similarity_weights: Generating similarity weights for calculating_similarity
In imbs-hl/MDRDist: This Package calculates something like a distance

Description Usage Arguments Value Examples

This function can be used together with create_weighting_list to manually assemble similarity weights, which can be used in function calculate_similarity. Some usual weightings aur already collected in the function MDRDist_weights().

1	similarity_weights(numerator, denominator)

numerator

list of class "HLO_weighting" [which is output of create_weighting_list], that indicates, how similar classifications should be rewarded

denominator

list of class "HLO_weighting" [which is output of create_weighting_list], that indicates, which classification relations should be punished

@details Please note the following - For reasons of scaling, it is recommented, that for each element numerator < denominator [or at least numerator < denominator * C] for some constant. - For symmetry, it should not matter, whether the first or the second observation is L or H. Therefore you should set H_L = L_H. This applies to H_O = O_H and L_O = O_L as well. - Further it does not make sense, if a thing is to something more similar than to itself. So H_same and L_same should be maximal.

The function returns a list object with numerator, denuminator and class label "similarity_weights".