Two locus general matrix#16

Open

lkirk wants to merge 11 commits intoupstream-mainfrom

two-locus-general-matrix

Commits on Oct 26, 2025

initial stab at a general matrix (no normalisation)
lkirk
committed
added dimension dropping, but I think transposing is better -- we don't have to add a dimension at the end for scalar operations
lkirk
committed

Commits on Dec 7, 2025

do not test equality, this was useful on my local machine but is problematic in practice
lkirk
committed