mescla.uncertainty.identifiability.endmember_separation

mescla.uncertainty.identifiability.endmember_separation(endmembers, standardize=True)

Pairwise distance between end-members, normalised by their overall spread.

End-members that sit close together cannot be told apart by any amount of data. The relative column is the pair distance divided by the largest pair distance: values below about 0.2 mean that pair is effectively one source.

Parameters:

• endmembers (Any)

• standardize (bool)

Return type:

pd.DataFrame