Consider adding ignore_diags to sample

[Phlya]

This would allow one to more fairly sample libraries in cases of divergence in very short range contact frequency (very typical, and accounts for a large % of total contacts).

Should this be done only in the case of specified cis_count or for global count as well?

[gfudenberg]

I could see it being useful for frac as well

[Phlya]

How would it work for frac?..

[gfudenberg]

I'd imagine just the fraction of the total counts after ignoring some # diagonals

…

On Thu, Mar 31, 2022 at 10:11 AM Ilya Flyamer ***@***.***> wrote: How would it work for frac?.. — Reply to this email directly, view it on GitHub <#347 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEV7GZOUCFVD7CMF7OBN5GTVCXMFRANCNFSM5SFVOTSQ> . You are receiving this because you commented.Message ID: ***@***.***>

[Phlya]

But fraction is independent of the total number of contacts...

[gfudenberg]

not if exact is used?

cooltools/cooltools/api/sample.py

Line 93 in 2583684

count = np.round(frac * clr.info["sum"])

[Phlya]

Sorry, I still don't understand...

[gfudenberg]

consider merging with issue #458