{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"toc": true
},
"source": [
"Table of Contents
\n",
""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Labeling and MetaLabeling"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Overview\n",
"\n",
"In this chapter of the book AFML, De Prado introduces several novel techniques for labeling returns for the purposes of supervised machine learning. \n",
"\n",
"First he identifies the typical issues of fixed-time horizon labeling methods - primarily that it is easy to mislabel a return due to dynamic nature of volatility throughout a trading period.\n",
"\n",
"More importantly he addresses a major overlooked aspect of the financial literature. He emphasizes that every investment strategy makes use of stop-loss limits of some kind, whether those are enforced by a margin call, risk department or self-imposed. He highlights how unrealistic it is to test/implement/propagate a strategy that profits from positions that would have been stopped out. \n",
"\n",
"> That virtually no publication accounts for that when labeling observations tells you something about the current state of financial literature.\n",
">\n",
"> -De Prado, \"Advances in Financial Machine Learning\", pg.44\n",
"\n",
"He also introduces a technique called metalabeling, which is used to augment a strategy by improving recall while also reducing the likelihood of overfitting."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:13.812868Z",
"start_time": "2019-03-01T17:54:08.567716Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"2019-03-01T10:54:08-07:00\n",
"\n",
"CPython 3.7.2\n",
"IPython 6.5.0\n",
"\n",
"compiler : GCC 7.3.0\n",
"system : Linux\n",
"release : 4.19.11-041911-generic\n",
"machine : x86_64\n",
"processor : x86_64\n",
"CPU cores : 12\n",
"interpreter: 64bit\n",
"\n",
"pandas 0.23.4\n",
"pandas_datareader 0.7.0\n",
"dask 1.0.0\n",
"numpy 1.15.4\n",
"sklearn 0.20.2\n",
"statsmodels 0.9.0\n",
"scipy 1.1.0\n",
"ffn (0, 3, 4)\n",
"matplotlib 3.0.2\n",
"seaborn 0.9.0\n"
]
}
],
"source": [
"%load_ext watermark\n",
"%watermark\n",
"\n",
"%load_ext autoreload\n",
"%autoreload 2\n",
"\n",
"# import standard libs\n",
"from IPython.display import display\n",
"from IPython.core.debugger import set_trace as bp\n",
"from pathlib import PurePath, Path\n",
"import sys\n",
"import time\n",
"from collections import OrderedDict as od\n",
"import re\n",
"import os\n",
"import json\n",
"\n",
"# import python scientific stack\n",
"import pandas as pd\n",
"import pandas_datareader.data as web\n",
"pd.set_option('display.max_rows', 100)\n",
"from dask import dataframe as dd\n",
"from dask.diagnostics import ProgressBar\n",
"from multiprocessing import cpu_count\n",
"pbar = ProgressBar()\n",
"pbar.register()\n",
"import numpy as np\n",
"import scipy.stats as stats\n",
"import statsmodels.api as sm\n",
"from numba import jit\n",
"import math\n",
"import ffn\n",
"\n",
"# import visual tools\n",
"import matplotlib as mpl\n",
"import matplotlib.pyplot as plt\n",
"import matplotlib.gridspec as gridspec\n",
"%matplotlib inline\n",
"import seaborn as sns\n",
"\n",
"plt.style.use('seaborn-talk')\n",
"plt.style.use('bmh')\n",
"#plt.rcParams['font.family'] = 'DejaVu Sans Mono'\n",
"plt.rcParams['font.size'] = 9.5\n",
"plt.rcParams['font.weight'] = 'medium'\n",
"plt.rcParams['figure.figsize'] = 10,7\n",
"blue, green, red, purple, gold, teal = sns.color_palette('colorblind', 6)\n",
"\n",
"# import util libs\n",
"from tqdm import tqdm, tqdm_notebook\n",
"import warnings\n",
"warnings.filterwarnings(\"ignore\")\n",
"import missingno as msno\n",
"from src.utils.utils import *\n",
"import src.features.bars as brs\n",
"import src.features.snippets as snp\n",
"\n",
"RANDOM_STATE = 777\n",
"\n",
"pdir = get_relative_project_dir('Adv_Fin_ML_Exercises', partial=False)\n",
"data_dir = pdir / 'data'\n",
"\n",
"print()\n",
"%watermark -p pandas,pandas_datareader,dask,numpy,sklearn,statsmodels,scipy,ffn,matplotlib,seaborn"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Code Snippets\n",
"\n",
"Below I reproduce all the relevant code snippets found in the book that are necessary to work through the excercises found at the end of chapter 3."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Symmetric CUSUM Filter [2.5.2.1]"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:14.136573Z",
"start_time": "2019-03-01T17:54:13.814830Z"
}
},
"outputs": [],
"source": [
"def getTEvents(gRaw, h):\n",
" tEvents, sPos, sNeg = [], 0, 0\n",
" diff = np.log(gRaw).diff().dropna()\n",
" for i in tqdm(diff.index[1:]):\n",
" try:\n",
" pos, neg = float(sPos+diff.loc[i]), float(sNeg+diff.loc[i])\n",
" except Exception as e:\n",
" print(e)\n",
" print(sPos+diff.loc[i], type(sPos+diff.loc[i]))\n",
" print(sNeg+diff.loc[i], type(sNeg+diff.loc[i]))\n",
" break\n",
" sPos, sNeg=max(0., pos), min(0., neg)\n",
" if sNeg<-h:\n",
" sNeg=0;tEvents.append(i)\n",
" elif sPos>h:\n",
" sPos=0;tEvents.append(i)\n",
" return pd.DatetimeIndex(tEvents)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Daily Volatility Estimator [3.1]"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:14.430421Z",
"start_time": "2019-03-01T17:54:14.138108Z"
}
},
"outputs": [],
"source": [
"def getDailyVol(close,span0=100):\n",
" # daily vol reindexed to close\n",
" df0=close.index.searchsorted(close.index-pd.Timedelta(days=1))\n",
" df0=df0[df0>0] \n",
" df0=(pd.Series(close.index[df0-1], \n",
" index=close.index[close.shape[0]-df0.shape[0]:])) \n",
" try:\n",
" df0=close.loc[df0.index]/close.loc[df0.values].values-1 # daily rets\n",
" except Exception as e:\n",
" print(f'error: {e}\\nplease confirm no duplicate indices')\n",
" df0=df0.ewm(span=span0).std().rename('dailyVol')\n",
" return df0"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Triple-Barrier Labeling Method [3.2]"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:14.746120Z",
"start_time": "2019-03-01T17:54:14.432530Z"
}
},
"outputs": [],
"source": [
"def applyPtSlOnT1(close,events,ptSl,molecule):\n",
" # apply stop loss/profit taking, if it takes place before t1 (end of event)\n",
" events_=events.loc[molecule]\n",
" out=events_[['t1']].copy(deep=True)\n",
" if ptSl[0]>0: pt=ptSl[0]*events_['trgt']\n",
" else: pt=pd.Series(index=events.index) # NaNs\n",
" if ptSl[1]>0: sl=-ptSl[1]*events_['trgt']\n",
" else: sl=pd.Series(index=events.index) # NaNs\n",
" for loc,t1 in events_['t1'].fillna(close.index[-1]).iteritems():\n",
" df0=close[loc:t1] # path prices\n",
" df0=(df0/close[loc]-1)*events_.at[loc,'side'] # path returns\n",
" out.loc[loc,'sl']=df0[df0pt[loc]].index.min() # earliest profit taking\n",
" return out"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Gettting Time of First Touch (getEvents) [3.3], [3.6]"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:15.090397Z",
"start_time": "2019-03-01T17:54:14.748256Z"
}
},
"outputs": [],
"source": [
"def getEvents(close, tEvents, ptSl, trgt, minRet, numThreads, t1=False, side=None):\n",
" #1) get target\n",
" trgt=trgt.loc[tEvents]\n",
" trgt=trgt[trgt>minRet] # minRet\n",
" #2) get t1 (max holding period)\n",
" if t1 is False:t1=pd.Series(pd.NaT, index=tEvents)\n",
" #3) form events object, apply stop loss on t1\n",
" if side is None:side_,ptSl_=pd.Series(1.,index=trgt.index), [ptSl[0],ptSl[0]]\n",
" else: side_,ptSl_=side.loc[trgt.index],ptSl[:2]\n",
" events=(pd.concat({'t1':t1,'trgt':trgt,'side':side_}, axis=1)\n",
" .dropna(subset=['trgt']))\n",
" df0=mpPandasObj(func=applyPtSlOnT1,pdObj=('molecule',events.index),\n",
" numThreads=numThreads,close=close,events=events,\n",
" ptSl=ptSl_)\n",
" events['t1']=df0.dropna(how='all').min(axis=1) # pd.min ignores nan\n",
" if side is None:events=events.drop('side',axis=1)\n",
" return events"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Adding Vertical Barrier [3.4]"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:15.400850Z",
"start_time": "2019-03-01T17:54:15.092280Z"
}
},
"outputs": [],
"source": [
"def addVerticalBarrier(tEvents, close, numDays=1):\n",
" t1=close.index.searchsorted(tEvents+pd.Timedelta(days=numDays))\n",
" t1=t1[t1minPct or df0.shape[0]<3:break\n",
" print('dropped label: ', df0.argmin(),df0.min())\n",
" events=events[events['bin']!=df0.argmin()]\n",
" return events"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Linear Partitions [20.4.1]"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:16.614458Z",
"start_time": "2019-03-01T17:54:16.306320Z"
}
},
"outputs": [],
"source": [
"def linParts(numAtoms,numThreads):\n",
" # partition of atoms with a single loop\n",
" parts=np.linspace(0,numAtoms,min(numThreads,numAtoms)+1)\n",
" parts=np.ceil(parts).astype(int)\n",
" return parts"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:16.943498Z",
"start_time": "2019-03-01T17:54:16.616483Z"
}
},
"outputs": [],
"source": [
"def nestedParts(numAtoms,numThreads,upperTriang=False):\n",
" # partition of atoms with an inner loop\n",
" parts,numThreads_=[0],min(numThreads,numAtoms)\n",
" for num in range(numThreads_):\n",
" part=1+4*(parts[-1]**2+parts[-1]+numAtoms*(numAtoms+1.)/numThreads_)\n",
" part=(-1+part**.5)/2.\n",
" parts.append(part)\n",
" parts=np.round(parts).astype(int)\n",
" if upperTriang: # the first rows are heaviest\n",
" parts=np.cumsum(np.diff(parts)[::-1])\n",
" parts=np.append(np.array([0]),parts)\n",
" return parts"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### multiprocessing snippet [20.7]"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:17.274840Z",
"start_time": "2019-03-01T17:54:16.945515Z"
}
},
"outputs": [],
"source": [
"def mpPandasObj(func,pdObj,numThreads=24,mpBatches=1,linMols=True,**kargs):\n",
" '''\n",
" Parallelize jobs, return a dataframe or series\n",
" + func: function to be parallelized. Returns a DataFrame\n",
" + pdObj[0]: Name of argument used to pass the molecule\n",
" + pdObj[1]: List of atoms that will be grouped into molecules\n",
" + kwds: any other argument needed by func\n",
" \n",
" Example: df1=mpPandasObj(func,('molecule',df0.index),24,**kwds)\n",
" '''\n",
" import pandas as pd\n",
" #if linMols:parts=linParts(len(argList[1]),numThreads*mpBatches)\n",
" #else:parts=nestedParts(len(argList[1]),numThreads*mpBatches)\n",
" if linMols:parts=linParts(len(pdObj[1]),numThreads*mpBatches)\n",
" else:parts=nestedParts(len(pdObj[1]),numThreads*mpBatches)\n",
" \n",
" jobs=[]\n",
" for i in range(1,len(parts)):\n",
" job={pdObj[0]:pdObj[1][parts[i-1]:parts[i]],'func':func}\n",
" job.update(kargs)\n",
" jobs.append(job)\n",
" if numThreads==1:out=processJobs_(jobs)\n",
" else: out=processJobs(jobs,numThreads=numThreads)\n",
" if isinstance(out[0],pd.DataFrame):df0=pd.DataFrame()\n",
" elif isinstance(out[0],pd.Series):df0=pd.Series()\n",
" else:return out\n",
" for i in out:df0=df0.append(i)\n",
" df0=df0.sort_index()\n",
" return df0"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### single-thread execution for debugging [20.8]"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:17.601019Z",
"start_time": "2019-03-01T17:54:17.277000Z"
}
},
"outputs": [],
"source": [
"def processJobs_(jobs):\n",
" # Run jobs sequentially, for debugging\n",
" out=[]\n",
" for job in jobs:\n",
" out_=expandCall(job)\n",
" out.append(out_)\n",
" return out"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Example of async call to multiprocessing lib [20.9]"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:17.934204Z",
"start_time": "2019-03-01T17:54:17.602756Z"
}
},
"outputs": [],
"source": [
"import multiprocessing as mp\n",
"import datetime as dt\n",
"\n",
"#________________________________\n",
"def reportProgress(jobNum,numJobs,time0,task):\n",
" # Report progress as asynch jobs are completed\n",
" msg=[float(jobNum)/numJobs, (time.time()-time0)/60.]\n",
" msg.append(msg[1]*(1/msg[0]-1))\n",
" timeStamp=str(dt.datetime.fromtimestamp(time.time()))\n",
" msg=timeStamp+' '+str(round(msg[0]*100,2))+'% '+task+' done after '+ \\\n",
" str(round(msg[1],2))+' minutes. Remaining '+str(round(msg[2],2))+' minutes.'\n",
" if jobNum\n",
"DatetimeIndex: 941297 entries, 2009-09-28 09:30:00 to 2018-02-26 18:30:00\n",
"Data columns (total 6 columns):\n",
"price 941297 non-null float64\n",
"bid 941297 non-null float64\n",
"ask 941297 non-null float64\n",
"size 941297 non-null float64\n",
"v 941297 non-null float64\n",
"dv 941297 non-null float64\n",
"dtypes: float64(6)\n",
"memory usage: 50.3 MB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"infp = PurePath(data_dir/'processed'/'IVE_dollarValue_resampled_1s.parquet')\n",
"df = pd.read_parquet(infp)\n",
"cprint(df)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## [3.1] Form Dollar Bars"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:19.835164Z",
"start_time": "2019-03-01T17:54:19.142594Z"
}
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 941297/941297 [00:00<00:00, 2874489.73it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" price bid ask size \\\n",
"2018-02-26 15:31:06 115.29 115.280000 115.290000 2022.000000 \n",
"2018-02-26 15:40:15 115.41 115.400000 115.410000 723.000000 \n",
"2018-02-26 15:49:42 115.20 115.176667 115.186667 4487.166667 \n",
"2018-02-26 15:59:04 115.27 115.260000 115.270000 300.000000 \n",
"2018-02-26 16:16:14 115.30 114.720000 115.620000 778677.000000 \n",
"\n",
" v dv \n",
"2018-02-26 15:31:06 2022.000000 2.331164e+05 \n",
"2018-02-26 15:40:15 723.000000 8.344143e+04 \n",
"2018-02-26 15:49:42 4487.166667 5.171190e+05 \n",
"2018-02-26 15:59:04 300.000000 3.458100e+04 \n",
"2018-02-26 16:16:14 778677.000000 8.978146e+07 \n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 30860 entries, 2009-09-28 09:53:49 to 2018-02-26 16:16:14\n",
"Data columns (total 6 columns):\n",
"price 30860 non-null float64\n",
"bid 30860 non-null float64\n",
"ask 30860 non-null float64\n",
"size 30860 non-null float64\n",
"v 30860 non-null float64\n",
"dv 30860 non-null float64\n",
"dtypes: float64(6)\n",
"memory usage: 1.6 MB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\n"
]
}
],
"source": [
"dbars = brs.dollar_bar_df(df, 'dv', 1_000_000).drop_duplicates().dropna()\n",
"cprint(dbars)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (a) Run cusum filter with threshold equal to std dev of daily returns"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:20.241528Z",
"start_time": "2019-03-01T17:54:19.837138Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" dailyVol\n",
"2018-02-26 15:31:06 0.006852\n",
"2018-02-26 15:40:15 0.006893\n",
"2018-02-26 15:49:42 0.006889\n",
"2018-02-26 15:59:04 0.006894\n",
"2018-02-26 16:16:14 0.006902\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 30843 entries, 2009-09-29 10:03:18 to 2018-02-26 16:16:14\n",
"Data columns (total 1 columns):\n",
"dailyVol 30842 non-null float64\n",
"dtypes: float64(1)\n",
"memory usage: 481.9 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"close = dbars.price.copy()\n",
"dailyVol = getDailyVol(close)\n",
"cprint(dailyVol.to_frame())"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:20.789871Z",
"start_time": "2019-03-01T17:54:20.243279Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"f,ax=plt.subplots()\n",
"dailyVol.plot(ax=ax)\n",
"ax.axhline(dailyVol.mean(),ls='--',color='r')"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:23.002687Z",
"start_time": "2019-03-01T17:54:20.791410Z"
}
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 30858/30858 [00:01<00:00, 16495.86it/s]\n"
]
},
{
"data": {
"text/plain": [
"DatetimeIndex(['2009-09-29 09:33:01', '2009-09-30 09:45:21',\n",
" '2009-09-30 13:31:12', '2009-10-01 09:43:58',\n",
" '2009-10-01 11:12:07', '2009-10-02 09:44:14',\n",
" '2009-10-02 10:35:05', '2009-10-05 09:51:42',\n",
" '2009-10-05 14:55:48', '2009-10-06 09:29:52',\n",
" ...\n",
" '2018-02-16 14:23:51', '2018-02-20 09:30:00',\n",
" '2018-02-20 15:21:07', '2018-02-21 14:04:12',\n",
" '2018-02-21 15:12:30', '2018-02-22 12:18:21',\n",
" '2018-02-22 14:56:14', '2018-02-23 11:37:32',\n",
" '2018-02-23 15:58:39', '2018-02-26 13:06:34'],\n",
" dtype='datetime64[ns]', length=2278, freq=None)"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"tEvents = getTEvents(close,h=dailyVol.mean())\n",
"tEvents"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (b) Add vertical barrier"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:23.325936Z",
"start_time": "2019-03-01T17:54:23.005620Z"
},
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"2009-09-29 09:33:01 2009-09-30 09:45:21\n",
"2009-09-30 09:45:21 2009-10-01 10:00:48\n",
"2009-09-30 13:31:12 2009-10-01 13:33:25\n",
"2009-10-01 09:43:58 2009-10-02 09:44:14\n",
"2009-10-01 11:12:07 2009-10-02 11:50:21\n",
"2009-10-02 09:44:14 2009-10-05 09:51:42\n",
"2009-10-02 10:35:05 2009-10-05 09:51:42\n",
"2009-10-05 09:51:42 2009-10-06 10:16:02\n",
"2009-10-05 14:55:48 2009-10-06 15:35:49\n",
"2009-10-06 09:29:52 2009-10-07 09:47:16\n",
"2009-10-06 11:32:02 2009-10-07 11:48:22\n",
"2009-10-06 14:07:37 2009-10-07 14:22:36\n",
"2009-10-08 09:29:51 2009-10-09 09:31:12\n",
"2009-10-12 09:31:02 2009-10-13 09:47:54\n",
"2009-10-13 10:52:10 2009-10-14 11:12:03\n",
"2009-10-14 09:29:52 2009-10-15 09:37:24\n",
"2009-10-14 15:30:48 2009-10-15 15:57:25\n",
"2009-10-16 09:55:03 2009-10-19 09:37:41\n",
"2009-10-16 15:40:15 2009-10-19 09:37:41\n",
"2009-10-19 11:39:38 2009-10-20 11:50:28\n",
"2009-10-20 11:50:28 2009-10-21 12:44:38\n",
"2009-10-21 10:11:57 2009-10-22 10:47:06\n",
"2009-10-21 15:32:09 2009-10-22 15:49:30\n",
"2009-10-22 09:55:51 2009-10-23 10:03:53\n",
"2009-10-22 14:33:52 2009-10-23 14:49:39\n",
"2009-10-23 10:57:52 2009-10-26 09:52:17\n",
"2009-10-26 09:52:17 2009-10-27 09:57:46\n",
"2009-10-26 11:32:02 2009-10-27 12:04:42\n",
"2009-10-26 11:59:14 2009-10-27 12:04:42\n",
"2009-10-27 13:37:35 2009-10-28 14:04:15\n",
"2009-10-28 10:00:16 2009-10-29 10:00:59\n",
"2009-10-28 14:41:52 2009-10-29 15:00:53\n",
"2009-10-29 09:32:01 2009-10-30 09:43:02\n",
"2009-10-29 13:40:22 2009-10-30 13:54:51\n",
"2009-10-30 09:58:07 2009-11-02 09:51:15\n",
"2009-10-30 11:51:20 2009-11-02 09:51:15\n",
"2009-10-30 12:57:50 2009-11-02 09:51:15\n",
"2009-10-30 15:06:13 2009-11-02 09:51:15\n",
"2009-10-30 15:44:12 2009-11-02 09:51:15\n",
"2009-11-02 10:17:36 2009-11-03 10:42:33\n",
"2009-11-02 12:23:50 2009-11-03 12:24:26\n",
"2009-11-02 12:58:06 2009-11-03 13:10:26\n",
"2009-11-02 14:07:16 2009-11-03 14:22:31\n",
"2009-11-02 14:55:04 2009-11-03 15:18:16\n",
"2009-11-03 14:22:31 2009-11-04 14:41:42\n",
"2009-11-04 09:34:15 2009-11-05 09:59:36\n",
"2009-11-04 15:46:56 2009-11-05 16:09:46\n",
"2009-11-05 09:59:36 2009-11-06 10:06:33\n",
"2009-11-05 16:09:46 2009-11-09 09:54:17\n",
"2009-11-09 09:54:17 2009-11-10 10:09:52\n",
" ... \n",
"2018-02-06 09:36:34 2018-02-07 09:43:03\n",
"2018-02-06 09:58:38 2018-02-07 10:04:28\n",
"2018-02-06 10:18:08 2018-02-07 10:22:20\n",
"2018-02-06 10:38:41 2018-02-07 10:39:35\n",
"2018-02-06 11:35:33 2018-02-07 11:46:44\n",
"2018-02-06 11:53:57 2018-02-07 11:57:50\n",
"2018-02-06 12:32:24 2018-02-07 12:42:28\n",
"2018-02-06 13:04:03 2018-02-07 13:08:44\n",
"2018-02-06 14:19:57 2018-02-07 14:20:37\n",
"2018-02-06 14:49:56 2018-02-07 14:53:22\n",
"2018-02-06 15:05:41 2018-02-07 15:11:44\n",
"2018-02-06 15:42:53 2018-02-07 15:47:02\n",
"2018-02-07 09:43:03 2018-02-08 09:57:38\n",
"2018-02-07 11:15:27 2018-02-08 11:18:31\n",
"2018-02-07 13:16:25 2018-02-08 13:17:26\n",
"2018-02-07 15:28:09 2018-02-08 15:33:11\n",
"2018-02-07 15:58:58 2018-02-08 15:59:48\n",
"2018-02-08 10:33:27 2018-02-09 10:41:40\n",
"2018-02-08 12:29:28 2018-02-09 12:40:46\n",
"2018-02-08 13:45:14 2018-02-09 13:52:34\n",
"2018-02-08 15:07:57 2018-02-09 15:09:17\n",
"2018-02-08 15:45:50 2018-02-09 15:47:50\n",
"2018-02-09 09:30:00 2018-02-12 09:30:00\n",
"2018-02-09 10:41:40 2018-02-12 09:30:00\n",
"2018-02-09 12:05:08 2018-02-12 09:30:00\n",
"2018-02-09 13:27:21 2018-02-12 09:30:00\n",
"2018-02-09 13:52:34 2018-02-12 09:30:00\n",
"2018-02-09 14:11:06 2018-02-12 09:30:00\n",
"2018-02-09 15:05:41 2018-02-12 09:30:00\n",
"2018-02-09 15:29:15 2018-02-12 09:30:00\n",
"2018-02-09 15:47:50 2018-02-12 09:30:00\n",
"2018-02-12 09:30:00 2018-02-13 09:30:00\n",
"2018-02-12 10:25:02 2018-02-13 10:36:48\n",
"2018-02-12 12:12:51 2018-02-13 12:34:24\n",
"2018-02-13 09:30:00 2018-02-14 09:30:00\n",
"2018-02-13 13:43:37 2018-02-14 13:53:59\n",
"2018-02-14 10:30:48 2018-02-15 10:42:27\n",
"2018-02-14 13:36:02 2018-02-15 13:42:09\n",
"2018-02-15 09:31:56 2018-02-16 09:42:36\n",
"2018-02-15 14:05:41 2018-02-16 14:15:08\n",
"2018-02-16 11:11:50 2018-02-20 09:30:00\n",
"2018-02-16 14:23:51 2018-02-20 09:30:00\n",
"2018-02-20 09:30:00 2018-02-21 09:34:28\n",
"2018-02-20 15:21:07 2018-02-21 15:22:14\n",
"2018-02-21 14:04:12 2018-02-22 14:20:25\n",
"2018-02-21 15:12:30 2018-02-22 15:16:50\n",
"2018-02-22 12:18:21 2018-02-23 12:30:16\n",
"2018-02-22 14:56:14 2018-02-23 15:02:21\n",
"2018-02-23 11:37:32 2018-02-26 09:30:00\n",
"2018-02-23 15:58:39 2018-02-26 09:30:00\n",
"Length: 2277, dtype: datetime64[ns]"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"t1 = addVerticalBarrier(tEvents, close, numDays=1)\n",
"t1"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (c) Apply triple-barrier method where `ptSl = [1,1]` and `t1` is the series created in `1.b`"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:24.141594Z",
"start_time": "2019-03-01T17:54:23.327391Z"
}
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"2019-03-01 10:54:23.908117 9.09% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.03 minutes.\r",
"2019-03-01 10:54:23.918158 18.18% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:23.919821 27.27% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:23.921352 36.36% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:23.921910 45.45% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:23.923873 54.55% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:23.924893 63.64% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:23.935960 72.73% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:23.940660 81.82% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:23.972865 90.91% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:23.983755 100.0% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" t1 trgt\n",
"2018-02-13 13:43:37 2018-02-14 13:53:59 0.014365\n",
"2018-02-14 10:30:48 2018-02-15 09:31:56 0.012136\n",
"2018-02-14 13:36:02 2018-02-15 13:42:09 0.011688\n",
"2018-02-15 09:31:56 2018-02-16 09:42:36 0.011244\n",
"2018-02-15 14:05:41 2018-02-16 12:05:18 0.010183\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 929 entries, 2009-10-05 14:55:48 to 2018-02-15 14:05:41\n",
"Data columns (total 2 columns):\n",
"t1 929 non-null datetime64[ns]\n",
"trgt 929 non-null float64\n",
"dtypes: datetime64[ns](1), float64(1)\n",
"memory usage: 21.8 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"# create target series\n",
"ptsl = [1,1]\n",
"target=dailyVol\n",
"# select minRet\n",
"minRet = 0.01\n",
"\n",
"# Run in single-threaded mode on Windows\n",
"import platform\n",
"if platform.system() == \"Windows\":\n",
" cpus = 1\n",
"else:\n",
" cpus = cpu_count() - 1\n",
" \n",
"events = getEvents(close,tEvents,ptsl,target,minRet,cpus,t1=t1)\n",
"\n",
"cprint(events)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (d) Apply `getBins` to generate labels"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:24.550849Z",
"start_time": "2019-03-01T17:54:24.143356Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" ret bin\n",
"2018-02-13 13:43:37 0.010108 1.0\n",
"2018-02-14 10:30:48 0.015045 1.0\n",
"2018-02-14 13:36:02 0.005056 1.0\n",
"2018-02-15 09:31:56 0.003964 1.0\n",
"2018-02-15 14:05:41 0.010431 1.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 929 entries, 2009-10-05 14:55:48 to 2018-02-15 14:05:41\n",
"Data columns (total 2 columns):\n",
"ret 929 non-null float64\n",
"bin 929 non-null float64\n",
"dtypes: float64(2)\n",
"memory usage: 61.8 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n",
" 1.0 523\n",
"-1.0 406\n",
"Name: bin, dtype: int64\n"
]
}
],
"source": [
"labels = getBins(events, close)\n",
"cprint(labels)\n",
"print(labels.bin.value_counts())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## [3.2] Use snippet 3.8 to drop under-populated labels"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:24.976600Z",
"start_time": "2019-03-01T17:54:24.552234Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" ret bin\n",
"2018-02-13 13:43:37 0.010108 1.0\n",
"2018-02-14 10:30:48 0.015045 1.0\n",
"2018-02-14 13:36:02 0.005056 1.0\n",
"2018-02-15 09:31:56 0.003964 1.0\n",
"2018-02-15 14:05:41 0.010431 1.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 929 entries, 2009-10-05 14:55:48 to 2018-02-15 14:05:41\n",
"Data columns (total 2 columns):\n",
"ret 929 non-null float64\n",
"bin 929 non-null float64\n",
"dtypes: float64(2)\n",
"memory usage: 61.8 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"clean_labels = dropLabels(labels)\n",
"cprint(clean_labels)"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:25.770917Z",
"start_time": "2019-03-01T17:54:24.979035Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
" 1.0 523\n",
"-1.0 406\n",
"Name: bin, dtype: int64"
]
},
"execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"clean_labels.bin.value_counts()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## [3.3] Adjust the `getBins` function to return a `0` whenever the vertical barrier is the one touched first."
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:29.719766Z",
"start_time": "2019-03-01T17:54:25.772774Z"
}
},
"outputs": [],
"source": [
"def getBinsNew(events, close, t1=None):\n",
" '''\n",
" Compute event's outcome (including side information, if provided).\n",
" events is a DataFrame where:\n",
" -events.index is event's starttime\n",
" -events['t1'] is event's endtime\n",
" -events['trgt'] is event's target\n",
" -events['side'] (optional) implies the algo's position side\n",
" -t1 is original vertical barrier series\n",
" Case 1: ('side' not in events): bin in (-1,1) <-label by price action\n",
" Case 2: ('side' in events): bin in (0,1) <-label by pnl (meta-labeling)\n",
" '''\n",
" #1) prices aligned with events\n",
" events_=events.dropna(subset=['t1'])\n",
" px=events_.index.union(events_['t1'].values).drop_duplicates()\n",
" px=close.reindex(px,method='bfill')\n",
" #2) create out object\n",
" out=pd.DataFrame(index=events_.index)\n",
" out['ret']=px.loc[events_['t1'].values].values/px.loc[events_.index]-1\n",
" if 'side' in events_:out['ret']*=events_['side'] # meta-labeling\n",
" out['bin']=np.sign(out['ret'])\n",
" \n",
" if 'side' not in events_:\n",
" # only applies when not meta-labeling\n",
" # to update bin to 0 when vertical barrier is touched, we need the original\n",
" # vertical barrier series since the events['t1'] is the time of first \n",
" # touch of any barrier and not the vertical barrier specifically. \n",
" # The index of the intersection of the vertical barrier values and the \n",
" # events['t1'] values indicate which bin labels needs to be turned to 0\n",
" vtouch_first_idx = events[events['t1'].isin(t1.values)].index\n",
" out.loc[vtouch_first_idx, 'bin'] = 0.\n",
" \n",
" if 'side' in events_:out.loc[out['ret']<=0,'bin']=0 # meta-labeling\n",
" return out"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## [3.4] Develop moving average crossover strategy. For each obs. the model suggests a side but not size of the bet"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:32.472289Z",
"start_time": "2019-03-01T17:54:29.721345Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" price fast slow\n",
"2018-02-26 15:31:06 115.29 115.227691 115.057569\n",
"2018-02-26 15:40:15 115.41 115.273268 115.101623\n",
"2018-02-26 15:49:42 115.20 115.254951 115.113920\n",
"2018-02-26 15:59:04 115.27 115.258713 115.133430\n",
"2018-02-26 16:16:14 115.30 115.269035 115.154251\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 30860 entries, 2009-09-28 09:53:49 to 2018-02-26 16:16:14\n",
"Data columns (total 3 columns):\n",
"price 30860 non-null float64\n",
"fast 30860 non-null float64\n",
"slow 30860 non-null float64\n",
"dtypes: float64(3)\n",
"memory usage: 964.4 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"fast_window = 3\n",
"slow_window = 7\n",
"\n",
"close_df = (pd.DataFrame()\n",
" .assign(price=close)\n",
" .assign(fast=close.ewm(fast_window).mean())\n",
" .assign(slow=close.ewm(slow_window).mean()))\n",
"cprint(close_df)"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:33.757620Z",
"start_time": "2019-03-01T17:54:32.473676Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 29,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"def get_up_cross(df):\n",
" crit1 = df.fast.shift(1) < df.slow.shift(1)\n",
" crit2 = df.fast > df.slow\n",
" return df.fast[(crit1) & (crit2)]\n",
"\n",
"def get_down_cross(df):\n",
" crit1 = df.fast.shift(1) > df.slow.shift(1)\n",
" crit2 = df.fast < df.slow\n",
" return df.fast[(crit1) & (crit2)]\n",
"\n",
"up = get_up_cross(close_df)\n",
"down = get_down_cross(close_df)\n",
"\n",
"f, ax = plt.subplots(figsize=(11,8))\n",
"\n",
"close_df.loc['2014':].plot(ax=ax, alpha=.5)\n",
"up.loc['2014':].plot(ax=ax,ls='',marker='^', markersize=7,\n",
" alpha=0.75, label='upcross', color='g')\n",
"down.loc['2014':].plot(ax=ax,ls='',marker='v', markersize=7, \n",
" alpha=0.75, label='downcross', color='r')\n",
"\n",
"ax.legend()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (a) Derive meta-labels for `ptSl = [1,2]` and `t1` where `numdays=1`. Use as `trgt` dailyVol computed by snippet 3.1 (get events with sides)"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:34.189161Z",
"start_time": "2019-03-01T17:54:33.760127Z"
},
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" 0\n",
"2018-02-21 11:10:00 1\n",
"2018-02-21 15:12:30 -1\n",
"2018-02-22 11:48:39 1\n",
"2018-02-22 13:34:29 -1\n",
"2018-02-23 10:01:41 1\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 1712 entries, 2009-09-30 09:45:21 to 2018-02-23 10:01:41\n",
"Data columns (total 1 columns):\n",
"0 1712 non-null int64\n",
"dtypes: int64(1)\n",
"memory usage: 26.8 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"side_up = pd.Series(1, index=up.index)\n",
"side_down = pd.Series(-1, index=down.index)\n",
"side = pd.concat([side_up,side_down]).sort_index()\n",
"cprint(side)"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:36.918898Z",
"start_time": "2019-03-01T17:54:34.191595Z"
}
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 30858/30858 [00:01<00:00, 17281.84it/s]\n",
"2019-03-01 10:54:36.730759 100.0% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes..\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" side t1 trgt\n",
"2018-02-13 13:43:37 NaN 2018-02-14 13:53:59 0.014365\n",
"2018-02-14 10:30:48 NaN 2018-02-15 10:42:27 0.012136\n",
"2018-02-14 13:36:02 NaN 2018-02-15 13:42:09 0.011688\n",
"2018-02-15 09:31:56 NaN 2018-02-16 09:42:36 0.011244\n",
"2018-02-15 14:05:41 NaN 2018-02-16 14:15:08 0.010183\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 929 entries, 2009-10-05 14:55:48 to 2018-02-15 14:05:41\n",
"Data columns (total 3 columns):\n",
"side 102 non-null float64\n",
"t1 929 non-null datetime64[ns]\n",
"trgt 929 non-null float64\n",
"dtypes: datetime64[ns](1), float64(2)\n",
"memory usage: 29.0 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"minRet = .01 \n",
"ptsl=[1,2]\n",
"\n",
"dailyVol = getDailyVol(close_df['price'])\n",
"tEvents = getTEvents(close_df['price'],h=dailyVol.mean())\n",
"t1 = addVerticalBarrier(tEvents, close_df['price'], numDays=1)\n",
"\n",
"ma_events = getEvents(close_df['price'],tEvents,ptsl,target,minRet,cpus,\n",
" t1=t1,side=side)\n",
"cprint(ma_events)"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:37.223058Z",
"start_time": "2019-03-01T17:54:36.920513Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
" 1.0 53\n",
"-1.0 49\n",
"Name: side, dtype: int64"
]
},
"execution_count": 32,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"ma_events.side.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:37.538874Z",
"start_time": "2019-03-01T17:54:37.225750Z"
}
},
"outputs": [],
"source": [
"ma_side = ma_events.dropna().side"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:37.883132Z",
"start_time": "2019-03-01T17:54:37.540690Z"
},
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" ret bin\n",
"2016-07-07 14:28:00 -0.018703 0.0\n",
"2016-07-08 09:30:57 0.010571 1.0\n",
"2018-02-06 10:18:08 -0.026702 0.0\n",
"2018-02-07 15:28:09 -0.030792 0.0\n",
"2018-02-13 09:30:00 -0.001803 0.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 102 entries, 2009-10-29 13:40:22 to 2018-02-13 09:30:00\n",
"Data columns (total 2 columns):\n",
"ret 102 non-null float64\n",
"bin 102 non-null float64\n",
"dtypes: float64(2)\n",
"memory usage: 2.4 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"ma_bins = getBinsNew(ma_events,close_df['price'], t1).dropna()\n",
"cprint(ma_bins)"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:38.190267Z",
"start_time": "2019-03-01T17:54:37.886249Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" ret bin side\n",
"2016-07-07 14:28:00 -0.018703 0.0 -1\n",
"2016-07-08 09:30:57 0.010571 1.0 1\n",
"2018-02-06 10:18:08 -0.026702 0.0 -1\n",
"2018-02-07 15:28:09 -0.030792 0.0 1\n",
"2018-02-13 09:30:00 -0.001803 0.0 -1\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 102 entries, 2009-10-29 13:40:22 to 2018-02-13 09:30:00\n",
"Data columns (total 3 columns):\n",
"ret 102 non-null float64\n",
"bin 102 non-null float64\n",
"side 102 non-null int64\n",
"dtypes: float64(2), int64(1)\n",
"memory usage: 3.2 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"Xx = pd.merge_asof(ma_bins, side.to_frame().rename(columns={0:'side'}),\n",
" left_index=True, right_index=True, direction='forward')\n",
"cprint(Xx)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (b) Train Random Forest to decide whether to trade or not `{0,1}` since underlying model (crossing m.a.) has decided the side, `{-1,1}`"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:38.523365Z",
"start_time": "2019-03-01T17:54:38.192661Z"
}
},
"outputs": [],
"source": [
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.metrics import roc_curve, classification_report"
]
},
{
"cell_type": "code",
"execution_count": 37,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:44.603030Z",
"start_time": "2019-03-01T17:54:38.525191Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" precision recall f1-score support\n",
"\n",
" 0.0 0.00 0.00 0.00 24\n",
" 1.0 0.53 1.00 0.69 27\n",
"\n",
" micro avg 0.53 0.53 0.53 51\n",
" macro avg 0.26 0.50 0.35 51\n",
"weighted avg 0.28 0.53 0.37 51\n",
"\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"X = ma_side.values.reshape(-1,1)\n",
"#X = Xx.side.values.reshape(-1,1)\n",
"y = ma_bins.bin.values\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5,\n",
" shuffle=False)\n",
"\n",
"n_estimator = 10000\n",
"rf = RandomForestClassifier(max_depth=2, n_estimators=n_estimator,\n",
" criterion='entropy', random_state=RANDOM_STATE)\n",
"rf.fit(X_train, y_train)\n",
"\n",
"# The random forest model by itself\n",
"y_pred_rf = rf.predict_proba(X_test)[:, 1]\n",
"y_pred = rf.predict(X_test)\n",
"fpr_rf, tpr_rf, _ = roc_curve(y_test, y_pred_rf)\n",
"print(classification_report(y_test, y_pred))\n",
"\n",
"plt.figure(1)\n",
"plt.plot([0, 1], [0, 1], 'k--')\n",
"plt.plot(fpr_rf, tpr_rf, label='RF')\n",
"plt.xlabel('False positive rate')\n",
"plt.ylabel('True positive rate')\n",
"plt.title('ROC curve')\n",
"plt.legend(loc='best')\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## [3.5] Develop mean-reverting Bollinger Band Strategy. For each obs. model suggests a side but not size of the bet."
]
},
{
"cell_type": "code",
"execution_count": 38,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:44.934155Z",
"start_time": "2019-03-01T17:54:44.604514Z"
}
},
"outputs": [],
"source": [
"def bbands(price, window=None, width=None, numsd=None):\n",
" \"\"\" returns average, upper band, and lower band\"\"\"\n",
" ave = price.rolling(window).mean()\n",
" sd = price.rolling(window).std(ddof=0)\n",
" if width:\n",
" upband = ave * (1+width)\n",
" dnband = ave * (1-width)\n",
" return price, np.round(ave,3), np.round(upband,3), np.round(dnband,3) \n",
" if numsd:\n",
" upband = ave + (sd*numsd)\n",
" dnband = ave - (sd*numsd)\n",
" return price, np.round(ave,3), np.round(upband,3), np.round(dnband,3)"
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:45.316561Z",
"start_time": "2019-03-01T17:54:44.935929Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" price ave upper lower\n",
"2018-02-26 15:31:06 115.29 114.005 114.959 113.051\n",
"2018-02-26 15:40:15 115.41 114.069 115.008 113.129\n",
"2018-02-26 15:49:42 115.20 114.124 115.047 113.202\n",
"2018-02-26 15:59:04 115.27 114.183 115.083 113.282\n",
"2018-02-26 16:16:14 115.30 114.231 115.125 113.338\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 30811 entries, 2009-10-01 15:51:02 to 2018-02-26 16:16:14\n",
"Data columns (total 4 columns):\n",
"price 30811 non-null float64\n",
"ave 30811 non-null float64\n",
"upper 30811 non-null float64\n",
"lower 30811 non-null float64\n",
"dtypes: float64(4)\n",
"memory usage: 1.2 MB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"window=50\n",
"bb_df = pd.DataFrame()\n",
"bb_df['price'],bb_df['ave'],bb_df['upper'],bb_df['lower']=bbands(close, window=window, numsd=1)\n",
"bb_df.dropna(inplace=True)\n",
"cprint(bb_df)"
]
},
{
"cell_type": "code",
"execution_count": 40,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:45.892818Z",
"start_time": "2019-03-01T17:54:45.319710Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 40,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"f,ax=plt.subplots(figsize=(11,8))\n",
"bb_df.loc['2014'].plot(ax=ax)"
]
},
{
"cell_type": "code",
"execution_count": 41,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:46.625944Z",
"start_time": "2019-03-01T17:54:45.894447Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 41,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"def get_up_cross(df, col):\n",
" # col is price column\n",
" crit1 = df[col].shift(1) < df.upper.shift(1) \n",
" crit2 = df[col] > df.upper\n",
" return df[col][(crit1) & (crit2)]\n",
"\n",
"def get_down_cross(df, col):\n",
" # col is price column \n",
" crit1 = df[col].shift(1) > df.lower.shift(1) \n",
" crit2 = df[col] < df.lower\n",
" return df[col][(crit1) & (crit2)]\n",
"\n",
"bb_down = get_down_cross(bb_df, 'price')\n",
"bb_up = get_up_cross(bb_df, 'price') \n",
"\n",
"f, ax = plt.subplots(figsize=(11,8))\n",
"\n",
"bb_df.loc['2014':].plot(ax=ax, alpha=.5)\n",
"bb_up.loc['2014':].plot(ax=ax, ls='', marker='^', markersize=7,\n",
" alpha=0.75, label='upcross', color='g')\n",
"bb_down.loc['2014':].plot(ax=ax, ls='', marker='v', markersize=7, \n",
" alpha=0.75, label='downcross', color='r')\n",
"ax.legend()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (a) Derive meta-labels for `ptSl=[0,2]` and `t1` where `numdays=1`. Use as `trgt` dailyVol."
]
},
{
"cell_type": "code",
"execution_count": 42,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:47.459211Z",
"start_time": "2019-03-01T17:54:46.627389Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" 0\n",
"2018-02-22 13:34:29 1\n",
"2018-02-22 14:20:25 1\n",
"2018-02-22 14:44:33 1\n",
"2018-02-23 13:41:26 -1\n",
"2018-02-23 14:40:49 -1\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 2040 entries, 2009-10-06 09:29:52 to 2018-02-23 14:40:49\n",
"Data columns (total 1 columns):\n",
"0 2040 non-null int64\n",
"dtypes: int64(1)\n",
"memory usage: 31.9 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"2019-03-01 10:54:47.213061 9.09% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.03 minutes.\r",
"2019-03-01 10:54:47.217601 18.18% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:47.219915 27.27% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:47.222889 36.36% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:47.224198 45.45% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:47.224620 54.55% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:47.226273 63.64% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:47.227236 72.73% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:47.230049 81.82% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:47.235371 90.91% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:47.236315 100.0% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" side t1 trgt\n",
"2018-02-13 13:43:37 -1.0 2018-02-14 13:53:59 0.014365\n",
"2018-02-14 10:30:48 NaN 2018-02-15 10:42:27 0.012136\n",
"2018-02-14 13:36:02 NaN 2018-02-15 13:42:09 0.011688\n",
"2018-02-15 09:31:56 NaN 2018-02-16 09:42:36 0.011244\n",
"2018-02-15 14:05:41 NaN 2018-02-16 14:15:08 0.010183\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 929 entries, 2009-10-05 14:55:48 to 2018-02-15 14:05:41\n",
"Data columns (total 3 columns):\n",
"side 139 non-null float64\n",
"t1 929 non-null datetime64[ns]\n",
"trgt 929 non-null float64\n",
"dtypes: datetime64[ns](1), float64(2)\n",
"memory usage: 29.0 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n",
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" side\n",
"2016-07-07 10:17:10 -1.0\n",
"2016-07-08 09:30:57 -1.0\n",
"2018-02-06 10:18:08 1.0\n",
"2018-02-06 14:19:57 1.0\n",
"2018-02-13 13:43:37 -1.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 139 entries, 2009-10-06 09:29:52 to 2018-02-13 13:43:37\n",
"Data columns (total 1 columns):\n",
"side 139 non-null float64\n",
"dtypes: float64(1)\n",
"memory usage: 2.2 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"bb_side_up = pd.Series(-1, index=bb_up.index) # sell on up cross for mean reversion\n",
"bb_side_down = pd.Series(1, index=bb_down.index) # buy on down cross for mean reversion\n",
"bb_side_raw = pd.concat([bb_side_up,bb_side_down]).sort_index()\n",
"cprint(bb_side_raw)\n",
"\n",
"minRet = .01 \n",
"ptsl=[0,2]\n",
"bb_events = getEvents(close,tEvents,ptsl,target,minRet,cpus,t1=t1,side=bb_side_raw)\n",
"cprint(bb_events)\n",
"\n",
"bb_side = bb_events.dropna().side\n",
"cprint(bb_side)"
]
},
{
"cell_type": "code",
"execution_count": 43,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:47.820661Z",
"start_time": "2019-03-01T17:54:47.460801Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
" 1.0 72\n",
"-1.0 67\n",
"Name: side, dtype: int64"
]
},
"execution_count": 43,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"bb_side.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 44,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:48.192517Z",
"start_time": "2019-03-01T17:54:47.823058Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" ret bin\n",
"2016-07-07 10:17:10 -0.003791 0.0\n",
"2016-07-08 09:30:57 -0.010571 0.0\n",
"2018-02-06 10:18:08 0.025085 1.0\n",
"2018-02-06 14:19:57 0.028779 1.0\n",
"2018-02-13 13:43:37 -0.010108 0.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 139 entries, 2009-10-06 09:29:52 to 2018-02-13 13:43:37\n",
"Data columns (total 2 columns):\n",
"ret 139 non-null float64\n",
"bin 139 non-null float64\n",
"dtypes: float64(2)\n",
"memory usage: 3.3 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n",
"0.0 79\n",
"1.0 60\n",
"Name: bin, dtype: int64\n"
]
}
],
"source": [
"bb_bins = getBins(bb_events,close).dropna()\n",
"cprint(bb_bins)\n",
"print(bb_bins.bin.value_counts())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### (b) train random forest to decide to trade or not. Use features: volatility, serial correlation, and the crossing moving averages from exercise 2."
]
},
{
"cell_type": "code",
"execution_count": 45,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:48.508454Z",
"start_time": "2019-03-01T17:54:48.195468Z"
}
},
"outputs": [],
"source": [
"def returns(s):\n",
" arr = np.diff(np.log(s))\n",
" return (pd.Series(arr, index=s.index[1:]))\n",
"\n",
"def df_rolling_autocorr(df, window, lag=1):\n",
" \"\"\"Compute rolling column-wise autocorrelation for a DataFrame.\"\"\"\n",
"\n",
" return (df.rolling(window=window)\n",
" .corr(df.shift(lag))) # could .dropna() here\n",
"\n",
"#df_rolling_autocorr(d1, window=21).dropna().head()"
]
},
{
"cell_type": "code",
"execution_count": 46,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:48.863084Z",
"start_time": "2019-03-01T17:54:48.511289Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" srl_corr\n",
"2018-02-26 15:31:06 0.028037\n",
"2018-02-26 15:40:15 0.015957\n",
"2018-02-26 15:49:42 0.032877\n",
"2018-02-26 15:59:04 0.046014\n",
"2018-02-26 16:16:14 0.109129\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 30859 entries, 2009-09-28 10:06:04 to 2018-02-26 16:16:14\n",
"Data columns (total 1 columns):\n",
"srl_corr 30809 non-null float64\n",
"dtypes: float64(1)\n",
"memory usage: 482.2 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"srl_corr = df_rolling_autocorr(returns(close), window=window).rename('srl_corr')\n",
"cprint(srl_corr)"
]
},
{
"cell_type": "code",
"execution_count": 47,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:49.215766Z",
"start_time": "2019-03-01T17:54:48.864737Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" vol ma_side srl_corr\n",
"2016-07-07 14:28:00 0.012624 -1.0 0.251865\n",
"2016-07-08 09:30:57 0.011944 1.0 0.238590\n",
"2018-02-06 10:18:08 0.013317 -1.0 0.123961\n",
"2018-02-07 15:28:09 0.024870 1.0 -0.005597\n",
"2018-02-13 09:30:00 0.017363 -1.0 0.198935\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 102 entries, 2009-10-29 13:40:22 to 2018-02-13 09:30:00\n",
"Data columns (total 3 columns):\n",
"vol 102 non-null float64\n",
"ma_side 102 non-null float64\n",
"srl_corr 102 non-null float64\n",
"dtypes: float64(3)\n",
"memory usage: 3.2 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"features = (pd.DataFrame()\n",
" .assign(vol=bb_events.trgt)\n",
" .assign(ma_side=ma_side)\n",
" .assign(srl_corr=srl_corr)\n",
" .drop_duplicates()\n",
" .dropna())\n",
"cprint(features)"
]
},
{
"cell_type": "code",
"execution_count": 48,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:49.534835Z",
"start_time": "2019-03-01T17:54:49.218405Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" vol ma_side srl_corr bin\n",
"2016-07-07 14:28:00 0.012624 -1.0 0.251865 0.0\n",
"2016-07-08 09:30:57 0.011944 1.0 0.238590 0.0\n",
"2018-02-06 10:18:08 0.013317 -1.0 0.123961 1.0\n",
"2018-02-07 15:28:09 0.024870 1.0 -0.005597 0.0\n",
"2018-02-13 09:30:00 0.017363 -1.0 0.198935 0.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 102 entries, 2009-10-29 13:40:22 to 2018-02-13 09:30:00\n",
"Data columns (total 4 columns):\n",
"vol 102 non-null float64\n",
"ma_side 102 non-null float64\n",
"srl_corr 102 non-null float64\n",
"bin 102 non-null float64\n",
"dtypes: float64(4)\n",
"memory usage: 4.0 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n"
]
}
],
"source": [
"Xy = (pd.merge_asof(features, bb_bins[['bin']], \n",
" left_index=True, right_index=True, \n",
" direction='forward').dropna())\n",
"cprint(Xy)"
]
},
{
"cell_type": "code",
"execution_count": 49,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:49.873144Z",
"start_time": "2019-03-01T17:54:49.536775Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
"0.0 60\n",
"1.0 42\n",
"Name: bin, dtype: int64"
]
},
"execution_count": 49,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"Xy.bin.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 50,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:54:56.011226Z",
"start_time": "2019-03-01T17:54:49.875226Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" precision recall f1-score support\n",
"\n",
" no_trade 0.47 0.73 0.58 26\n",
" trade 0.36 0.16 0.22 25\n",
"\n",
" micro avg 0.45 0.45 0.45 51\n",
" macro avg 0.42 0.45 0.40 51\n",
"weighted avg 0.42 0.45 0.40 51\n",
"\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEfCAYAAABSy/GnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzsnXl8VNX5/99PNhJDErIZVBbL4toqKiqIC4paxVqQH9YN9WtFsFJAVhfUonVlB7XuGxV36wJiQUFssCAq7lYqKCBgQjaykWSS8Pz+mMk0xkkySe69s9zzfr3ua+aee+49zydnMs+c9RFVxWAwGAyG1ogJtQEGg8FgiAyMwzAYDAZDUBiHYTAYDIagMA7DYDAYDEFhHIbBYDAYgsI4DIPBYDAEhXEYBoPBYAgK4zAMUY2IPC0i6jvqRWSHiCwWkYMC5M0RkftFZKuIeESkQEReEZF+AfLGich4EdkgIuUiUioin4rIDBFJd0adweAsxmEY3EAucADQA7gUOAZ4uXEGEekOfAycBPwJ6AOcB9QC60XknEZ544G3gLuAl4AzgKOBGcAA4Ep75fwcEUlwsjyDexGz0tsQzYjI00A3VT2zUdp4YBGQpqplvrQ3gROAQxrSGuVfDhwL/EpVq0RkCjAbGKSq6wKUma6qJc3YEwfcjNepdAMKgX+o6njfdQUuV9VnG93zLrBDVf/Pd74VeBbIAC4Cvgf+C2Sr6tlNynsbKFXVi33nZwEzfXqKgZXAVFUtav6vaDB4MS0Mg6sQkQOBkUC978DXhXQe8EBTZ+HjHiAHOMt3fjmwOpCzAGjOWfh4Avgz3i/tI4D/h/cLv61MAHYDA/E6n8XAkMZdbSLSYPMzvvMzgDeAF4CjgOHAwcBrIiLtsMHgMuJCbYDB4ACDRaQC7w+kJF/aXFWt9L3v67v2dTP3N6Qf6ns9BPhXW40QkT7AFcCFqvqKL3kLsL6tzwI+UtWZjZ79LZAHjALu8yVfBhTgbUUA3AYsUtX7G913JbANb5faZ+2ww+AiTAvD4AY+BPrh7XL6K94v6FsbXW/t13XTflsJkBYMx/peV7aYKzg2ND5R1X3AErytnwYuB5aoar3v/HjgehGpaDiAb3zX+lpgkyHKMS0MgxuoUtXNvvdficghwIPAH31p3wH7gF8DrwW4/9e+102NXo+0yVbllw4sPkC+ygBpzwDTROQ4oAavk2w8AB+Dt/Xx9wD35rXdVIPbMC0MgxuZCVwpIv0BVLUYeBsYJyKpAfLfDOQD7/jOnwXOEJGBgR7ewrTajb7Xs5u5Dt5xiQMbPasT3rGOVlHVr31lXOE7PlPVLxpl+Rg4UlU3BzgqginD4G6MwzC4DlX9FliGdzC7gXF4B8FXi8g5ItJdRI4XkeeA04H/U9UqX96FwCpghYhMFZH+ItLTd9/reL+sA5W7GW+30d9EZJSI9PaVMbFRtneBa0VkoIj8GngaaMu02WeAS/COXyxucu02YJiIzBeRfr7yzxGRJ0Qk6RdPMhiaYByGwa3MAs4UkSEAqroN6I93vOMRvIPRbwOdgIGq+s+GG1W1FjgX7zjIxcD7wJd4HdAGfLOSmuEq3/PvBP6DtwvsV42uTwW+Alb4yv8X8FEbdD0HdAH29733o6rv4V0z8hu8a1O+AOYD5XjXmxgMLWLWYRgMBoMhKEwLw2AwGAxBYRyGwWAwGILCOAyDwWAwBIVxGAaDwWAIiqhduLdmzRrt1KlTu+6tq6sjLi5q/zQBMZrdgdHsDjqiee/evYVDhgzJDnQtav+KnTp14rDDDmvXvTU1NbTX2UQqRrM7MJrdQUc0b9y4cVtz10yXVADy8ty3S4LR7A6MZndgl2bjMAIQHx9o657oxmh2B0azO7BLs3EYAUhLSwu1CY5jNLsDo9kd2KXZOIwAFBYWhtoExzGa3YHR7A7s0mwcRgDMLxJ3YDS7A6PZOhybJSUiF+PdEfRoYD9VbbFs39bTf8Mbi+An4C+N4xx3BFWloqKC5vbR8ng8lJUFitQZuagqiYmJzc6c8Hg8DlsUeoxmd2A0W4eT02pL8DqAJODRljKKSBrenTrnAKcAp+KNO7yluTjKbaGiooJOnTqRkBB41+hOnTpF3TQ8VWXv3r3U1dWRnJz8i+tVVVUB7opujGZ3YDRbh2MOQ1VXAIjI4CCyjwCqgFnqbQa8IyKvAWOAoBxGXV0dmzdv9p9nZGSQkZHRYEuzzgKic1aFiJCcnNxsy6lr164OWxR6jGZ34ITmW1ZsYcOP4dUrsXJ0T8ufGa4L944GNurP+4w28vN4xS2ye/duRo4c6T8fO3YsM2fOJC8vj6SkJPbbbz/q6+uJi4ujvt4b8jguLo7a2lr27dtHfHw89fX1xMfHU1dXB0BsbCx1dXXExsaiqv58tbW1iEjQ1+Pi4ti3b9/PrsfExBATE+O/Xl9fj6r+7LqI/Mzm1q431hQbGwtAZWUlCQkJFBQUICJkZGRQUFBAVVUVWVlZVFZW0rVrV/Ly8oiPjyctLY3CwkLS0tLweDxUVVX5ryckJJCSkkJRURHp6elUVVVRXV3tv56YmEhSUhIlJSVkZmZSXl6Ox+PxX09KSiIhIYHS0lKysrIoLS2ltrbWfz05OZnY2FjKysrIzs6muLgYVSU7O5v8/Hw6d+4MeFuMOTk5v9CUmppKfX19s5qqq6v9dkeLptbqqaamhrS0tKjS1Fo9eTweOnfubKumcHMWALt27WqXppZwPB6Gr4XxbktjGCLyBBCnqlc2SrsKmKGqfYIpJzc3VxMTE/3njVsYZWVlpKYGisTpxePxtNgCiWSa0/7TTz9xwAEHhMCi0GE0uwMnNJ/9+KcArBx9jK3ltMR9993Hgw8+yG233ca5557LQQcd1K7nbNy48ZMhQ4b0D3QtXFsY5cDBTdK6AEG78bi4OPr0Ccq3/IKGX+NuIiUlJdQmOI7R7A6iVfOPP/7Itm3bOPnkkwGYNGkSl112Gd26daOiwp4Q7eE6rfZzoKmrPsaXbjsNXVBuoqioKNQmOI7R7A6iTfO+fft4/PHHGTRoEH/84x/9+hISEujWrRtgn2bHHIaIxIpIIr6A9iKS6DskQPbXgP1EZJqIJPjiLo+gldlVVhEuO1uef/75dO3ale7du9OzZ09OPfVUXn/99YDXG44JEya0q6z09HSrzI4YjGZ3EE2av/vuO8477zymT59ORUUFJ510UsDlAXZpdrKFcTnemU8rgFjf+yqgp4icIiIVItIDQFX3AEOBC4FS4DHgWium1AbDvn37nCgmKKZOncqPP/7Ili1buOSSSxgzZgzff//9L643HIsWLWpXOWbqoTswmiOT2tpa5s+fz6mnnsqHH35ITk4OzzzzDE8//TRZWVm/yB8N02qfBp5u5vJW4GdD9Kr6EXCCrUb5aBiwcor2DIzFxcVxxRVXMGPGDL788kt69eplqU3V1dWWPi8SMJrdQTRoHjNmDG+88QYAl112GX/961/p0qVLs/nt0hwefS+GVvF4PDz55JMA7R7MbwkzP98dGM2RyejRo/n888+ZN28egwcPbjW/XZqNw+CXv/jDKeDKvHnzeOCBB6ioqCA+Pp6FCxdy5JFH/uJ6Ay+//DLHH398m8vJy8ujZ0/rF/qEM0azO4hEzevXr+eDDz5gypQpAAwaNIgPP/ww6EXFdmkO11lSISUmJnz+LJMnT2br1q1s3ryZs846i9zc3IDXG472OAuAxmtW3ILR7A4iSXN5eTk33HAD5513HnfddRfr16/3X2vLDhR2aQ6fb8YwIpwcRgNdunRh4cKFvPPOOyxfvtzy5yclJVn+zHDHaHYHkaJ51apVDBo0iMcee4zY2FimTJnCMce0byGgXZrD75sxDAjXdRjp6elcd911/PWvf7V8JldJSYmlz4sEjGZ3EO6aS0pKGDduHBdeeCE7duygX79+rF69mhkzZrS7a9wuzcZhBCBc1mEEYuzYseTn5/PCCy9Y+tzMzExLnxcJGM3uINw133vvvTz//PMkJiYyc+ZMVq5cya9//esOPdMuzeH7zRhC6uvrw2J7kKVLl/4iLTU11b8O49JLL7WsrPLy8qA2H4smjGZ3EI6aVZWGNcs33HADP/30E7fddptlMyDt0mxaGAFwekPGcMAEmXEHRnNoUVWWLFnC0KFDqampAbwboy5evNjS6fLREEApYojGeBitEQ1z1duK0ewO2qPZjvgW27dv5/rrr2fNmjUAvPbaa1x88cWWltGAXfVsWhgBqK2tDbUJjpOXlxdqExzHaHYH7dHcHmdxQvfAIRPq6+t55JFHGDRoEGvWrCEjI4NHHnmEiy66qM1lBItd9WxaGAEIx2m1dhMpUw+txGh2Bx3R3NH4Fps2bWLixIls2LABgAsuuIB7772X7OzsDj23NeyqZ+MwAhB4A93oJloDRrWE0ewOQql548aNbNiwga5duzJnzhyGDh3qSLl2aXalwxCRFqPqNYQ5jSZUlb179zY7+6u0tLTFzcyiEaPZHTitubi42B/d8+KLL2bPnj1ceumlpKWlOWaDXZqj61sxSDp37kxFRUWzOzrW1dX5ZzBEC6pKYmJiswuBAm2RHO0Yze7AKc1VVVXMmjWLJ554gtWrV9OnTx9EhD/96U+OlN8YuzS70mGISIthG3ft2sWBBx7ooEWhp7S0lOTk5FCb4ShGsztwQvO///1vJk6cyJYtW4iJiSE3N9eWXaWDxS7NrnQYreHGWVJGszswmq2lrKyMO+64wx964NBDD+X++++nf//+tpUZDHZpNg4jAGauujswmiOXtq+TKLbchvXr13PNNdewc+dO4uLimDx5MpMmTQqL0AhmHYaDmLnq7sBojlysXlQXiObWVTSQmZlJYWEhxxxzDGvWrOHGG28MC2cBZh2Go7itjxeMZrcQbZqDWSdRWFhoySCwqpKbm8spp5yCiNC3b1+WLVvG0UcfHXazKu2qZ9PCCEA4bDzoNEazOzCa28dPP/3E5ZdfzvDhw3+2U/Rxxx0Xds4C7Ktn4zACUFZmf3M33DCa3YHR3DZUlcWLFzNw4ECWL19OSkpKROwEYVc9h59rDAPsXrYfjhjN7sBoDp6tW7dy/fXX869//QuA3/72t8yZM4eDDjrISvNswa56Dn9XGQKKi62fURHuGM3uwGgOjg8//JBBgwbxr3/9i8zMTB577DGee+65iHAWYF89mxZGANwYD8NodgdGc3D069eP7t27c9RRR3HPPfeEfdS+pthVz8ZhBMA0292B0Rxe2BGDAoLT7PF4eOihh7jiiitIT0+nU6dOrFy5ktTUlqfWhiumS8pB8vPzQ22C4xjN7iCcNbfVWbS2TqKB1jRv3LiR008/ndtvv50ZM2b40yPVWYB99WxaGAEIt/i/TmA0u4NI0NzRGBRNaU7z3r17ueeee3jooYfYt28fvXr14tJLL7W07FBhVz0bh2EwGFzH2rVrmThxIj/88AMxMTGMHz+eG264gf322y/UpoU1pksqABUVFaE2wXGMZndgNMPmzZsZNmwYP/zwA0cccQTvvPMOt99+e1Q5C7vq2bQwApCTkxNqExzHaHYHRjP06dOH0aNHk5WVxcSJE6MyCqFd9exYC0NEYkVktogUiEi5iLwqIs1u8CIiU0Vkiy/vdyJynVO2FhQUOFVU2GA0uwM3at60aRNjxozh3//+tz/tvvvuY9q0aVHpLMC+enayhXEjMAw4ESgCngT+DpzbNKOI/B64HRiiqutFZCDwroh8p6rv2G2oG2N6G83uwE2aVZVXX32V6dOns2fPHjZt2sSaNWtc8TewS6OTDmMMcIeqfg8gItOBzSJysKpubZK3D/C5qq4HUNV1IvIFcDQQlMOoq6tj8+bN/vOMjAx/nN3WCDZfNGE0uwMnNdu1riIYdu7cydSpU1mxYgUAp512GvPnz3eFswD76tkRhyEiaUAP4JOGNFXdIiJlwFHA1ia3vAD8UUQGAeuAQcAhwD+DLXP37t2MHDnSfz527FhmzpxJXl4eycnJxMbGUlZWRnZ2NsXFxagq2dnZ5OfnU1VVRXZ2NhUVFeTk5FBQUICIkJGRQUFBAampqdTX11NZWUnXrl3Jy8sjPj6etLQ0CgsLSUtLw+PxUFVV5b+ekJBASkoKRUVFpKenU1VVRXV1tf96YmIiSUlJlJSUkJmZSXl5OR6Px389KSmJhIQESktLycrKorS0lNraWv/11jQ1TLNrTlNVVRVZWVlRpam1eqqurvbbHS2aWqunmpoa0tLSHNHUHmfR/6DObNu2rd2fvf3335+//e1vLFiwgMrKSjp37szEiRO55JJLqK2tpaamJiLqqaOfPY/HQ3Jycrs0tYQ4sVWAiHQHtgO9VPWHRunbgBmq+myT/HHALcDN/G+c5XpVfSDYMnNzczUxMdF/3pYWRklJCenp6cEWFRUYze7ASc1nP/4pYP26ipYoKSnhhBNOoKioiKFDhzJ79mwSExNNPbeBjRs3fjJkyJCAMWad6pIq972mNUnvAgT6GXIrcAnQD/gPcATwpohUqeoTwRQYFxfX7iDs9fX17bovkjGa3UE0aq6rq0NViY+PJz09nQULFuDxeBg+fDgiQmFhYahNdBy76tmRWVKqugdvC+PYhjQR6QWkAl8EuOU44DVV/Ua9fA28DvzOCXsrKyudKCasMJrdQbRp/uabbzjnnHOYP3++P+28887jggsu8I9XRJvmYLBLs5ML9x4FbhCRX4lIKnAfsCLAgDfAB8BwEekLICKHA8OBjU4YalcA9XDGaHYH0aK5pqaGe+65h8GDB7Nx40ZefPFFampqAuaNFs1twS7NTjqMe4GlwEfATiAWGAUgIpeJSOOlibOB14B3fOkr8LYw7nXCULsCqIczRrM7iAbNH330EYMHD2b27NnU1dVx9dVX895779GpU6eA+aNBc1uxS7Nj02pVtR6Y6juaXlsCLGl0Xod33caNTtnXmPj4+FAUG1KMZncQyZo9Hg+33347Dz/8MKpK7969WbhwISeddFKL90Wy5vZil+agHYZv5tLxwEGq+oqIJAGoapUtloWQtLSmY/PRj9HsDjqiOZTrKsD7JfjNN98QExPDn//8Z6ZPn05SUlKr95l6to6gHIaIHAm84TvtCrwCDAEuwzubKaooLCwkOTk51GY4itHsDjqiuT3OItiYFc1RWlpKRUUFBx10ECLCggULKCkpoV+/fkE/w9SzdQTbwngIuFNVnxaREl/aGuBhyy0KA8wvEndgNLcPp9ZVLF++nKlTp9K7d2/eeOMNYmJi6NmzJz179mzTc0w9W0ewg96/AZ7xvVcAVa0Aomc/4EZ4PJ5Qm+A4RrM7iATNu3fv5o9//COjRo0iLy8Pj8dDaWlpu58XCZqtxi7NwTqMbcDPflaISH9gi+UWhQFVVVE3LNMqRrM7CGfNqspLL73EwIEDef3110lOTubee+9l+fLlHVqpHc6a7cIuzcF2Sd0GvCUifwMSRGQaMA74ky1WhRgzb9sdGM3hg6py5ZVXsmzZMgAGDx7MggUL6NGjR4efHa6a7SSk6zBU9U3g90B3vIvqDgX+oKpv22JViDHztt2B0Rw+iAhHH300aWlpPPDAA7z66quWOAsIX812EtJ1GCJygaq+hnfRXeP04ar6ui2WhZBoDarSEkazOwgnzZs3b2b79u2cccYZAEyYMIFRo0ZZHi0unDQ7hV2agx3DeKaZ9CetMiScSElJCbUJjmM0u4Nw0FxXV8fChQs55ZRTGDNmjD86XHx8vC2hRcNBs9PYpbnFFoaINLQJY3xblDeOPtILCLx5S4RTVFQU1N7w0YTR7A5Crfmrr75i/PjxfP755wCMGDHC9pXYodYcCuzS3FqX1FZ802jxzpRqTCHwF6sNCgfctnc+GM1uIVSaq6urmTt3LgsXLqSuro7u3bszf/58f3eUnZh6to7WuqTigU7Aet/7hiNOVfdX1YdssSrEmGl47sBodo5rrrmGuXPnUl9fz5gxY/jggw8ccRZg6tlKWmxh+DYMBG+IVNdQXV0dahMcx2h2B6HSPG7cODZv3sz8+fMZMGCAo2WberaOYGdJxQJjgdOALBqNZaiqMz8THMTM23YHRrN9rF69mnXr1jFjxgwABgwYwNq1a4mNjXWk/MaYeraOYGdJzQMmABuAE4G3gG7AWlusCjFm3rY7MJqtZ8+ePYwbN46RI0cyd+5c1q1b578WCmcBpp6tJFiHMRI4R1XnAvW+12HAqbZYFWISExNDbYLjGM3uwE7NS5cuZeDAgTz//PN06tSJv/zlLxx//PG2lRcspp6tI9itQfbjf7Ok9opIkqr+R0SObemmSCWYPfajDaPZHdihOT8/n+nTp7N06VLA2/20cOFC+vbta3lZ7cHUs3UE28L4Fujve/8JcJuI3AjsssWqEFNSUtJ6pijDaHYHdmieN28eS5cupXPnzsyaNYtly5aFjbMAU89WEmwLYxKwz/d+CvAIkAJca4dRoSYzMzPUJjiO0ewOrNK8b98+YmK8vzdvvvlmSktLmTFjBt27d7fk+VZi6tk6Wm1h+GZIHQJ8BaCqm1R1sKoep6prbLEqxJSXl4faBMcxmt2BFZofe+wxzjzzTP/UzbS0NB5++OGwdBZg6tlKWnUYvrUY96tqVG4DEggTcMUdGM3t44YbbuCzzz7zj1mEO6aerSPYMYy3RGSoLRaEIWbetjswmoOjtraWefPm+c9zcnJYvHgxF154oZWm2YapZ+sI1mHEAP8QkXdF5CkRebLhsMWqEGPmbbsDo7l1vvjiC84880zuvPNOf9q6dev43e9+Z7VptmHq2TqCHfT+DphtiwVhiJmG5w6MZrhlxRY2/FjW4j2dLl/gnyIJ0KVLFxsssw9Tz9YRlMNQ1VttKT1MMQFX3IHRTKvOoikndE+10hxHMPVsHcG2MFxFaWlpxP2K6ihGsztoTvPK0cdQXl7OX//6V5599lnee+89Dj300BBYaD2mnq0j2DEMV5GVlRVqExzHaHYHzWl+9913GTRoEI8//jh1dXV8/PHHDltmH6aercM4jACUlpaG2gTHMZrdQXOa//CHP7Bjxw769evH6tWrueyyyxy2zD5MPVuH6ZIKQG1tbahNcByj2R001rx27VogGfBuVnfjjTdy3XXXERcXXV8Lbq9nKwm6hSEiV4nIShH51Hd+ioiMbMP9sSIyW0QKRKRcRF4VkWbbTSKyv4g8IyJFIlImIp+JyIHBltcRzLxtd+B2zQce+L9/p9zcXCZMmBB1zgJMPVtJUA5DRG4HrgMWA7/yJe8CbmpDWTfi3RL9RLyxNAD+3kx5icAqwAMcCnQBLgMq2lBeuzHztt2B2zSrKi+99BKqCkCvXr3813r37h0qs2zHbfUMoV+HcRXQX1V3i8j9vrTvgV4t3NOUMcAdqvo9gIhMBzaLyMGqurVJ3ivxOonrVLWhbfV1G8qirq6OzZs3+88zMjLIyMgI6t7k5OS2FBUVGM3RzbZt27j++ut5//33Abj88stDbJFzuKmeG7BLc7AOIw5oGEVR32tngvzFLyJpQA+8W6N7H6K6RUTKgKOArU1uOR34BnhERIYBBcCjqjqPINm9ezcjR/6vx2zs2LHMnDmTvLw8kpOTiY2NpaysjOzsbIqLi1FVsrOzyc/PR0QoKiqioqKCnJwcCgoKEBEyMjIoKCggNTWV+vp6Kisr6dq1K3l5ecTHx5OWlkZhYSFpaWl4PB6qqqr81xMSEkhJSaGoqIj09HSqqqqorq72X09MTCQpKYmSkhIyMzMpLy/H4/H4ryclJZGQkEBpaSlZWVmUlpZSW1vrv96aps6dOwM0qykmJobCwsKo0tRaPcXGxrJ79+6o0tS0nnbu3Mkrr7zCwoULqaqqokuXLuzbt49t27b9rNuirKwsYjS1tZ7i4+P56aefokpTa/WUkJDArl272qWpxe/yhuZpi5m8W4BUApOBfFXNEJG5QLKqtrrFuYh0B7YDvVT1h0bp24AZqvpsk/zvAkOA64GH8DqVfwITVXVJqwYDubm52jjqVFtaGNu2baNnz55B5Y0WjObo49tvv2XixIl89NFHAIwYMYJx48ZxzDHH+POc/fingHcdRrQS7fUciI5o3rhx4ydDhgzpH+haW+JhLMHbyujkaxm8D4wK8v6GvXbTmqR3AQItNS0HdqrqQt/5xyLyLN4xkKAcRlxcHH369AnSvJ+TnZ3drvsiGaM5uli3bh0XXHABHo+HAw44gDlz5nDuueeyd+/eUJvmONFcz81hl+agBr1VtVRVfwf0BU4GjlDV81U1qMm+qroHbwvDH9JVRHoBqcAXAW75jP91ff3sUcGU11GKi4udKCasMJqji+OOO46+fftyxRVXsG7dOs4991wgujU3h9FsHcHOkpojIseo6k5VXaeqO9pR1qPADSLyKxFJBe4DVgQY8AZ4GsgUkXG+6bhH450l9Y92lNtmgummizaM5simqqqKu+++m6KiIsC7l9CKFStYsGABqan/2/8pmjQHi9FsHcGuw0gC/iki34rIrSLSnjl49wJLgY+AnUAsvi4tEblMRPwD6Kq6DRgKjMbbZfUKMFNVX2xHuW3GNGHdQbRo/uCDDzjllFOYM2cON998sz99v/32+0XeaNHcFoxm6wi2S2occCDesYy+wEYR+VBEJgRbkKrWq+pUVc1S1RRVHaGqhb5rS1S1c5P8a1T1GFVNVtW+qvpg8LI6Rn5+vlNFhQ1Gc+RRVlbGlClTOP/88/n+++85/PDDGTNmTIv3RLrm9mA0W0fQyzp9oVrfBt4WkYOAp4D5wCJbLAshwUwvizaM5vCk1XgVh4+i/6z/zT256VPg009beaq7+vQjoZ6txi7NbdkaJElELhaRN4D/+pKvtsUqg8EAtD1eRXuIxBgXhtAQVAtDRJ4HzgO+Ap4HrlHV3XYaFkoqKirIzMwMtRmOYjSHN03XSdx2223sv//+/OlPfyI2Njbo57hxTUIk1bNV2KU52C6pr4CbGy+6i2ZycnJCbYLjGM3hzahRo7jmmms47bTTALjjjjva9ZxI0mwVRrN1BDvofZdbnAVAQUFBqE1wHKM5/Gg8NXL58uX85S9/6fB0yXDXbAeANL9jAAAgAElEQVRGs3U06zBE5MtG738Qke8DHbZYFWJEJNQmOI7RHF788MMPDB8+3H9+zjnn8Nxzz3XY5nDWbBdGs3W01CX150bvR9tSepgS7J5T0YTRHB7U19fz8MMPc/fdd1NVVUX/873pS5YsseRLIBw1243RbB3NtjBU9f1Gp6mquqrpAaTYYlWIMU1YdxCOmsvLy7n//vupqqriwgsv9Kdb9YsxHDXbjdFsHcFOq32mmfQnrTIknGi8lYJbMJpDh8fjwePxANClSxcWLVrECy+8wCOPPGJ5WeGi2UmMZuto0WGISA8R6QHEiEj3hnPfMRioscWqEFNfXx9qExzHaA4Nn3zyCaeffjpz5szxp5199tmcffbZtpQXDpqdxmi2jtZaGFuBH/DuJbXNd95wvAS0b25fmFNZWRlqExzHaHaWvXv3cuutt/Lb3/6W//znP7z55pv+VoadmHp2B3Zpbs1hxAOdgPW+9w1HnKrur6oP2WJViDFB491BqDTn5uZy8skn8+CD3u3Rxo8fz+rVq0lISLC9bFPP7sAuzS06DN+GgXWqOsj3vuHYZ4s1YYIJGu8OnNZcU1PDpEmTGDZsGFu3buWII47gnXfe4fbbbw+4s6wdmHp2B3ZpbnZarYi8parn+d6/RzPBi1T1DFssCyHx8fGhNsFxjGb7SUhIYMeOHcTHxzN16lQmTpzoSKuiMaae3YFdmltah9E49sSzzeaKQtLSmkaSjX6MZnsoLCxk79699OjRAxFh/vz5VFRUcNhhh9lediBMPbsDuzQ36zBUdXGj90/YUnqYUlhYSHJycqjNcBSj2VpUlVdffZUbb7yRQw45hGXLlhETE0O3bt1sKS9YTD27A7s0Bxui9Q8icpjvfV8RWSUiK0XkEMstCgPMLxJ3YJfmHTt2cMkllzBmzBiKi4tJSEigrMz+bcqDwdSzO7BLc7AL9+4GSnzv5wJfABuAqJwl5cT0xnDDaO44+/bt4+mnn+akk05i5cqVpKamsnDhQl577TW6dOliaVntxdSzO7BLc7Dbm++vqvki0gk4BbgQqAOics19VVVVqE1wHKO5Y6gqF110EatWrQLgvPPOY9asWRxwwAGWlWEFpp7dgV2ag21hFIlIL+Ac4GNVrQESgKjcBtLM23YHVmoWEU455RSys7N58sknWbx4cdg5CzD17BZCsg6jEXcCG/HG8W7Yw2AI3q6pqMPM23YHHdX89ddfs2LFCv/5ddddx/r16xk+fHjYbqlt6tkdOL4OozGq+oSIvOR7X+5L/hi41BarQozTc+PDAaM5eGpqapg7dy4LFiwgOTmZ9evXk5OTQ1xcHOnp6RZbaS2mnt2BXZqDHcMAb2vkPBE5CNgJLFfVPbZYFWJSUqJy1/YWMZqD46OPPmLChAls2rQJgJEjR5KUlGS1abZh6tkd2KU52Gm1JwBbgOuBE4CJwGYROdEWq0JMUVFRqE1wHKO5ZSorK7n55ps555xz2LRpE7179+att95i9uzZEbV9tqlnd2CX5mBbGIuAiaq6pCFBRC71pUed0wj3bgU7MJpbZvTo0axYsYLY2FjGjx/PtGnTIqpl0YCpZ3dgl+ZgB70PBZ5vkvYiEJUL98w0PHfQFs1Tpkzh6KOP5t133+W2226LSGcBpp7dQqin1W4GRjZJGwF8b6054UF1dXWoTXAco/nnLF++nFtvvdV/3r9/f1avXs3RRx/thGm2YerZHdilOdguqUnAMhGZgDeQ0sHAEcDvbbEqxJh52+4gkObdu3dz44038vrrrwMwdOhQBg4cCFgXVzuUmHp2ByFdh6Gqa4E+wOPA18BjQF9VzbXFqhBj5m27g8aaVZUXX3yRgQMH8vrrr5OcnMy9997LiSdG1xCd2+vZLYR0HQaAqhaKyNvAgcAuVS20xaIwIDExMdQmOI6bNe/YsYPJkyfz7rvvAjB48GAWLFhAjx49QmmeLbi5nt2EXZqDnVbbzRdEaSewCtgpImtEpHuwBYlIrIjMFpECESkXkVdFJCuI+/4kIioitwRbVkeJ1AHNjuBmzQ888ADvvvsuXbp04cEHH+TVV1+NSmcB7q5nN2GX5mAHvZ/B2xWVoaoZQCbwJbC4xbt+zo3AMLzTcBuCAvy9pRtEpCcwxVeWY5SUlLSeKcpwm+a6ujq/5ptvvpkrr7ySdevWcckll0TFWEVzuK2ewWi2kmAdxvHAJFUtA1DVUmAqcFwbyhoD3Keq3/vunw6cIyIHt3DPE8AMoLgN5QDeL4TNmzf7j+Li4B+RmZnZ1uIiHrdorqurY8GCBZx22mn+ONqpqanMnz+fnJycEFtnP26p58YYzdYR7BjGBrzOYX2jtH6+9FYRkTSgB/BJQ5qqbhGRMuAoYGuAe8YCe1X1RRH5U5B2+tm9ezcjR/5vJvDYsWOZOXMmeXl5JCcnExsbS1lZGdnZ2RQXF6OqZGdnk5+fT21tLenp6VRUVJCTk0NBQQEiQkZGBgUFBaSmplJfX09lZSVdu3YlLy+P+Ph40tLSKCwsJC0tDY/HQ1VVlf96QkICKSkpFBUVkZ6eTlVVFdXV1f7riYmJJCUlUVJSQmZmJuXl5Xg8Hv/1pKQkEhISKC0tJSsri9LSUmpra/3XW9PUuXNngGY11dXVUV1dHVWamtbT+++/zx133ME333wDwKuvvsqIESPCWlMDe/futeSzt2/fPiorK8O6nqz+7Klq1GlqrZ5UlbKysnZpaglR1dYziTwAXAYsBX4EugPn44317Y+Joap3NHN/d2A70EtVf2iUvg2YoarPNsnfA/gAGKCqO0VkDfCuqt7ZqrE+cnNztfHAT0ZGBhkZGUHdu23bNnr27BlsUVFBNGuurq5mzpw5LFy4kPr6erp37868efPo06dP2Gs++/FPAVg5+hhLnhfN9dwcRnPb2Lhx4ydDhgzpH+hasC2MNGAZ3vgXDaOBy4AuvgOgJc/TsMNt07iBXYBAsSsfB+5U1Z1B2vcL4uLi6NOnT7vuNfO2o4cNGzYwfvx4vvvuO0SEMWPGcMstt9C5c2dqampCbZ7jRGs9t4TRbB3Bbm9+eUcKUdU9IrIdOBb4DMAXkCmVwDE1zgKOE5G7fOdpwPEi8ltVPaUjtgRDXl6e636RRKvmnTt38t1339G3b18WLlzIgAED/NeiVXNLGM3uwC7NbdnevKM8Ctzgm55bBNwHrFDVrQHyNp2u+zKQizeeuO2YaXiRzfbt2/3TYocPH47H42HYsGG/mJseTZqDxWh2B6GeVmsF9+IdA/kI73qOWGAUgIhcJiIVDRlVdUfjA6gBylQ13wlDTcCVyKSkpIRx48Zx/PHH+we2RYSLLroo4EKmaNDcVoxmd2CXZscchqrWq+pUVc1S1RRVHdGwWlxVl6hqs0P0qjq4LQPeHaW0tNSposKGSNf85ptvMnDgQJ5//nliYmL46quvWr0n0jW3B6PZHdil2ckuqYghK6vVBehRR6Rqzs/PZ/r06SxduhSAAQMGsHDhQvr27dvqvaHQfMuKLWz4MdA8D2eI1HruCEazdQTdwhCR00XkERF53Xd+rIicZotVIcb8IokMVq9ezcCBA1m6dCmdO3dm9uzZLFu2LChnAaHR3B5ncUJ36yL6RWI9dxSj2TqCamGIyHV4V3Y/CVzsS/YAdwEn22JZCKmtrQ21CY4TiZp79+6Nx+NhyJAhzJ8/n27durV+UyNCqdmqdRVtJRLruaMYzdYRbAtjCnCmbxxhny/tP8DhtlgVYsy87fBk3759vPbaa+zb5/0I9uzZk/fee4+XXnqpzc4CIkOz1RjN7iCk8TCAFLyBk+B/C/Ti8LYyog6zf374sWnTJoYOHcrVV1/N4sX/2/Oyb9++7d4sMNw124HR7A7s0hysw1iLt0uqMeOA9601JzxITk4OtQmOE66aa2trmTdvHqeddhobNmwgJyfHsl9P4arZToxmd2CX5mBnSY3HG6L1GiBFRL7G27oYaotVISY2NjbUJjhOOGr+/PPPmTBhAl9+6d3dftSoUdxxxx106dKllTuDIxw1243R7A7s0hzs1iA7ReQ4YCDevaR+BNapar0tVoWYsrIy0tPTQ22Go4Sb5rVr13LBBRdQX19Pz549mT9/PoMHD7a0jHDT7ARGszuwS3NbQrTuw7uD7AeWWxFmZGdnh9oExwk3zSeeeCK/+c1vGDBgADNmzLCliW2V5lCvrWgL4VbPTmA0W0ew02p/oJndaFW1l6UWhQHFxcX+4DpuIdSay8vLmTVrFuPHj2f//fcnPj6ef/7zn7Zu62CV5rY6CyvXVbSVUNdzKDCarSPYFsboJucH4B3XeN5ac8KDYGKERBuh1PzOO+8wefJkdu7cSV5eHo899hhg/x5AVmsO1dqKtmA+2+7ALs3BjmGsapomIquA5cACq40KNaYJ6wzFxcXMmDGDF198EYB+/fpx/fXXO1a+qWd3YDRbR0c2H6wCoq47Crz7E7kNJzWrKq+//joDBw7kxRdfJDExkZkzZ7Jy5UqOPPJIx+ww9ewOjGbrCHYM47YmSfsB5wErLbcoDAgmtm204aTm//73v1x99dWoKoMGDWLBggX07t3bsfIbMPXsDoxm6wh2DKPpbm6VwIPA05ZaY4haVNW/IvvQQw9l8uTJHHTQQVxxxRXExDgZlsVgMLSXVh2GiMQC7wAvqWq1/SaFnoqKCjIzM0NthqPYqXnr1q1MmjSJ8ePHc8YZZwAwY8YMW8oKRMvTXrc7Zkc4YD7b7sAuza06DFWtF5H7VXVxa3mjhZycnFCb4Dh2aK6vr+fRRx/lrrvuYu/evZSWlnL66ae3e++n9uLEGolQTpVtC+az7Q7s0hxsl9RbIjJUVZfbYkWYUVBQQPfuTcOKRzdWa/7222+ZMGECH3/8MQAjRozgnnvucdxZNKbptNcff/zR1LMLMJqtI1iHEQP8Q0TW4t0WxD/JV1X/aLlVISaUX2qhwirNHo+HhQsXMnfuXDweDwcccABz5szh3HPPteT5VmLq2R0YzdYRrMP4DphtiwVhSEZGRqhNcByrNFdVVfHUU0/h8Xi44ooruP3220lLS7Pk2VZj6tkdGM3W0aLDEJFLVPV5Vb3VltLDlIKCAnr27BlqMxylI5r37t1LTEwMiYmJpKWl8cADDxAXF8epp55qsZXWYurZHRjN1tHafMZHLC8xAkhNjYwBTCtpr+YPPviAU089lVmzZvnTzjjjjLB3FmDq2S0YzdbRmsNwX+cf3tk9bqOtmsvKypgyZQrnn38+33//Pe+++y4eT2QFYDT17A6MZutobQwjVkROpwXHoaqrrTUp9FRWVpKVlRVqMxylLZpXrlzJ5MmT2bVrF/Hx8UyaNInJkyfbvllgY6zYUtzUszswmq2jNYfRCXiC5h2GEoX7SZmg8YGprq5m4sSJvPzyywAce+yxLFq0iCOOOMJu836BFVuKm3p2B0azdbTmMCqjMd5Fa+Tl5blukCwYzZ06daKsrIykpCRuvvlmrr322pCHv+zIluKmnt2B0WwdQUfccxPx8fGhNsFxmtO8a9cuampq+NWvfoWIMHfuXKqrq+nVK/J/R5h6dgdGs3WYQe8AhOu6ATtpqllVeeaZZxg4cCDXXnutfxDtwAMPjApnAaae3YLRbB0tOgxVTbGl1DCnsLAw1CY4TmPNP/zwA8OHD2fSpEmUl5eTmZlJZWVlCK2zB7fXs1swmq3DsX2lRSRWRGaLSIGIlIvIqyIScBhfRIaKyGoRKRSREhHJFZFTnLLVrb9I6uvrefDBBzn55JPJzc0lKyuLxx9/nCVLlkTlXHa31rPbMJqtw8lABDcCw4ATgW6+tL83kzcduB/oA2QDzwFvi4gjO4hF2noCK6ipqWHEiBHceuutVFVVceGFF7Ju3TpGjBgRtXvxuLGejWZ3YJdmJwe9xwB3qOr3ACIyHdgsIger6tbGGVV1SZN7HxKRO4D+eDc/bJW6ujo2b97sP8/IyAh6f5Wqqqqg8kUT1dXVnH322WzZsoV58+Zx9tlnd+h5VqyTsBs31rPR7A7s0uyIwxCRNKAH8ElDmqpuEZEy4Chgayv3HwVkAl8FW+bu3bsZOXKk/3zs2LHMnDmTvLw8kpOTiY2NpaysjOzsbIqLi1FVsrOzyc/PJzExkaKiIioqKsjJyaGgoAARISMjg4KCAlJTU6mvr6eyspKuXbuSl5dHfHw8aWlpFBYWkpaWhsfjoaqqyn89ISGBlJQUioqKSE9Pp6qqiurqav/1xMREkpKSKCkpITMzk/Lycjwej/96UlISCQkJlJaWkpWVRWlpKbW1tf7rrWlqCNnYWNNXX31FRUUFJ5xwAklJSYwcOZLBgwfTp08ftm3b1iFNTjiLfjkdq6f99tuP3bt3h309WfnZS05OJj8/P6o0tVZPKSkp/PTTT1GlqbV6Sk1NZdeuXe3S1BKiqq1m6ii+rqTtQC9V/aFR+jZghqo+28K9+wNrgX+o6o3Blpmbm6uJiYn+87a0MLZt2xbV87b37t3L3XffzcMPP0xycjLr1q2jtrbWUs1nP/4p0LF1EnYT7fUcCKPZHXRE88aNGz8ZMmRI/0DXnOqSKve9Nh2J6QI0+1NURA7EGx52JXBTWwqMi4ujT58+bbnFj5NbXDhNbm4uEydOZOvWrcTExHDVVVeRnp7Onj17Qm2a40RzPTeH0ewO7NLsiMNQ1T0ish04FvgMQER6AanAF4HuEZGDgVXAa6o61Qk7G0hJib7ZxGVlZdx2220sXuyNtHvEEUdw//33c8wx3haAGzdoi8Z6bg2j2R3YpdnJWVKPAjeIyK9EJBW4D1jRdMAbQEQOw9sN9bzTzgKgqKjI6SJtZ/To0SxevJj4+HhuuukmVq9e7XcWEJ2aW8NodgdGs3U46TDuBZYCHwE7gVhgFICIXCYiFY3y3gAcBFwvIhWNjsucMDQ9Pd2JYhzl5ptv5sQTT2TNmjVMmzbtF03WaNTcGkazOzCarcMxh6Gq9ao6VVWzVDVFVUeoaqHv2hJV7dwo71WqKqraucnRdLqtLUT6NDxV5eWXX2batGn+tH79+rF8+XIOP/zwgPdEuub2YDS7A6PZOszmgwGorq4OtQntZseOHUydOpWVK1cCMGLECAYOHAi0HBg+kjW3F6PZHRjN1uFkl1TEEIn75+/bt4+nnnqKk046iZUrV5KamsqiRYsYMGBAUPdHouaOYjS7A6PZOozDCEBeXl6oTWgTW7ZsYdiwYUyZMoWKigqGDh3KunXrGDVqVNDbekSaZiswmt2B0WwdpksqAI0X/EUCjz32GB988AHZ2dncd999DBs2rM37P0WaZiswmt2B0WwdxmEEICkpKdQmtEpNTQ2dOnUCYMaMGYgI06ZNC3o1e1MiQbPVGM3uwGi2DtMlFYCSkpJQm9AsNTU13HXXXZxyyin+GBUpKSncc8897XYWEN6a7cJodgdGs3UYhxGAzMzMUJsQkA0bNnDaaacxd+5cNm/ezKpVqyx7drhqthOj2R0YzdZhHEYAysvLW8/kIBUVFdx0002ce+65/Pe//6V379689dZb/P73v7esjHDT7ARGszswmq3DjGEEIJwCrqxdu5Y///nPbN++ndjYWMaPH8+0adMC9lF2PAaFu2aThFM9O4XR7A6iIYBSxBBO87ZLSkrYvn07v/nNb1i0aBFHH310s3nDLWDRCd3DO6xrONWzUxjN7sAuzcZhBCAvLy+k++f/97//5ZBDDgHg/PPP56mnnmLo0KHEx8cHdX97YlC4MWZAqOs5FBjN7sAuzWYMIwChmoa3e/durrrqKk4++WS+/PJLf/qwYcOCdhbtxUw9dAdGszsw02odxOmAK6rKCy+8wIABA3jjjTfo1KkTW7ZscdQGE2TGHRjN7sAuzcZhBKC0tNSxsn788UcuvPBCrrvuOvbs2cPpp5/OBx98wPDhwx2zAZzVHC4Yze7AaLYOM4YRgKysLEfKWblyJaNHj6aiooIuXbpw1113cfHFF7d5Ww8rcEpzOGE0uwOj2TpMCyMATv0iOeKIIwDvwPa6deu45JJLQuIswPwKcwtGszswLQwHqa2tte25L7/8MhdffDExMTF069aNtWvX0qNHD1vKa6ttbsNodgdGs3UYhxEAO+Ywf/HFF0yYMIEvvviCyspKrrnmGoCwcBZg5qq7BaPZHZh4GA5i5V7y1dXV3HnnnQwZMoQvvviC7t2706dPH8uebxUmZoA7MJrdgYmH4SDJycmWPGf9+vVMnDiR7777DhFhzJgx3HLLLXTu3Ln1mx3GKs2RhNHsDoxm6zAOIwCxsbEdfsb777/PiBEjUFX69u3LwoULgw6XGgqs0BxpGM3uwGi2DtMlFYCyso7vyTRo0CD69+/P5MmTef/998PaWYA1miMNo9kdGM3WYVoYAcjOzm7zPSUlJdx1111MmTKFAw44gLi4ON566y3i4iLjT9wezZGO0ewOjGbrMC2MABQXF7cp/5tvvsnAgQN58sknmTFjhj89UpwFtF1zNGA0uwOj2Toi5xvNQVQ1qHx5eXlMnz6dZcuWATBgwABuuukmO02zjWA1RxNGszswmq3DtDAC0FpzTlV57rnnGDhwIMuWLaNz587Mnj2bZcuW0bdvX4estBbTbHcHRrM7MF1SDpKfn9/i9U2bNjFhwgRKS0sZMmQI//73v7n66quJiYncP2drmqMRo9kdGM3WYbqkAhBonYSq+vd5Ouyww7jhhhvo0aMHf/jDH0K2/5OVhOPaELsxmt2B0WwdkfuT2EE2bdrEueeey8qVK/1p06ZN46KLLooKZ2EwGAzBYBxGACoqKgDvBl5z587ltNNOY8OGDcyaNStqB9AaNLsJo9kdGM3W4ZjDEJFYEZktIgUiUi4ir4pIs5u2i8g5IvK1iFSJyFcicrZTtubk5PDZZ58xZMgQ7rrrLjweD6NGjeKVV16J2hZFTk5OqE1wHKPZHRjN1uFkC+NGYBhwItDNl/b3QBlFpBfwD+AeIM33+pqIHGy3kVVVVcyYMYOzzjqLr776ip49e/KPf/yDRYsW0aVLF7uLDxkFBQWhNsFxjGZ3YDRbh5OD3mOAO1T1ewARmQ5sFpGDVXVrk7xXAp+o6rO+8yUicq0v/fZgCqurq2Pz5s3+84yMDDIyMlq9b9iSb+HwURxzzyh/2t2b4e7NnwZTbMQSrS2nljCa3YHRbB2OOAwRSQN6AJ80pKnqFhEpA44Ctja55ejGeX1s9KUHxe7duxk5cqT/fOzYscycOZO8vDySk5OJjY2lrKyM7OxsiouLUdWomK/9m6x49u7d+zNN+fn5/lkTFRUV5OTkUFBQgIiQkZFBQUEBnTp1orCwkMrKSrp27UpeXh7x8fGkpaVRWFhIWloaHo+Hqqoq//WEhARSUlIoKioiPT2dqqoqqqur/dcTExNJSkqipKSEzMxMysvL8Xg8/utJSUkkJCRQWlpKVlYWpaWl1NbW+q+3VE/BaEpNTaW+vr5ZTYmJiezevTuqNLVWT0lJSeTn50eVptbqKTk5mZ9++imqNLVWT507d2bXrl3t0tQS4sQgroh0B7YDvVT1h0bp24AZjVoSDemrgLWq+pdGabcDg1T1zGDKzM3N1cTERP95sC0MgG3bttGzZ8+g8kYLRrM7MJrdQUc0b9y48ZMhQ4b0D3TNqS6pct9rWpP0LkCgbRXL25A3IHFxce0OVJSamtqu+yIZo9kdGM3uwC7Njgx6q+oevC2MYxvSfAPbqcAXAW75vHFeH8f40m2nvr7eiWLCCqPZHRjN7sAuzU7OknoUuEFEfiUiqcB9wIoAA94Ai4H+InKJiMSLyCXAccAzThhaWVnpRDFhhdHsDoxmd2CXZicdxr3AUuAjYCcQC4wCEJHLRMS/0kRVtwAjgFvwdkPdAlzQjHOxHBM03h0Yze7AaLYOxxyGqtar6lRVzVLVFFUdoaqFvmtLVLVzk/z/VNUjVTXJ97oy8JOtxwSNdwdGszswmq3DbA3ShOLiYh599FFXBV0xmt2B0ewO7NRsHEYTiouLeeihh1z3ATOaox+j2R3Yqdk4DIPBYDAEhSML90LBqlWrCoBtbb2vrq4utri4OCcjIyM/Li7OFfPxjGajOVoxmtulueeQIUMCbnsRtQ7DYDAYDNZiuqQMBoPBEBTGYRgMBoMhKIzDMBgMBkNQGIdhMBgMhqAwDsNgMBgMQWEchsFgMBiCwjgMg8FgMASFcRgGg8FgCArjMAwGg8EQFK50GCISKyKzRaRARMpF5FURyWoh/zki8rWIVInIVyJytpP2WkFbNIvIUBFZLSKFIlIiIrkicorTNneUttZzo/v+JCIqIrc4YadVtONzvb+IPCMiRSJSJiKficiBTtrcUdqheaqIbPHl/U5ErnPSXisQkYt9/5NlIlIXRP7+IrJBRPb6tI9qb9mudBjAjcAw4ESgmy/t74Ey+kLJ/gO4B2+c8XuA10TkYNuttJagNQPpwP1AHyAbeA54W0S6222kxbRFMwAi0hOYAnxpr2m20JbPdSKwCvAAhwJdgMuAikD5w5i2aP49cDtwmaqmAFcAs0XkLCcMtZAS4G/A9a1lFJE04G3gVbz/19cCD4vIwHaVrKquO/BuSnh1o/PegAIHB8h7O5DbJC0X+EuoddiluZn7C/BGPQy5Fjs1A+8CFwFrgFtCrcEuvcBY4EcgPtR2O6h5MvDvJmnrgKmh1tFO7YOBulbyXAVsx7dvoC/t78BT7SnTdS0Mn8ftAXzSkKbekLBlwFEBbjm6cV4fG33pEUE7NDe9/yggE/jKLhutpj2aRWQssFdVX3TESAtph97TgW+AR3xdUt+KyGRHjLWIdmh+AUgVkUEiEuPrZj0E+KcT9oaIo4GN6vMUPtr9/RVniUmRRarvtbRJ+p5G1xqT0kzeIy22y07aqtmPiOwPvALMUtXvbLDNLtqkWUR64I0dP8Bmu+yirXWcBZXob7wAAAdYSURBVAzB261xLd4v2H+KSL6qLrHNSmtpq+bdeD/L7/G/7vjrVTVifgi1g+a+v1r8v28O17UwgHLfa1qT9C54f5kEyh9s3nClrZoB8A2AvgesBG6yxzTbaKvmx4E7VXWnrVbZR3s+1ztVdaGqelT1Y+BZvOMBkUJbNd8KXAr0A+Lx/sqeJCJX22Zh6LH0+8t1DkNV9+Dt0zu2Ic03sJ0KfBHgls8b5/VxjC89ImiHZnyD+rnA26r65yZN2rCnHZrPAu72zQwrBAYBN4lIrhP2dpR26P0Mb1//Lx5li4E20A7NxwGvqeo36uVr4HXgd07YGyI+x/t91Zj2f3+FeuAmRINFM4BNwK/wfrheBv7ZTN7ewF7gEry/Si4BKglysDhcjjZqPgzYgfcXd8htd0hztybHOmAWkBNqHTbp7en7XI8DYvH+2i4ALgq1Dhs13+TL29d3fjiwBbg11DraqDkWSATOBup87xNpNLDdKG8XX71OAxLwdkNWAAPbVXaoxYfwDz4HKMTbZPsHkOW7dhlQ0ST/OcDXQJXv9exQa7BTM/AU3l+aFU2Oy0Ktw856bnLvGiJvllRbP9eDgU/x/gD6DhgXag12asY7ZnsvsNX3ed4OzCXCZooB/+f7/2x6HAyc4tPWo1H+44ENvu+v74FR7S3bhGg1GAwGQ1C4bgzDYDAYDO3DOAyDwWAwBIVxGAaDwWAICuMwDAaDwRAUxmEYDAaDISiMwzAYDAZDUBiHYYg6RORZEZkZajtaQ0SuFJG3W7g+WES+dtImg6EljMMwhC0istUXtKqi0RFRAX5aQlWfUdVzAUQkzhe06eBG19eoasg3uQxkm8GdGIdhCHfOV9XOjY5doTYo2hARN+5abWgHxmEYIg5fLINXRCRPRPaIyBoRObyZvPuLyHJfvmIR+Veja91E5DVfeM8fRGRcC2U+KyIPisgqX3jP9xpHIBSRk0XkYxEp9YXDPLHRtat9raVyEfleRC72pY8WkTW+bA12fe1rSf0/ETlTRLb68t4iIi80selBEZnne99FRJ4SkZ9EZIeI3CEiAf+/ReROEXlRRJ4XkXJglIgMFJH1vr/TTyKySETim7PN95zfi8jnvnvWisivm/v7GaID4zAMkcoyoC/QFW9gp+ZCr07Du39Oti/vreCNBe17xkfAQXh3q50mIkNaKHMUcBveWBLfNJQp3hjSb+HdlygTWAQsF5F0EUkF5gFnqTcs6CAC76R6qu/1SF9L6tUm158Dficiyb4y44ALfeng3Zq8Cu9mmf2B8/BGW2uOC3z3pgEv4t3EbqJP2yC8+6eNbc42ETkeeAwY7dP8JPCGiCS0UKYhwjEOwxDuvO77BbtHRF4HUNV9qvq0qparajUwEziu4cu0CbXAgXg3Y/Oo6vu+9AFAqqre7UvfDDwBXNyCLUtV9QNVrQFuBk4VkQOA84GvVfV5Va1T1WfxOqnzfPcp8GsRSVTVn1T1m7b+EVT1e7yOsSFexVnAHlX9WEQOwrsL6SRV3auqecCCVrSsVdWlvr9llap+pKof+uz/HngUOK2F+8cAf/PdV6+qT/rSj2+rNkPkYByGIdwZrqpdfMdw8LYORGSWr3unDNjsy5sV4P578cZ9XiUiW0Rkmi+9J9CjkTPaA0zH2wppjh8b3qhqKd5IZgf6jm1N8m4DDlLVMrxb4o8D8kRkmYgc0gb9jXnO9yzwBgJqiIzXE+gE5DfS8iCQE4wWABE5TETe8nXzlQF3EPjv2UBP4IYmf78D8LbWDFGKcRiGSOQKYChwBt4ulT6+dGmaUVXLVHWSqh4MDMf7JXca3i/M7xo5oy6qmqKq57dQbuMxizRf2bt8R88meXsAO302vK2qZ+L9Qt0MPBLg2cFsG/0icKaIdMPb0mjojvoRb2yLjEZaUlW1pXjtTct7BG8Lpo+qpuLtepNm8jaUeXuTv99+qvpSEDoMEYpxGIZIJAWoAYqA/YC7mssoIueLSG8REbwtgnrfsQ7wiMgUkf/f3h26RBQEcRz/TlJUrltFMIgW0WAwGdVgEUQxixYRMWowWkyixXLcn2AUhFPwDzAoNoOIhyaLeIxh9jg47j22HR6/T73H7rxy82Z2YWwwVS1TZjZTsu9yOhweAE6Its4bcRYyaWZr6QrqOpHErs1sNMUwBPwQsyeanQu7ezO9z1jR5u7+DtSJeSVPnmasu/srcAucmlklXQoYN7OForW6aM1+/k4XCFrnF0WxXQI7ZjZrYSS9Z7e2oPQJJQz5j65of9k/Avclz04AN8RQmTvgzN3r7v5LVClzxECdBvGVXSlZq0okigYwDWwCuPsHsAIcEn+se8CSu38SA34OgLf02zywW7D+EVBLLZ7VgmdqwCLt6qJlAxgmDuO/iMlzZe21TvvAFjGE6IKoZgpjc/cHYBs4T/s9pxikj2mAkkgGM6sCL+5+3OtYRHpFFYaIiGRRwhARkSxqSYmISBZVGCIikkUJQ0REsihhiIhIFiUMERHJooQhIiJZ/gAJ8DFy0TBz/AAAAABJRU5ErkJggg==\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"X = Xy.drop('bin',axis=1).values\n",
"y = Xy['bin'].values\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5,\n",
" shuffle=False)\n",
"\n",
"n_estimator = 10000\n",
"rf = RandomForestClassifier(max_depth=2, n_estimators=n_estimator,\n",
" criterion='entropy', random_state=RANDOM_STATE)\n",
"rf.fit(X_train, y_train)\n",
"\n",
"# The random forest model by itself\n",
"y_pred_rf = rf.predict_proba(X_test)[:, 1]\n",
"y_pred = rf.predict(X_test)\n",
"fpr_rf, tpr_rf, _ = roc_curve(y_test, y_pred_rf)\n",
"print(classification_report(y_test, y_pred, target_names=['no_trade','trade']))\n",
"\n",
"plt.figure(1)\n",
"plt.plot([0, 1], [0, 1], 'k--')\n",
"plt.plot(fpr_rf, tpr_rf, label='RF')\n",
"plt.xlabel('False positive rate')\n",
"plt.ylabel('True positive rate')\n",
"plt.title('ROC curve')\n",
"plt.legend(loc='best')\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {
"ExecuteTime": {
"end_time": "2018-06-12T14:52:40.235268Z",
"start_time": "2018-06-12T14:52:39.912077Z"
}
},
"source": [
"### (c) What is accuracy of predictions from primary model if the secondary model does not filter bets? What is classification report?"
]
},
{
"cell_type": "code",
"execution_count": 51,
"metadata": {
"ExecuteTime": {
"end_time": "2019-03-01T17:55:02.893334Z",
"start_time": "2019-03-01T17:54:56.012559Z"
},
"scrolled": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"2019-03-01 10:54:56.611727 9.09% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.03 minutes.\r",
"2019-03-01 10:54:56.612903 18.18% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:56.613181 27.27% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.01 minutes.\r",
"2019-03-01 10:54:56.615975 36.36% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:56.624309 45.45% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:56.626470 54.55% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:56.629972 63.64% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:56.632532 72.73% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:56.633112 81.82% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:56.639612 90.91% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\r",
"2019-03-01 10:54:56.657881 100.0% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.0 minutes.\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" t1 trgt\n",
"2018-02-13 13:43:37 2018-02-14 13:53:59 0.014365\n",
"2018-02-14 10:30:48 2018-02-15 10:42:27 0.012136\n",
"2018-02-14 13:36:02 2018-02-15 13:42:09 0.011688\n",
"2018-02-15 09:31:56 2018-02-16 09:42:36 0.011244\n",
"2018-02-15 14:05:41 2018-02-16 14:15:08 0.010183\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 929 entries, 2009-10-05 14:55:48 to 2018-02-15 14:05:41\n",
"Data columns (total 2 columns):\n",
"t1 929 non-null datetime64[ns]\n",
"trgt 929 non-null float64\n",
"dtypes: datetime64[ns](1), float64(1)\n",
"memory usage: 21.8 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n",
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" ret bin\n",
"2018-02-13 13:43:37 0.010108 1.0\n",
"2018-02-14 10:30:48 0.010876 1.0\n",
"2018-02-14 13:36:02 0.005056 1.0\n",
"2018-02-15 09:31:56 0.003964 1.0\n",
"2018-02-15 14:05:41 0.004842 1.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 929 entries, 2009-10-05 14:55:48 to 2018-02-15 14:05:41\n",
"Data columns (total 2 columns):\n",
"ret 929 non-null float64\n",
"bin 929 non-null float64\n",
"dtypes: float64(2)\n",
"memory usage: 21.8 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n",
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" vol ma_side srl_corr\n",
"2016-07-07 14:28:00 0.012624 -1.0 0.251865\n",
"2016-07-08 09:30:57 0.011944 1.0 0.238590\n",
"2018-02-06 10:18:08 0.013317 -1.0 0.123961\n",
"2018-02-07 15:28:09 0.024870 1.0 -0.005597\n",
"2018-02-13 09:30:00 0.017363 -1.0 0.198935\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 102 entries, 2009-10-29 13:40:22 to 2018-02-13 09:30:00\n",
"Data columns (total 3 columns):\n",
"vol 102 non-null float64\n",
"ma_side 102 non-null float64\n",
"srl_corr 102 non-null float64\n",
"dtypes: float64(3)\n",
"memory usage: 3.2 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n",
"-------------------------------------------------------------------------------\n",
"dataframe information\n",
"-------------------------------------------------------------------------------\n",
" vol ma_side srl_corr bin\n",
"2016-07-07 14:28:00 0.012624 -1.0 0.251865 1.0\n",
"2016-07-08 09:30:57 0.011944 1.0 0.238590 1.0\n",
"2018-02-06 10:18:08 0.013317 -1.0 0.123961 1.0\n",
"2018-02-07 15:28:09 0.024870 1.0 -0.005597 -1.0\n",
"2018-02-13 09:30:00 0.017363 -1.0 0.198935 1.0\n",
"--------------------------------------------------\n",
"\n",
"DatetimeIndex: 102 entries, 2009-10-29 13:40:22 to 2018-02-13 09:30:00\n",
"Data columns (total 4 columns):\n",
"vol 102 non-null float64\n",
"ma_side 102 non-null float64\n",
"srl_corr 102 non-null float64\n",
"bin 102 non-null float64\n",
"dtypes: float64(4)\n",
"memory usage: 4.0 KB\n",
"None\n",
"-------------------------------------------------------------------------------\n",
"\n",
" precision recall f1-score support\n",
"\n",
" no_trade 0.39 0.43 0.41 21\n",
" trade 0.57 0.53 0.55 30\n",
"\n",
" micro avg 0.49 0.49 0.49 51\n",
" macro avg 0.48 0.48 0.48 51\n",
"weighted avg 0.50 0.49 0.49 51\n",
"\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"minRet = .01 \n",
"ptsl=[0,2]\n",
"bb_events = getEvents(close,tEvents,ptsl,target,minRet,cpus,t1=t1)\n",
"cprint(bb_events)\n",
"\n",
"bb_bins = getBins(bb_events,close).dropna()\n",
"cprint(bb_bins)\n",
"\n",
"features = (pd.DataFrame()\n",
" .assign(vol=bb_events.trgt)\n",
" .assign(ma_side=ma_side)\n",
" .assign(srl_corr=srl_corr)\n",
" .drop_duplicates()\n",
" .dropna())\n",
"cprint(features)\n",
"\n",
"Xy = (pd.merge_asof(features, bb_bins[['bin']], \n",
" left_index=True, right_index=True, \n",
" direction='forward').dropna())\n",
"cprint(Xy)\n",
"\n",
"### run model ###\n",
"X = Xy.drop('bin',axis=1).values\n",
"y = Xy['bin'].values\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5,\n",
" shuffle=False)\n",
"\n",
"n_estimator = 10000\n",
"rf = RandomForestClassifier(max_depth=2, n_estimators=n_estimator,\n",
" criterion='entropy', random_state=RANDOM_STATE)\n",
"rf.fit(X_train, y_train)\n",
"\n",
"# The random forest model by itself\n",
"y_pred_rf = rf.predict_proba(X_test)[:, 1]\n",
"y_pred = rf.predict(X_test)\n",
"fpr_rf, tpr_rf, _ = roc_curve(y_test, y_pred_rf)\n",
"print(classification_report(y_test, y_pred, \n",
" target_names=['no_trade','trade']))\n",
"\n",
"plt.figure(1)\n",
"plt.plot([0, 1], [0, 1], 'k--')\n",
"plt.plot(fpr_rf, tpr_rf, label='RF')\n",
"plt.xlabel('False positive rate')\n",
"plt.ylabel('True positive rate')\n",
"plt.title('ROC curve')\n",
"plt.legend(loc='best')\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python [conda env:py37]",
"language": "python",
"name": "conda-env-py37-py"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.2"
},
"toc": {
"base_numbering": 1,
"nav_menu": {
"height": "490px",
"width": "355px"
},
"number_sections": true,
"sideBar": true,
"skip_h1_title": false,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": true,
"toc_position": {
"height": "calc(100% - 180px)",
"left": "10px",
"top": "150px",
"width": "346px"
},
"toc_section_display": true,
"toc_window_display": true
},
"varInspector": {
"cols": {
"lenName": 16,
"lenType": 16,
"lenVar": 40
},
"kernels_config": {
"python": {
"delete_cmd_postfix": "",
"delete_cmd_prefix": "del ",
"library": "var_list.py",
"varRefreshCmd": "print(var_dic_list())"
},
"r": {
"delete_cmd_postfix": ") ",
"delete_cmd_prefix": "rm(",
"library": "var_list.r",
"varRefreshCmd": "cat(var_dic_list()) "
}
},
"types_to_exclude": [
"module",
"function",
"builtin_function_or_method",
"instance",
"_Feature"
],
"window_display": false
}
},
"nbformat": 4,
"nbformat_minor": 2
}