search

apcamargo / pycoverm

Simple Python interface to CoverM's fast coverage estimation functions
GNU General Public License v3.0
7 stars 2 forks source link

min_identity does not seem to work #6

Closed jakobnissen closed 2 years ago

jakobnissen commented 2 years ago

With a test BAM file, I can't actually manage to get any rows filtered away using min_identity. How does it work? Also, are you sure it's a float between 0 and 1, given that its internal Rust name includes "percent"?

Can you make it work on your side?

apcamargo commented 2 years ago

Thanks for the report!

Despite the name, the "percent" parameters work in the [0, 1] range. CoverM detects if the value is greater than 1 and then divides it by 100.

I remember that in my tests the output of pyCoverM matched those of CoverM perfectly, but I didn't put together automated tests... I might be wrong, though. Unfortunately I can't investigate the issue within the next couple of days. I'll try to look into it next week.

jakobnissen commented 2 years ago

I was being stupid - I thought rows in the returned matrix would be removed if all reads fell below the minimum read identity. But of course this is not the behaviour (nor should it be).

Closing as solved.

apcamargo commented 2 years ago

No problem :) Thanks for the update!