New option for function make tile: keep_chroms

[prauten]

I added a new optional option to the make_tile function that allows you to select those chromosomes from which you want to keep the fragments. If this option is enabled, the remove_chroms option is ignored. If keep_chroms is not used as an argument, the function runs as before.

Unfortunately, opening the files in PyCharm seems to have changed some of the indentation and spaces, too.

I read the contributing guidelines but could not comply with all of them yet.

For reference, they are:

1. Include passing tests (run tox);
2. Update documentation when there's new API, functionality etc.
3. Add a note to CHANGELOG.rst about the changes.
4. Add yourself to AUTHORS.rst.

Regarding 1: I ran tox, not all tests succeeded, but this is not due to the novel added functionality but must stem from previous issues.

Regarding 2: I am happy to add a line about this in the jupyter notebook if you want to.

Regarding 3: The CHANGELOG.rst is currently not up to date, such that I disregarded this for now.

Regarding 4: Done.

[wkopp]

Hi Pia, thank you for your contribution and request. In general, I'm fine with adding this functionality.

I'd have two small requests:

1. Could you add the additional argument in the docstring of the function so that it's clear what input is expected. From what I understand, I'd assume keep_chroms: None or list(str), right?
2. There is a bit of repetitive code by using if keep_chrs is not None: ... else ... now. In general, it's better to keep repetitive code to a minimum amount, if possible. Here, I think, this would be possible by simply making the selection of the keep_chroms before the for chrom in genomesize: loop on the genomesize dictionary. That is by putting genomesize = {chrom: genomesize[chrom] for chrom in keep_chroms} if keep_chroms is not None before the loop. That way the rest of the function would stay the same and be a bit slimmer. Would you mind adapting it that way? Thank you very much!

Best, Wolfgang

[prauten]

Hi Wolfgang,

Thank you very much for the quick review and for agreeing with the addition of this feature.

1. Done!
2. Great suggestion! I changed it the way you proposed which slims it down a lot.

Just one more thing I noticed, the docstring and default of the make_counting_bins function option remove_chroms is an empty list. However, the way the command line parsing is set up, this will be ['chrM', 'chrY', 'chrX'] unless you specify the remove_chroms argument without input on the command line. Maybe this is what you want, but it is not intuitive for me that I have to add --remove_chroms NOTHING, in case I don't want to filter any chromosomes on the command line.

Best, Pia

[wkopp]

Thank you! Regarding the default parameters. As long as the defaults are documented at all levels, I'd say it's fine. Even if they differ. True, you might need to specify a dummy chromosome to prevent the filtering from the command line. However, I never really needed to do that previously (e.g. to keep mitochondrial reads). In my experience, more often than not, you would want to exclude some chromosomes/contigs.