def Main(edgeFile=None,outEdgeFile=None,nodeFile=None,nodeList=None,comp=None,subType=None,sep=None,log=None):
""" Main program """
i0 = time.clock()
inext = i0
### Argument processing ========================================
## Filename definitions ======================================
i0 = time.clock()
## Lecture des fichiers ========================================
if nodeFile:
subnodes = ut.file2list(nodeFile)
elif nodeList:
subnodes = nodeList.strip().split(",")
elif comp:
compFile,compID = comp.strip().split(",")
subnodes = ut.getNodes(compFile,compID)
else:
subnodes = None
## Corps du programme =========================================== The dictionary old_nodes --> new_nodes obtained as the completion of the clustering file
ut.inducedSubgraph(edgeFile,subnodes=subnodes,nodeType=subType,outFile=outEdgeFile,sep=sep)
## Sortie ======================================================
inext = myTimer(inext,handle=log)
## Ending ==============================================================
prog = myModule()
if prog == "__main__.py":
prog = sys.argv[0].split("/")[-1]
inext = myTimer(i0,"Total computing time for %s" % prog,handle=log)
return
def test_str2tmp():
txt = "This is a test string"
filename = utils.str2tmp(txt)
assert os.path.isfile(filename)
assert open(filename).read() == txt + "\n"
assert utils.file2str(filename) == txt
# TODO: should this really return
# an empty line at the end?
assert utils.file2list(filename) == [txt, '']
os.unlink(filename)
def grammarLoader():
gram_in = [
x.rstrip() for x in utils.file2list('grammar.txt')
if not re.match('\\n|#', x)
]
terS.extend(gram_in[0].split(','))
i_nTerS.extend(gram_in[1].split(','))
relations.update({
k.rstrip(): (v.strip()).split('|')
for k, v in map(lambda x: x.split('->'), gram_in[2:])
})
o_relations.update({f'{x}_f': set(x) for x in terS})
def random_sample(segment_per_class, sample_range):
""" Returns the list of the random sample
(new_indexes, time_series, label)
Argument
---------
segment_per_class
The list of time_series
The list of the corresponding indexes
"""
dataset = {}
poses = []
if load_history_matrix:
poses = file2list(POS_FILE)
if poses:
for k in segment_per_class.keys():
dataset[k] = ([], [])
for item in segment_per_class[k]:
if item[1] in poses:
dataset[k][0].append(item[0])
dataset[k][1].append(item[1])
else:
for k in segment_per_class.keys():
array = list2array(segment_per_class[k])
if sample_range > 0:
if len(segment_per_class[k]) > sample_range:
list_a = random.sample(segment_per_class[k], sample_range)
array = list2array(list_a)
dataset[k] = (array[:, 0].tolist(), array[:, 1].tolist())
dataset = intercept_part_dataSet(dataset)
ts = []
for label in dataset:
(list_time_series, list_indexes) = dataset[label]
for time_series, indexes in zip(list_time_series, list_indexes):
# time_series = preprocessing(time_series)
time_series.insert(0, data_index[label])
ts.append(time_series)
return ts
def model_test(model_path,
clf=None,
filename='out.txt',
tcName=['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']):
path = sys.path[0]
tbasePath = os.path.join(path, "mnist/test/")
#tcName = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']
tst = time.time()
f = open(filename, "w")
feature_cnt = len(tcName)
allErrCount = 0
allErrorRate = 0.0
allScore = 0.0
ErrCount = np.zeros(feature_cnt, int)
TrueCount = np.zeros(feature_cnt, int)
predict_list = []
true_list = []
if clf == None:
clf = joblib.load(model_path)
for tcn in tcName:
testPath = tbasePath + tcn
tflist = utils.file2list(testPath)
tdataMat, tdataLabel = utils.read2convert(tflist)
print("test dataMat shape: {0}, test dataLabel len: {1} ".format(
tdataMat.shape, len(tdataLabel)),
file=f)
pre_st = time.time()
preResult = clf.predict(tdataMat)
pre_et = time.time()
predict_list.append(preResult)
true_list.append([tcn] * len(preResult))
print("Recognition " + tcn + " spent {:.4f}s.".format(
(pre_et - pre_st)),
file=f)
print("predict result: {}".format(len(preResult)))
errCount = len([x for x in preResult if x != tcn])
ErrCount[int(tcn)] = errCount
TrueCount[int(tcn)] = len(tdataLabel) - errCount
print("errorCount: {}.".format(errCount), file=f)
allErrCount += errCount
score_st = time.time()
score = clf.score(tdataMat, tdataLabel)
score_et = time.time()
print("computing score spent {:.6f}s.".format(score_et - score_st),
file=f)
allScore += score
print("score: {:.6f}.".format(score), file=f)
print("error rate is {:.6f}.".format((1 - score)), file=f)
tet = time.time()
print("Testing All class total spent {:.6f}s.".format(tet - tst), file=f)
print("All error Count is: {}.".format(allErrCount), file=f)
avgAccuracy = allScore / (feature_cnt * 1.0)
print("Average accuracy is: {:.6f}.".format(avgAccuracy), file=f)
print("Average error rate is: {:.6f}.".format(1 - avgAccuracy), file=f)
print("number", " TrueCount", " ErrCount", file=f)
for tcn in tcName:
tcn = int(tcn)
print(tcn, " ", TrueCount[tcn], " ", ErrCount[tcn], file=f)
plt.figure(figsize=(12, 6))
x = list(range(feature_cnt))
plt.plot(x, TrueCount, color='blue', label="TrueCount") # 将正确的数量设置为蓝色
plt.plot(x, ErrCount, color='red', label="ErrCount") # 将错误的数量为红色
plt.legend(loc='best') # 显示图例的位置，这里为右下方
plt.title('Projects')
plt.xlabel('number') # x轴标签
plt.ylabel('count') # y轴标签
plt.xticks(np.arange(feature_cnt), tcName)
plt.savefig('./accuracy_%s.jpg' % filename.split('_')[1])
plt.show()
f.close()
true_list = np.array(true_list).flatten()
predict_list = np.array(predict_list).flatten()
return [predict_list, true_list]