Skip to content

clam and -t (--threads) #340

Description

@brantfaircloth

Hi y'all,

As noted in #333, I've been testing clam w/ loci having highly variable coverage. One thing that I've noticed in these runs (and during snparcher runs, more generally) is that clam seems to mostly run in single-threaded mode. As a result, test runs inputting different parameters for clam are [really] slow.

After activating the correct conda environment, i've been running clam manually:

clam loci -o bcf-callable_loci.INF.zarr -t 24 --per-sample -m 10 depths.zarr/ &> bcf-callable_loci.zarr.INF.log

At first, there seems to be some activity involving multiple cores/threads, but after a while, CPU activity drops from >100% (usually around 400-600%) to 95-100%, where it stays. I haven't investigate the clam code - so this may be expected behavior (although it seems suspect).

This is clam 1.1.4 which was packaged up in snparcher v2.1. Also happy to move this over to the clam page, but I figured I would report here given #333.

Thanks!
-b

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions