この手法は機械学習ではなく、情報検索に近いといえる。
履歴を使っているので一般的な意味での学習とはいえるが、
全く処理を施していないので機械学習とは異なる。
（精度が上がる保証がある手法のことを機械学習と呼ぶのか）
import matplotlib import matplotlib.pyplot as plt import matplotlib.ticker as mticker import matplotlib.dates as mdates import numpy as np import time #np.seterr(divide='ignore', invalid='ignore') totalStart = time.time() date,bid,ask = np.loadtxt('GBPUSD1d.txt',unpack=True,delimiter=',',converters={0:mdates.strpdate2num('%Y%m%d%H%M%S')}) def percentChange(startPoint,currentPoint): try: x = ((float(currentPoint)-startPoint)/abs(startPoint))*100.00 if x == 0.0: return 0.0000000001 else: return x except: return 0.000000001 def patternStorage(): patStartTime = time.time() x = len(avgLine)-60 y = 31 while y < x: pattern = [] p1 = percentChange(avgLine[y-30],avgLine[y-29]) p2 = percentChange(avgLine[y-30],avgLine[y-28]) p3 = percentChange(avgLine[y-30],avgLine[y-27]) p4 = percentChange(avgLine[y-30],avgLine[y-26]) p5 = percentChange(avgLine[y-30],avgLine[y-25]) p6 = percentChange(avgLine[y-30],avgLine[y-24]) p7 = percentChange(avgLine[y-30],avgLine[y-23]) p8 = percentChange(avgLine[y-30],avgLine[y-22]) p9 = percentChange(avgLine[y-30],avgLine[y-21]) p10 = percentChange(avgLine[y-30],avgLine[y-20]) p11 = percentChange(avgLine[y-30],avgLine[y-19]) p12 = percentChange(avgLine[y-30],avgLine[y-18]) p13 = percentChange(avgLine[y-30],avgLine[y-17]) p14 = percentChange(avgLine[y-30],avgLine[y-16]) p15 = percentChange(avgLine[y-30],avgLine[y-15]) p16 = percentChange(avgLine[y-30],avgLine[y-14]) p17 = percentChange(avgLine[y-30],avgLine[y-13]) p18 = percentChange(avgLine[y-30],avgLine[y-12]) p19 = percentChange(avgLine[y-30],avgLine[y-11]) p20 = percentChange(avgLine[y-30],avgLine[y-10]) p21 = percentChange(avgLine[y-30],avgLine[y-9]) p22 = percentChange(avgLine[y-30],avgLine[y-8]) p23 = percentChange(avgLine[y-30],avgLine[y-7]) p24 = percentChange(avgLine[y-30],avgLine[y-6]) p25 = percentChange(avgLine[y-30],avgLine[y-5]) p26 = percentChange(avgLine[y-30],avgLine[y-4]) p27 = percentChange(avgLine[y-30],avgLine[y-3]) p28 = percentChange(avgLine[y-30],avgLine[y-2]) p29 = percentChange(avgLine[y-30],avgLine[y-1]) p30 = percentChange(avgLine[y-30],avgLine[y]) outcomeRange = avgLine[y+20:y+30] currentPoint = avgLine[y] try: avgOutcome = reduce(lambda x, y: x+y, outcomeRange) / len(outcomeRange) except Exception, e: print str(e) avgOutcome=0 futureOutcome = percentChange(currentPoint,avgOutcome) pattern.append(p1) pattern.append(p2) pattern.append(p3) pattern.append(p4) pattern.append(p5) pattern.append(p6) pattern.append(p7) pattern.append(p8) pattern.append(p9) pattern.append(p10) pattern.append(p11) pattern.append(p12) pattern.append(p13) pattern.append(p14) pattern.append(p15) pattern.append(p16) pattern.append(p17) pattern.append(p18) pattern.append(p19) pattern.append(p20) pattern.append(p21) pattern.append(p22) pattern.append(p23) pattern.append(p24) pattern.append(p25) pattern.append(p26) pattern.append(p27) pattern.append(p28) pattern.append(p29) pattern.append(p30) patternAr.append(pattern) performanceAr.append(futureOutcome) y += 1 patEndTime = time.time() print len(patternAr) print len(performanceAr) print "Pattern storage took:",patEndTime-patStartTime,'seconds' def currentPattern(): cp1 = percentChange(avgLine[-31],avgLine[-30]) cp2 = percentChange(avgLine[-31],avgLine[-29]) cp3 = percentChange(avgLine[-31],avgLine[-28]) cp4 = percentChange(avgLine[-31],avgLine[-27]) cp5 = percentChange(avgLine[-31],avgLine[-26]) cp6 = percentChange(avgLine[-31],avgLine[-25]) cp7 = percentChange(avgLine[-31],avgLine[-24]) cp8 = percentChange(avgLine[-31],avgLine[-23]) cp9 = percentChange(avgLine[-31],avgLine[-22]) cp10 = percentChange(avgLine[-31],avgLine[-21]) cp11 = percentChange(avgLine[-31],avgLine[-20]) cp12 = percentChange(avgLine[-31],avgLine[-19]) cp13 = percentChange(avgLine[-31],avgLine[-18]) cp14 = percentChange(avgLine[-31],avgLine[-17]) cp15 = percentChange(avgLine[-31],avgLine[-16]) cp16 = percentChange(avgLine[-31],avgLine[-15]) cp17 = percentChange(avgLine[-31],avgLine[-14]) cp18 = percentChange(avgLine[-31],avgLine[-13]) cp19 = percentChange(avgLine[-31],avgLine[-12]) cp20 = percentChange(avgLine[-31],avgLine[-11]) cp21 = percentChange(avgLine[-31],avgLine[-10]) cp22 = percentChange(avgLine[-31],avgLine[-9]) cp23 = percentChange(avgLine[-31],avgLine[-8]) cp24 = percentChange(avgLine[-31],avgLine[-7]) cp25 = percentChange(avgLine[-31],avgLine[-6]) cp26 = percentChange(avgLine[-31],avgLine[-5]) cp27 = percentChange(avgLine[-31],avgLine[-4]) cp28 = percentChange(avgLine[-31],avgLine[-3]) cp29 = percentChange(avgLine[-31],avgLine[-2]) cp30 = percentChange(avgLine[-31],avgLine[-1]) patForRec.append(cp1) patForRec.append(cp2) patForRec.append(cp3) patForRec.append(cp4) patForRec.append(cp5) patForRec.append(cp6) patForRec.append(cp7) patForRec.append(cp8) patForRec.append(cp9) patForRec.append(cp10) patForRec.append(cp11) patForRec.append(cp12) patForRec.append(cp13) patForRec.append(cp14) patForRec.append(cp15) patForRec.append(cp16) patForRec.append(cp17) patForRec.append(cp18) patForRec.append(cp19) patForRec.append(cp20) patForRec.append(cp21) patForRec.append(cp22) patForRec.append(cp23) patForRec.append(cp24) patForRec.append(cp25) patForRec.append(cp26) patForRec.append(cp27) patForRec.append(cp28) patForRec.append(cp29) patForRec.append(cp30) print patForRec def patternRecognition(): predictedOutcomeAr = [] plotPatAr = [] patFound = 0 for eachPattern in patternAr: sim1 = 100.00 - abs(percentChange(eachPattern[0],patForRec[0])) sim2 = 100.00 - abs(percentChange(eachPattern[1],patForRec[1])) sim3 = 100.00 - abs(percentChange(eachPattern[2],patForRec[2])) sim4 = 100.00 - abs(percentChange(eachPattern[3],patForRec[3])) sim5 = 100.00 - abs(percentChange(eachPattern[4],patForRec[4])) sim6 = 100.00 - abs(percentChange(eachPattern[5],patForRec[5])) sim7 = 100.00 - abs(percentChange(eachPattern[6],patForRec[6])) sim8 = 100.00 - abs(percentChange(eachPattern[7],patForRec[7])) sim9 = 100.00 - abs(percentChange(eachPattern[8],patForRec[8])) sim10 = 100.00 - abs(percentChange(eachPattern[9],patForRec[9])) sim11 = 100.00 - abs(percentChange(eachPattern[10],patForRec[10])) sim12 = 100.00 - abs(percentChange(eachPattern[11],patForRec[11])) sim13 = 100.00 - abs(percentChange(eachPattern[12],patForRec[12])) sim14 = 100.00 - abs(percentChange(eachPattern[13],patForRec[13])) sim15 = 100.00 - abs(percentChange(eachPattern[14],patForRec[14])) sim16 = 100.00 - abs(percentChange(eachPattern[15],patForRec[15])) sim17 = 100.00 - abs(percentChange(eachPattern[16],patForRec[16])) sim18 = 100.00 - abs(percentChange(eachPattern[17],patForRec[17])) sim19 = 100.00 - abs(percentChange(eachPattern[18],patForRec[18])) sim20 = 100.00 - abs(percentChange(eachPattern[19],patForRec[19])) sim21 = 100.00 - abs(percentChange(eachPattern[20],patForRec[20])) sim22 = 100.00 - abs(percentChange(eachPattern[21],patForRec[21])) sim23 = 100.00 - abs(percentChange(eachPattern[22],patForRec[22])) sim24 = 100.00 - abs(percentChange(eachPattern[23],patForRec[23])) sim25 = 100.00 - abs(percentChange(eachPattern[24],patForRec[24])) sim26 = 100.00 - abs(percentChange(eachPattern[25],patForRec[25])) sim27 = 100.00 - abs(percentChange(eachPattern[26],patForRec[26])) sim28 = 100.00 - abs(percentChange(eachPattern[27],patForRec[27])) sim29 = 100.00 - abs(percentChange(eachPattern[28],patForRec[28])) sim30 = 100.00 - abs(percentChange(eachPattern[29],patForRec[29])) howSim = (sim1+sim2+sim3+sim4+sim5+sim6+sim7+sim8+sim9+sim10+ sim11+sim12+sim13+sim14+sim15+sim16+sim17+sim18+sim19+sim20+ sim21+sim22+sim23+sim24+sim25+sim26+sim27+sim28+sim29+sim30 )/30.00 if howSim > 70: patdex = patternAr.index(eachPattern) patFound = 1 ''' print '#############################' print '#############################' print patForRec print '=============================' print '=============================' print eachPattern print '-----------------------------' print 'predicted outcome', performanceAr[patdex] print '#############################' print '#############################' ''' xp = [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30] plotPatAr.append(eachPattern) predArray = [] if patFound == 1: #fig = plt.figure(figsize=(10,6)) for eachPatt in plotPatAr: futurePoints = patternAr.index(eachPatt) if performanceAr[futurePoints] > patForRec[29]: pcolor = '#24bc00' predArray.append(1.000) else: pcolor = '#d40000' predArray.append(-1.000) #plt.plot(xp,eachPatt) predictedOutcomeAr.append(performanceAr[futurePoints]) #plt.scatter(35,performanceAr[futurePoints],c=pcolor,alpha=.3) realOutcomeRange = allData[toWhat+20:toWhat+30] realAvgOutcome = reduce(lambda x, y: x+y, realOutcomeRange) / len(realOutcomeRange) realMovement = percentChange(allData[toWhat],realAvgOutcome) predictedAvgOutcome = reduce(lambda x, y: x+y, predictedOutcomeAr) / len(predictedOutcomeAr) print predArray predictionAverage = reduce(lambda x,y: x+y, predArray)/len(predArray) print predictionAverage if predictionAverage < 0: print 'drop predicted' print patForRec[29] print realMovement if realMovement < patForRec[29]: accuracyArray.append(100) else: accuracyArray.append(0) if predictionAverage > 0: print 'rise predicted' print patForRec[29] print realMovement if realMovement > patForRec[29]: accuracyArray.append(100) else: accuracyArray.append(0) #plt.scatter(40,realMovement,c='#54fff7',s=25) #plt.scatter(40,predictedAvgOutcome,c='b',s=25) #plt.plot(xp,patForRec,'#54fff7',linewidth=3) #plt.grid(True) #plt.title('Pattern Recognition') #plt.show() def graphRawFX(): fig = plt.figure(figsize=(10,7)) ax1 = plt.subplot2grid((40,40),(0,0),rowspan=40,colspan=40) ax1.plot(date,bid) ax1.plot(date,ask) plt.gca().get_yaxis().get_major_formatter().set_useOffset(False) ax1.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d %H-%M-%S')) for label in ax1.xaxis.get_ticklabels(): label.set_rotation(45) ax1_2 = ax1.twinx() ax1_2.fill_between(date,0,(ask-bid),facecolor='g',alpha=.3) plt.subplots_adjust(bottom=.23) plt.grid(True) plt.show() if __name__=="__main__": dataLength = int(bid.shape[0]) print 'data length is',dataLength toWhat = 37000 allData = ((bid+ask)/2) accuracyArray = [] samps = 0 while toWhat < dataLength: #avgLine = ((bid+ask)/2) avgLine = allData[:toWhat] patternAr = [] performanceAr = [] patForRec = [] patternStorage() currentPattern() patternRecognition() #graphRawFX() totalTime = time.time() - totalStart print 'Entire pricessing time took:',totalTime,'seconds' moveOn = raw_input('press ENTER to continue') samps += 1 toWhat += 1 accuracyAverage = reduce(lambda x,y: x+y, accuracyArray)/len(accuracyArray) print 'Backtested Acuracy is',str(accuracyAverage)+'% after',samps,'samples'
