NEW: Adds core beta diversity measures

[ChrisKeefe]

This PR includes:

• bray-curtis
• jaccard
• weighted and unweighted unifrac
• relevant citations

In addition, this PR attempts to impose consistent quoting across the package, with user-facing strings double-quoted, and non-user-facing strings, keys, etc single-quoted. See fd0e24e

[thermokarst]

Hey @ChrisKeefe, this PR is looking pretty good. I was poking around at this this morning and noticed a few things that I wanted to bring to your attention. There is a bit of an inconsistency between the util tests, the decorator utils, and the view functions (the actions). In particular, the inconsistency has to do with the "file-based" biom files. The decorators (and their tests) all are set up to work withstr filepaths, but the view annotations in your actions look like:

def faith_pd(table: BIOMV210Format, phylogeny: NewickFormat) -> pd.Series:

Note, the annotation for table is BIOMV210Format, not str. I think it would make sense to "consolidate" on one view type: BIOMV210Format. You can start to do that by following a few of the suggestions I have provided, inline.

[thermokarst]

Took a pass through the tests, they look great! I have a few minor requests (see inline). Thanks!

[ChrisKeefe]

@thermokarst, I'm running into some squirrelly issues (Multiple values for argument thread) that I thought were resolved when I run these methods through q2-diversity, and will have to do some troubleshooting tomorrow to determine the source of the issue. Sorry for any inconvenience.

[thermokarst]

Just a mis-wiring issue - the problem is in qiime2/q2-diversity#273, not in this PR. You just need to match up the right n_jobs vs threads params in the various methods:

applying this patch (below) to commit d2ec98091bc2c8c0ca05ec6c7c928dd5ae021d7b (in PR 273) fixes it up:

diff --git a/q2_diversity/_core_metrics.py b/q2_diversity/_core_metrics.py
index 47a7b89..4f1ab57 100644
--- a/q2_diversity/_core_metrics.py
+++ b/q2_diversity/_core_metrics.py
@@ -61,9 +61,9 @@ def core_metrics_phylogenetic(ctx, table, phylogeny, sampling_depth, metadata,
 
     dms = []
     dms += unweighted_unifrac(table=cr.rarefied_table, phylogeny=phylogeny,
-                              n_jobs=n_jobs)
+                              threads=n_jobs)
     dms += weighted_unifrac(table=cr.rarefied_table, phylogeny=phylogeny,
-                            n_jobs=n_jobs)
+                            threads=n_jobs)
 
     pcoas = []
     for dm in dms:

[thermokarst]

I'm going to merge this PR so that busywork can chew away at it, we'll address any fallout in additional PRs. Then, if everything goes to plan, you'll have a good dev env to work with in travis, which should help keep you lined up on qiime2/q2-diversity#273.

Also, I will do one sweeping pass over the whole repo tomorrow, to do a "holisitic" review.

[ChrisKeefe]

Thanks for pushing this through, @thermokarst. Totally possible it is a mis-wiring issue, but it may be elsewhere. The code I'm testing already looks like this:

I'll give it a thorough looking over when I'm fresh. :)

[thermokarst]

when I'm fresh

I think that sounds like a great idea.

---

Let's:

a) stop discussing here (a least for now), and instead move over to qiime2/q2-diversity#273
b) rather than sharing isolated snippets and error messages with no context, please use the tools we have available: push any problematic code up to github, that way we can discuss code & any relevant errors that crop up in the ci unit tests.

Thanks!

[ChrisKeefe]

For posterity, the multiple values bug mentioned above was addressed in #18.