-
Notifications
You must be signed in to change notification settings - Fork 0
/
generate_data.py
130 lines (91 loc) · 4.57 KB
/
generate_data.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
import os
import numpy
from sklearn.model_selection import KFold
from mlutils import datasets
from mlutils.benchmarks import Chrono
numpy.random.seed(1337)
def getSamples(data, w=None, pct=10):
if w is None:
sampleIndexes = numpy.random.choice(data.shape[0], size=int(data.shape[0] * (pct / 100.0)), replace=False, )
else:
sampleIndexes = numpy.random.choice(data.shape[0], size=int(data.shape[0] * (pct / 100.0)), p=w, replace=False)
sampleIndexes = numpy.sort(sampleIndexes)
sampleWeights = None
if w is not None:
sampleWeights = w[sampleIndexes]
sampleWeights = sampleWeights / numpy.sum(sampleWeights)
return (sampleIndexes, sampleWeights)
def coreset(dataIn, pct=10):
def getWeights(data):
q, _ = numpy.linalg.qr(data)
# get sample weights
qnorm = numpy.linalg.norm(q, 2, 1)
w = qnorm / numpy.sum(qnorm)
return w
(sampleIndexes, w) = getSamples(dataIn, getWeights(dataIn), pct)
return sampleIndexes, w
def mkfolders(fname):
os.makedirs(fname, exist_ok=True)
return fname
def prepareExp(folder, fold, train, test, weights, family, base_time, cmds, fulltrain):
prefix = folder + "fold_" + str(fold) + "_"
trainfname = prefix + "train.npz"
weightsfname = prefix + "weights.npz"
testfname = prefix + "test.npz"
coefffname = prefix + "coeff.mmp"
predfname = prefix + "pred.mmp"
timefname = prefix + "time.mmp"
trainllfname = prefix + "trainll.mmp"
testllfname = prefix + "testll.mmp"
fulltrainllfname = prefix + "fulltrainll.mmp"
fulltrainfname = prefix + "full_train.npz"
numpy.savez_compressed(trainfname, train)
numpy.savez_compressed(fulltrainfname, fulltrain)
numpy.savez_compressed(testfname, test)
weightscmd = ""
if weights is not None:
numpy.savez_compressed(weightsfname, weights)
weightscmd = "--weights '%s'" % (weightsfname)
numpy.memmap(coefffname, dtype='float32', mode='w+', shape=(train.shape[1], train.shape[1] + 1)).flush()
numpy.memmap(predfname, dtype='float32', mode='w+', shape=(test.shape[0], test.shape[1])).flush()
numpy.memmap(trainllfname, dtype='float32', mode='w+', shape=(train.shape[0], train.shape[1])).flush()
numpy.memmap(testllfname, dtype='float32', mode='w+', shape=(test.shape[0], test.shape[1])).flush()
numpy.memmap(fulltrainllfname, dtype='float32', mode='w+', shape=(fulltrain.shape[0], fulltrain.shape[1])).flush()
timetable = numpy.memmap(timefname, dtype='float32', mode='w+', shape=(train.shape[1], 1))
timetable += base_time
timetable.flush()
for feature_idx in range(train.shape[1]):
cmds.append(
"python3 learnDN.py --train '%s' --full_train '%s' --test '%s' %s --family '%s' --feature '%s' --out_coeff '%s' --out_preds '%s' --out_time '%s' --out_train_loglikelihood '%s' --out_test_loglikelihood '%s' --out_full_train_loglikelihood '%s'" % (
trainfname, fulltrainfname, testfname, weightscmd, family, feature_idx, coefffname, predfname, timefname, trainllfname, testllfname, fulltrainllfname))
def generateInput(dsname, data, family):
data = data[numpy.var(data, 1) > 0, :]
fullcmd = []
cscmd = []
rndcmd = []
for fold, (train_index, test_index) in enumerate(KFold(n_splits=10, random_state=1).split(data)):
train, test = data[train_index, :], data[test_index, :]
prepareExp(mkfolders("data/" + dsname + "/full/"), fold, train, test, None, family, 0.0, fullcmd, train)
for pct in [5, 10, 20, 30, 40]:
cschrono = Chrono().start()
sampleIndexes, csweight = coreset(train, pct)
cstrain = train[sampleIndexes, :]
cschrono.end()
prepareExp(mkfolders("data/" + dsname + "/cs/" + str(pct) + "/"), fold, cstrain, test, csweight, family, cschrono.elapsed(), cscmd, train)
rndchrono = Chrono().start()
sampleIndexes, _ = getSamples(train, None, pct)
rtrain = train[sampleIndexes, :]
rndchrono.end()
prepareExp(mkfolders("data/" + dsname + "/rnd/" + str(pct) + "/"), fold, rtrain, test, None, family, rndchrono.elapsed(), rndcmd, train)
allcmds = []
allcmds.extend(fullcmd)
allcmds.extend(cscmd)
allcmds.extend(rndcmd)
with open("cmds_"+dsname+".txt", "w") as text_file:
text_file.write("\n".join(allcmds))
(dsname, data, features, target) = datasets.getMNIST()
print(dsname)
generateInput(dsname, data, "gaussian")
#(dsname, data, features, times, hours) = datasets.getTrafficLarge()
#print(dsname)
#generateInput(dsname, data, "poisson")