xmipp3.protocols.protocol_consensus_classes module

class xmipp3.protocols.protocol_consensus_classes.XmippProtConsensusClasses(*args, **kwargs)[source]

Bases: ProtClassify3D

Compare several SetOfClasses. Return the consensus clustering based on a objective function that uses the similarity between clusters intersections and the entropy of the clustering formed.

METRICS = ['cosine', 'euclidean', 'cityblock', 'correlation']

findElbowsStep()[source]

intersectStep()[source]

mergeStep()[source]

referenceIntersectionStep()[source]