Semibin not respecting -p threads argument

[AroneyS]

Thanks for the fantastic binner! I'm getting many more bins from Semibin than any other binner.

Running Semibin within Aviary (https://github.com/rhysnewell/aviary/), I noticed that in the binning phase (logging: SemiBin[22080] INFO Start binning.), Semibin is using all the CPUs on the server (128) instead of 64.

Snakemake rule shell command (where threads=64):

SemiBin single_easy_bin -i {input.fasta} -b data/binning_bams/*.bam -o data/semibin_bins -p {threads} --self-supervised

[luispedro]

We cannot reproduce this, unfortunately

My guess is that this is some openMP/BLAS multithreading inside NumPy, which is a bit non-trivial to control, but I'll leave the issue open for now