Closed karchern closed 7 years ago
The call to the same function is correct. The difference is in the "S" matrix provided: TINA will take a SparCC-based S matrix, PINA will take one based on cophenetic distances. In the uploaded code this is indeed a bug. In the script used for the original paper, the correct S matrix was used.
Fixed by replacing S.sparcc
w/ S.phylo
in the code.
both call community.similarity.corr.par (which computes TINA). Needs fix.