import pandas as pd import numpy as np import pandas.io.data as web import datetime import math import sys from app.processor.stockprofiling.entity.StockProfilingDataBean import * from app.processor.stockprofiling.entity.StockReturnProfiles import * from app.processor.stockprofiling.entity.HistoricalVolatilities import * from app.processor.stockprofiling.entity.StandardDeviationBuckets import * # for debug from flask import json from app.resource.CustomerJSONEncoder import CustomerJSONEncoder class StockProfilingProcessor: @staticmethod def calculate(ticker): #ticker = 'SPY' endDate = datetime.date.today().strftime("%m/%d/%Y") if datetime.date.today().month == 2 and datetime.date.today().day == 29: startDate = datetime.date.today().replace(month = 2, day = 28, year=datetime.date.today().year - 2).strftime("%m/%d/%Y") else: startDate = datetime.date.today().replace(year=datetime.date.today().year - 2).strftime("%m/%d/%Y") try: stock = web.DataReader(ticker, 'yahoo', start = startDate, end =endDate) except Exception as e: print 'Error occur while reading stock data from Yahoo: ' + str(e) raise stock['Return'] = np.log(stock['Adj Close'] / stock['Adj Close'].shift(1)) Volume30 = stock['Volume'][:-31:-1].mean() stock['MA50'] = pd.rolling_mean(stock['Adj Close'], 50) stock['MA200'] = pd.rolling_mean(stock['Adj Close'], 200) Up = stock['Return'][stock['Return'] > 0].count() Down = stock['Return'][stock['Return'] < 0].count() UpDown = Up/Down + 0.0 MIN5 = stock['Adj Close'][:-6:-1].min() MIN30 = stock['Adj Close'][:-31:-1].min() MIN90 = stock['Adj Close'][:-91:-1].min() MIN252 = stock['Adj Close'][:-253:-1].min() MIN1YR = stock['Adj Close'][:-253:-1].min() MIN3YR = stock['Adj Close'].min() MAX5 = stock['Adj Close'][:-6:-1].max() MAX30 = stock['Adj Close'][:-31:-1].max() MAX90 = stock['Adj Close'][:-91:-1].max() MAX1YR = stock['Adj Close'][:-252:-1].max() MAX3YR = stock['Adj Close'].max() Volume5 = stock['Volume'][:-6:-1].mean() Volume90 = stock['Volume'][:-91:-1].mean() Volume252 = stock['Volume'][:-253:-1].mean() VolumePeriod = stock['Volume'].mean() Ret5 = math.log(stock['Adj Close'][-1] / stock['Adj Close'][-6]) Ret30 = math.log(stock['Adj Close'][-1] / stock['Adj Close'][-31]) Ret90 = math.log(stock['Adj Close'][-1] / stock['Adj Close'][-91]) Ret252 = math.log(stock['Adj Close'][-1] / stock['Adj Close'][-253]) RetPeriod = math.log(stock['Adj Close'][-1] / stock['Adj Close'][0]) stock['Gap Return'] = np.log(stock['Open'] / stock['Close'].shift(1)) Gap5 = stock['Gap Return'][:-6:-1].mean() Gap30 = stock['Gap Return'][:-31:-1].mean() Gap90 = stock['Gap Return'][:-91:-1].mean() Gap252 = stock['Gap Return'][:-253:-1].mean() GapPeriod = stock['Gap Return'].mean() Ret5MIN = stock['Return'][:-6:-1].min() Ret30MIN = stock['Return'][:-31:-1].min() Ret90MIN = stock['Return'][:-91:-1].min() Ret252MIN = stock['Return'][:-252:-1].min() RetPeriodMIN = stock['Return'].min() Ret5MAX = stock['Return'][:-6:-1].max() Ret30MAX = stock['Return'][:-31:-1].max() Ret90MAX = stock['Return'][:-91:-1].max() Ret252MAX = stock['Return'][:-253:-1].max() RetPeriodMAX = stock['Return'].max() AutoCorr5 = stock['Adj Close'][:-6:-1].autocorr() AutoCorr30 = stock['Adj Close'][:-31:-1].autocorr() AutoCorr90 = stock['Adj Close'][:-91:-1].autocorr() AutoCorr252 = stock['Adj Close'][:-253:-1].autocorr() AutoCorrPeriod = stock['Adj Close'].autocorr() stock['HV10'] = pd.rolling_std(stock['Return'],10) * np.sqrt(252) stock['HV20'] = pd.rolling_std(stock['Return'],20) * np.sqrt(252) stock['HV30'] = pd.rolling_std(stock['Return'],30) * np.sqrt(252) stock['HV60'] = pd.rolling_std(stock['Return'],60) * np.sqrt(252) stock['HV90'] = pd.rolling_std(stock['Return'],90) * np.sqrt(252) stock['xSTD10'] = stock['Return'] / (stock['HV10'].shift(1) / np.sqrt(252)) stock['xSTD20'] = stock['Return'] / (stock['HV20'].shift(1) / np.sqrt(252)) stock['xSTD30'] = stock['Return'] / (stock['HV30'].shift(1) / np.sqrt(252)) stock['xSTD60'] = stock['Return'] / (stock['HV60'].shift(1) / np.sqrt(252)) stock['xSTD90'] = stock['Return'] / (stock['HV90'].shift(1) / np.sqrt(252)) stockProfiling = StockProfilingDataBean() stockProfiling.lastPrice = stock['Adj Close'][-1] stockProfiling.change = stock['Return'][-1] stockProfiling.dayAverageVolume30 = Volume30 stockProfiling.dayAverageVolume50 = stock['MA50'][-1] stockProfiling.dayMovingAverage200 = stock['MA200'][-1] stockProfiling.upDownRatio = UpDown stockProfiling.stockReturnProfiles = [ StockProfilingProcessor.buildStockReturnProfiles("1 Week: ", MIN5, MAX5, Volume5, Ret5, Gap5, Ret5MIN, Ret5MAX, AutoCorr5), StockProfilingProcessor.buildStockReturnProfiles("30 Days: ", MIN30, MAX30, Volume30, Ret30, Gap30, Ret30MIN, Ret30MAX, AutoCorr30), StockProfilingProcessor.buildStockReturnProfiles("90 Days: ", MIN90, MAX90, Volume90, Ret90, Gap90, Ret90MIN, Ret90MAX, AutoCorr90), #StockProfilingProcessor.buildStockReturnProfiles("YTD: ",) StockProfilingProcessor.buildStockReturnProfiles("1 Year: ", MIN1YR, MAX1YR, Volume252, Ret252, Gap252, Ret252MIN, Ret252MAX, AutoCorr252), StockProfilingProcessor.buildStockReturnProfiles("3 Year: ", MIN3YR, MAX3YR, VolumePeriod,RetPeriod, GapPeriod, RetPeriodMIN, RetPeriodMAX, AutoCorrPeriod), ] stockProfiling.historicalVolatilities = [ StockProfilingProcessor.buildHistoricalVolatilities("Previous Close: ", stock['HV10'][-1], stock['HV20'][-1], stock['HV30'][-1], stock['HV60'][-1], stock['HV90'][-1]), StockProfilingProcessor.buildHistoricalVolatilities("1 Week Ago: ", stock['HV10'][-6], stock['HV20'][-6], stock['HV30'][-6], stock['HV60'][-6], stock['HV90'][-6]), StockProfilingProcessor.buildHistoricalVolatilities("1 Year High: ", stock['HV10'][:-253:-1].max(), stock['HV20'][:-253:-1].max(), stock['HV30'][:-253:-1].max(), stock['HV60'][:-253:-1].max(), stock['HV90'][:-253:-1].max()), StockProfilingProcessor.buildHistoricalVolatilities("1 Year Low: ", stock['HV10'][:-253:-1].min(), stock['HV20'][:-253:-1].min(), stock['HV30'][:-253:-1].min(), stock['HV60'][:-253:-1].min(), stock['HV90'][:-253:-1].min()), ] stockProfiling.standardDeviationBuckets = [ StockProfilingProcessor.buildStandardDeviationBuckets("HV10:", stock['xSTD10'][:-253:-1][stock['xSTD10'].abs() > 1].count(), stock['xSTD10'][:-253:-1][stock['xSTD10'].abs() > 2].count(), stock['xSTD10'][:-253:-1][stock['xSTD10'].abs() > 3].count(), stock['xSTD10'][:-253:-1][stock['xSTD10'].abs() > 4].count()), StockProfilingProcessor.buildStandardDeviationBuckets("HV20:", stock['xSTD20'][:-253:-1][stock['xSTD20'].abs() > 1].count(), stock['xSTD20'][:-253:-1][stock['xSTD20'].abs() > 2].count(), stock['xSTD20'][:-253:-1][stock['xSTD20'].abs() > 3].count(), stock['xSTD20'][:-253:-1][stock['xSTD20'].abs() > 4].count()), StockProfilingProcessor.buildStandardDeviationBuckets("HV30:", stock['xSTD30'][:-253:-1][stock['xSTD30'].abs() > 1].count(), stock['xSTD30'][:-253:-1][stock['xSTD30'].abs() > 2].count(), stock['xSTD30'][:-253:-1][stock['xSTD30'].abs() > 3].count(), stock['xSTD30'][:-253:-1][stock['xSTD30'].abs() > 4].count()), StockProfilingProcessor.buildStandardDeviationBuckets("HV60:", stock['xSTD60'][:-253:-1][stock['xSTD60'].abs() > 1].count(), stock['xSTD60'][:-253:-1][stock['xSTD60'].abs() > 2].count(), stock['xSTD60'][:-253:-1][stock['xSTD60'].abs() > 3].count(), stock['xSTD60'][:-253:-1][stock['xSTD60'].abs() > 4].count()), StockProfilingProcessor.buildStandardDeviationBuckets("HV90:", stock['xSTD90'][:-253:-1][stock['xSTD90'].abs() > 1].count(), stock['xSTD90'][:-253:-1][stock['xSTD90'].abs() > 2].count(), stock['xSTD90'][:-253:-1][stock['xSTD90'].abs() > 3].count(), stock['xSTD90'][:-253:-1][stock['xSTD90'].abs() > 4].count()), ] volChartData = stock.loc[:, ['HV10', 'HV20', 'HV30', 'HV60', 'HV90']] volChartData.loc[:,'Market Date'] = volChartData.index volChartData = volChartData[91:] chartColumns = ['Market Date', 'HV10', 'HV20', 'HV30', 'HV60', 'HV90'] chartValueDict = {'chartTitles':chartColumns, 'chartValues': volChartData, 'format':'date'} stockProfiling.volChart = ChartDataBean(**chartValueDict) stockPriceChartData = stock.loc[:, ['Adj Close']] stockPriceChartData.loc[:,'Market Date'] = stockPriceChartData.index stockPriceChartData = stockPriceChartData[91:] chartColumns = ['Market Date', 'Adj Close'] chartValueDict = {'chartTitles':chartColumns, 'chartValues': stockPriceChartData, 'format':'date'} stockProfiling.stockPriceChart = ChartDataBean(**chartValueDict) xSTDChartData = stock.loc[:, ['xSTD10', 'xSTD20', 'xSTD30', 'xSTD60', 'xSTD90']] xSTDChartData.loc[:,'Market Date'] = xSTDChartData.index xSTDChartData = xSTDChartData[91:] chartColumns = ['Market Date', 'xSTD10', 'xSTD20', 'xSTD30', 'xSTD60', 'xSTD90'] chartValueDict = {'chartTitles':chartColumns, 'chartValues': xSTDChartData, 'format':'date'} stockProfiling.xSTDChart = ChartDataBean(**chartValueDict) count, division = np.histogram(stock['Return'], bins = np.arange(round(stock['Return'].min(),2),round(stock['Return'].max(),2), round((stock['Return'].max() - stock['Return'].min())/16,3))) histogramDict = {'hist':[x for x in count], 'bin_edges':[x for x in division]} stockProfiling.histogramChart = histogramDict; return stockProfiling @staticmethod def buildStockReturnProfiles(period='', priceMin=0., priceMax=0., volume=0., periodReturn=0., gapReturnMean=0., returnMin=0., returnMax=0., autoCorrelation=0.): dict = { "period": period, "priceMin": priceMin, "priceMax": priceMax, "volume": volume, "periodReturn": periodReturn, "gapReturnMean": gapReturnMean, "returnMin": returnMin, "returnMax": returnMax, "autoCorrelation": autoCorrelation, } return StockReturnProfiles(**dict) @staticmethod def buildHistoricalVolatilities(period='', HV10=0., HV20=0., HV30=0., HV60=0., HV90=0.): dict = { "period": period, "HV10": HV10, "HV20": HV20, "HV30": HV30, "HV60": HV60, "HV90": HV90, } return HistoricalVolatilities(**dict) @staticmethod def buildStandardDeviationBuckets(period='', returnOver1STD=0., returnOver2STD=0., returnOver3STD=0., returnOver4STD=0.): dict = { "period": period, "returnOver1STD": returnOver1STD, "returnOver2STD": returnOver2STD, "returnOver3STD": returnOver3STD, "returnOver4STD": returnOver4STD, } return StandardDeviationBuckets(**dict)