HATCHet2 RDR/BAF locally-aware clustering error with sparse bins

[aganezov]

First of all, thank you to all on HATCHet2 team for the recent release and bringing on exciting development of locally-aware RDR/BAF bin clustering (vs global clustering in HATCHet1).

With that mentioned, when trying to run on HATCHet2 on some data that contains occasional gaps between bins on a given chromosomal arm (e.g., long stretches of homozygosity and/or identity by descent in non-human organisms, where BAF values are impossible to obtain; or cases of WES where gaps define intergenic regions) I get an error mentioning that there is more than 1 gap per chromosome (coming from this code: https://github.com/raphael-group/hatchet/blob/master/src/hatchet/utils/cluster_bins.py#L158C1-L159C1 presumably).

What would be the best way to utilize HATCHet2 locally-aware RDR/BAF bin clustering while allowing for non-contiguous bin runs within a given chromosomal arm?
I'd imagine that currently every chromosomal arm is treated independently as a contiguous stretch of boundary-touching bins, while, in principal, this can be generalized for for a case of having more than 1 "bin-run" along a chromosomal arm, where every such stretch of contiguous bins is treated independently as is currently the case for chromosomal arms.

Is it something that can be readily tried with the current HATCHet2 code base? Are there any workarounds for it?

Thanks!