"//encodingFile.h5",
mode="r")
featureRowMatrix = fileh.root.featureRowMatrix
sizeOfSeqList = featureRowMatrix.shape[1]
labelDict = algFile.readLabelDictFromFile()
labelVec = np.array(classifierFile.convertLabelDict2List(labelDict))
numOfLabelData = len(labelVec)
kFeatures = range(1, 60)
accuracyScoreList = []
accuracyScoreList2 = []
sumarizeAccuracy2 = []
loo = LeaveOneOut()
#variance
seqRowMatrixFull = algFile.readSeqRowMatFromFile()
sel = VarianceThreshold()
sel.fit_transform(seqRowMatrixFull)
sortedIndexVarList = np.flipud(np.argsort(sel.variances_))
newFeatureRowMatrix = np.transpose(seqRowMatrixFull)[sortedIndexVarList]
newSeqRowMatrix = np.transpose(newFeatureRowMatrix)
newSeqRowOfLabelData = newSeqRowMatrix[:numOfLabelData]
#leave one out #variance
for numOfFeature in kFeatures:
seqRowOfLabelData = newSeqRowOfLabelData[:, :numOfFeature]
tmpAccuracyScoreList2 = []
for train_index, test_index in loo.split(seqRowOfLabelData):
X_train, X_test = seqRowOfLabelData[
train_index], seqRowOfLabelData[test_index]
y_train, y_test = labelVec[train_index], labelVec[test_index]
labelVec = [None] * numOfLabelData
for label, listIndexSeq in labelDict.iteritems():
for index in listIndexSeq:
labelVec[index] = label
return labelVec
if __name__ == '__main__':
labelDict = algFile.readLabelDictFromFile()
# tmp = convertLabelDict2List(labelDict)
# tmp[1].remove(None)
# labelVec = tmp[1]
labelVec = convertLabelDict2List(labelDict)
labelVec.remove(None)
numOfLabelData = len(labelVec)
seqRowMatrix = algFile.readSeqRowMatFromFile(
'../outputfile/afterFeatureSelection.h5')
seqRowOfLabelData = seqRowMatrix[:numOfLabelData]
neigh = KNeighborsClassifier(n_neighbors=5, metric=cosine_distances)
tmp2 = neigh.fit(seqRowOfLabelData, labelVec)
labelPredict = neigh.predict(seqRowOfLabelData)
clf = svm.SVC()
clf.fit(seqRowOfLabelData, labelVec)
# seqRowMatrix = algFile.readSeqRowMatFromFile()
# clf = svm.SVC()
# clf.fit(seqRowMatrix[:numOfLabelData], labelVec)
# labelReal = labelVec[77]