search

skinnider / dismay

Calculation of distance metrics for matrices
MIT License
25 stars 5 forks source link

Implementation of Zero Inflated Kendall's Tau #3

Open abdkarr opened 3 years ago

abdkarr commented 3 years ago

First of all, thanks for this great library and associated library where data and evaluation pipeline are shared. My question is related to zero-inflated Kendall's Tau. I was wondering if its implementation is correct. In calculation of p1, X10 is compared to X11 element-wise, i.e. X10[i] to X11[i] forall i = 1, .., length(X10). Shouldn't the comparison be done between all pairs of X10 and X11, that is X10[i] should be compared to X11[j] for all i, j pairs for which X10[i] is nonzero and X11[j] is nonzero? Similarly, for p2. I hope my question is clear.

Close-your-eyes commented 1 year ago

Could you make up your mind about this issue? (@abdkarr and @skinnider) ??