I think there's a bug in the current export_recombinants function because we report the parents and the mutations from the zeroth element of the matches metadata array, and I think this is (sometimes?) the backwards pass, so doesn't correspond to the tree seq.
This PR uses the term "backward" and "forward" rather than indexes, so we don't get confused.
I think there's a bug in the current
export_recombinants
function because we report the parents and the mutations from the zeroth element of the matches metadata array, and I think this is (sometimes?) the backwards pass, so doesn't correspond to the tree seq.This PR uses the term "backward" and "forward" rather than indexes, so we don't get confused.