/
testalgos.py
232 lines (196 loc) · 8.41 KB
/
testalgos.py
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
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
import similarity
import display
import grouping
import tradingmeasure
import parameters as para
import util
import constants as const
#testOutputFilename = 'testresults.txt'
testOutputFilename = 'testresults_test.txt'
#testOutputFilename = 'results_confidenceSellOrKeep_fyh_nocond_byFirst.txt'
# test all algos
algosToTest = {
'acf': similarity.tsdist('acfDistance'),
'ar.lpc.ceps': similarity.tsdist('ar.lpc.cepsDistance'),
#'ar.mah': similarity.tsdist('ar.mahDistance'), #Need to retrieve p-value
'ar.pic': similarity.tsdist('ar.picDistance'),
'ccor': similarity.tsdist('ccorDistance'),
#'cdm': similarity.tsdist('cdmDistance'), # (USE) SLOW / INTERNAL ERROR 5 IN MEMCOMPRESS...?
'cid': similarity.tsdist('cidDistance'),
'cor': similarity.tsdist('corDistance'),
'cort': similarity.tsdist('cortDistance'),
'dissimapprox': similarity.tsdist('dissimapproxDistance'),
'dissim': similarity.tsdist('dissimDistance'),
'dtw': similarity.tsdist('dtwDistance'),
'edr_005': similarity.tsdist('edrDistance', 0.05),
'edr_01': similarity.tsdist('edrDistance', 0.1),
'edr_025': similarity.tsdist('edrDistance', 0.25),
'edr_05': similarity.tsdist('edrDistance', 0.5),
'erp_01': similarity.tsdist('erpDistance', 0.1),
'erp_05': similarity.tsdist('erpDistance', 0.5),
'erp_10': similarity.tsdist('erpDistance', 1),
'euclidean': similarity.tsdist('euclideanDistance'),
'fourier': similarity.tsdist('fourierDistance'),
#'frechet': similarity.tsdist('frechetDistance'), # (USE?) prints a lot of nonsense
'inf.norm': similarity.tsdist('inf.normDistance'),
'int.per': similarity.tsdist('int.perDistance'),
'lbKeogh_3': similarity.tsdist('lb.keoghDistance', 3),
'lcss_05': similarity.tsdist('lcssDistance', 0.05),
'lcss_15': similarity.tsdist('lcssDistance', 0.15),
'lcss_30': similarity.tsdist('lcssDistance', 0.3),
'lcss_50': similarity.tsdist('lcssDistance', 0.5),
'lp': similarity.tsdist('lpDistance'),
'manhattan': similarity.tsdist('manhattanDistance'),
'mindist.sax_1': similarity.tsdist('mindist.saxDistance',1),
'mindist.sax_2': similarity.tsdist('mindist.saxDistance',2),
'mindist.sax_4': similarity.tsdist('mindist.saxDistance',4),
'mindist.sax_8': similarity.tsdist('mindist.saxDistance',8),
'mindist.sax_16': similarity.tsdist('mindist.saxDistance',16),
'minkowski_25': similarity.lpNorms(2.5), #otherwise known as lp-norms
'minkowski_30': similarity.lpNorms(3),
'minkowski_05': similarity.lpNorms(0.5),
#'ncd': similarity.tsdist('ncdDistance'), # Unknown internal error
'pacf': similarity.tsdist('pacfDistance'),
'pdc': similarity.tsdist('pdcDistance'),
'per': similarity.tsdist('perDistance'),
#'pred': similarity.tsdist('predDistance'),
#'spec.glk': similarity.tsdist('spec.glkDistance'), # {USE} SLOW. Also, I'm getting strange L-BFGS-B errors.
#'spec.isd': similarity.tsdist('spec.isdDistance'), # {USE) SLOW. Also, I'm getting strange L-BFGS-B errors.
'spec.llr': similarity.tsdist('spec.llrDistance'),
'sts': similarity.tsdist('stsDistance'),
'tquest': similarity.tsdist('tquestDistance', tau=0.5), #seems to do nothing...?
'wav': similarity.tsdist('wavDistance'),
}
# test only a limited set
algosToTest = {
'sts': similarity.tsdist('stsDistance'),
'inf.norm': similarity.tsdist('inf.normDistance'),
'cort': similarity.tsdist('cortDistance'),
'lcss_05': similarity.tsdist('lcssDistance', 0.05),
'minkowski_25': similarity.lpNorms(2.5), #otherwise known as lp-norms
'minkowski_30': similarity.lpNorms(3),
'lbKeogh_3': similarity.tsdist('lb.keoghDistance', 3),
'dtw': similarity.tsdist('dtwDistance'),
'euclidean': similarity.tsdist('euclideanDistance'),
'fourier': similarity.tsdist('fourierDistance'),
'dissim': similarity.tsdist('dissimDistance'),
}
# test an even more limited set
algosToTest = {
'sts': similarity.tsdist('stsDistance'),
'mindist.sax_1': similarity.tsdist('mindist.saxDistance',1),
'dtw': similarity.tsdist('dtwDistance'),
}
# test only sts
algosToTest0 = {
'sts': similarity.tsdist('stsDistance'),
}
def testAlgo(algo, target):
# global data
# testAlgoOnData(algo, target, data['Close'])
#
#def testAlgoOnData(algo, target, dataList):
dates = data['Date']
groups = grouping.groupUp(data, data['Close'])
targetNext = target+const.ma
if targetNext >= len(groups):
return None
similarity._normalizeFuns = [similarity.byMean]
similarity._measureFun = algo
results = compareAllGroupsBefore(groups, target)
results2 = compareAllGroupsBefore(groups, targetNext)
results.reverse()
results.sort(key=lambda x : x[2])
results2.sort(key=lambda x : x[2])
### Uses Average Data: useOnlyAverageData = True
#tradePolicy = tradingmeasure.dontSell
#tradePolicy = tradingmeasure.sellOrKeep
#tradePolicy = tradingmeasure.riskAverseSellOrKeep
#tradePolicy = tradingmeasure.largestReturn
### Doesn't use Average Data: useOnlyAverageData = False
tradePolicy = tradingmeasure.confidenceFilter(0.2, tradingmeasure.sellOrKeep)
useOnlyAverageData = False
totalRank = 0
lpScore = 0
nResults = 10
for v in results[0:nResults]:
rank = getRank(results2, v[0]+const.ma)
totalRank += rank
lpScore += similarity.computeWith(groups[v[0]+const.ma], groups[targetNext], [similarity.byFirst], similarity.lpNorms(2))
dataLists = getDataLists(groups, results[0:nResults], const.ma)
if usingOnlyAverageData:
dataLists = tradingmeasure.averageData(dataLists)
money = tradingmeasure.computeWithFunOn(dataLists, groups[targetNext][2], tradePolicy)
#print(money)
totalRank *= 100 # normalize totalRank for equal weightage.
totalRank /= len(results2) # normalize totalRank for equal weightage.
return (lpScore/nResults, totalRank/nResults, money)
def compareAlgorithms(fileName):
global data, headers
data, headers = para.readFile(fileName)
if (len(data['Date']) == 0):
print('Empty File')
return
global algosToTest
for key in algosToTest.keys():
print('Testing ' + key)
algo = algosToTest[key]
result = testAlgo(algo, 404)
printResult(key, result)
def compareAlgorithmsWithData(testCases):
global algosToTest, data, headers, testOutputFilename
allResults = {}
for key in algosToTest.keys():
allResults[key] = []
for companyName in sorted(testCases.keys()):
print('Testing ' + companyName)
data, headers = para.readFile(display.nameToFile(companyName))
for key in sorted(algosToTest.keys()):
#print(' algo ' + key)
for target in testCases[companyName]:
algo = algosToTest[key]
result = testAlgo(algo, target)
if result != None:
allResults[key].append(result)
#printResult(key, result)
averageResults = {}
f = open(testOutputFilename, 'w+')
for key in allResults.keys():
averageResult = computeAverageResult(allResults[key])
s = formatResult(key, averageResult)
f.write(s+'\n')
f.close()
def formatResult(key, result):
s = map(str, [key] + list(result))
return '\t'.join(s)
def printResult(key, result):
print(formatResult(key, result))
def computeAverageResult(results):
import statistics
results = util.transposeLists(results)
print('Number of results: ' + str(len(results[0])))
return (statistics.mean(results[0]),
statistics.mean(results[1]),
statistics.mean(results[2]),
statistics.stdev(results[2]))
def getDataLists(groups, results, offset):
return list(map(lambda v : groups[v[0]+offset][2], results))
def getSimilarity(groups, i, j):
sim = similarity.compute(groups[i], groups[j])
return (i,j,sim)
def compareAllGroupsTo(groups, targetIndex):
results = []
print(dates[groups[targetIndex][0]])
print(dates[groups[targetIndex][1]])
for i in range(0,len(groups)):
if i != targetIndex:
results.append(getSimilarity(groups,i,targetIndex))
return results
def compareAllGroupsBefore(groups, targetIndex):
return list(map(lambda i : getSimilarity(groups, i, targetIndex), range(0,targetIndex)))
# assume results is sorted.
def getRank(results, index):
for i in range(0,len(results)):
if results[i][0] == index:
return i+1
return -1