[MRG] plot a random subsample with 'sourmash plot --subsample'.

[ctb]

This adds --subsample <N> and --subsample-seed <R> to sourmash plot, which will plot a randomly chosen subset of size N, chosen using Python's random.shuffle, seeded with --subsample-seed. Note that the seed defaults to 1, which intentionally gives stable results when used with the same inputs.

Fixes #221.

Also fixes #334, detecting multiple ksizes/moltypes earlier.

• Is it mergeable?
• make test Did it pass the tests?
• make coverage Is the new code covered?
• Did it change the command-line interface? Only additions are allowed
  without a major version increment. Changing file formats also requires a
  major version number increment.
• Was a spellchecker run on the source code and documentation after
  changes were made?

[ctb]

@taylorreiter comments & review welcome!

[ctb]

Note, had to fix the version of the random number seed for version compat - see https://stackoverflow.com/questions/11929701/why-is-seeding-the-random-generator-not-stable-between-versions-of-python

[codecov-io]

Codecov Report

Merging #343 into master will increase coverage by 0.02%.
The diff coverage is 92.59%.

@@            Coverage Diff             @@
##           master     #343      +/-   ##
==========================================
+ Coverage   86.96%   86.99%   +0.02%     
==========================================
  Files          13       13              
  Lines        2018     2037      +19     
  Branches       36       36              
==========================================
+ Hits         1755     1772      +17     
- Misses        262      264       +2     
  Partials        1        1

Impacted Files	Coverage Δ
sourmash_lib/commands.py	90.21% <92.59%> (-0.02%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 069fb51...8faa1b2. Read the comment docs.

[taylorreiter]

@ctb My use case was wanting to visualize divergent samples in a group of 11,000 samples that should have had similar tetranucleotide frequency throughout. I had no hypothesis as to which samples would not have similar tetranucleotide frequency, but suspected there would be some.

This provides the first step. If a random subsample is selected, I would expect that some of the time a non-similar sample would be plotted. The next interesting step I think would be to select the N most similar samples to sample X, and plot these from a sourmash compare matrix. I could do this in R pretty easily.

Ran the code and liked the output!
test_plot_random.pdf

[taylorreiter]

Very quick! Nice tool!

[ctb]

Ready for review & merge, @betatim @luizirber !

[ctb]

& thanks for trying it out, @taylorreiter :)

[ctb]

whups! I see @taylorreiter has already approved it, so I'll merge when the tests pass :)