Package CHEM :: Package ML :: Package featureAnalysis :: Module SimiMeasureAnalysis :: Class SimiMeasureAnalysis

Class SimiMeasureAnalysis

SimiMeasureAnalysis is used to compute different similarity matrices i.e., corrcoef, tanimoto, and etc, on selected features. Dump and load the similarity matrix (for windows). Plot the distribution(histogram) of the similarity values in the simi matrix. Show the grayscale inshow of the simi matrix (on windows).

Instance Methods

[hide private]

__init__(self, bitOrCount, selectionType, selectedFeatureNum, featureDictListFile)

computingSimiMeasureMatrix(self, simiMeasureType)

computingCorrSimiMatrix(self)

computingTaniSimiMatrix(self)

Tanimoto(self, A, B)

Instance Variables

[hide private]

selectedProbArray
f = open('dump1000bit.pkl','r') featureMatrix = zeros((49999,topFeatureNum),float) featureMatrix[0:10000,:] = pickle.load(f) featureMatrix[10001:20000,:] = pickle.load(f) featureMatrix[20001:30000,:] = pickle.load(f) featureMatrix[30001:40000,:] = pickle.load(f) featureMatrix[40001:,:] = pickle.load(f) y = pickle.load(f) z = pickle.load(f) f.close()