "\n",
      " Confusion Matrix\n",
      "=======================================================\n",
      "[[TN, FP]\n",
      " [FN, TP]]\n",
      "\n",
      "[[ 62  40]\n",
      " [ 47 102]]\n",
      "\n",
      " Accuracy Score\n",
      "=======================================================\n",
      "0.6533864541832669\n",
      "\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"
    }
   ],
   "source": [
    "y_pred1 = td_bag_cls.predict(X_test)\n",
    "y_prob1 = td_bag_cls.predict_proba(X_test)[:,1] #we only want true positive probability\n",
    "\n",
    "rs.report_matrix(y_test, y_pred1, y_prob1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Out-of-Bag score\n",
    "\n",
    "Original sklearn bagging method: Out-of-bag Score: 0.662093\n",
    "\n",
    "Sequential bootstrap bagging method:\n",
    "        \n",
    "1. without sample weight: Out-of-bag Score: 0.655232\n",
    "2. with abs return wght: Out-of-bag Score: 0.668954\n",
    "3. with time-decay wght: Out-of-bag Score: 0.662093\n",
    "\n",
    "Comparing the original sklearn bagging ensemble using decision tree classifier, our OOB score is lower.\n",
    "\n",
    "Once we included sample weights by either absolute returns or time-decay, our OOB score becomes higher or the equal with original sklearn OOB score.\n",
    "\n",
    "In fact, overall accuracy and f1-scores also increased with a better ROC graph. All of it with a less bias result.\n",
    "\n",
    "**Conclusion**\n",
    "\n",
    "Whenever we use bagging, we should try and include sample weights so that ML classification will improve the overall results.\n",
    "\n",
    "**Note**\n",
    "\n",
    "If you are wondering why I can get identical average uniqueness for all sample grps, just make sure you set random state."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}