@dwhswenson Do you have any list of functionality that you want to have tested via this notebook? (So I might exchange this notebook with proper unit-tests if I run into this when I have some more time)

Codecov Report

Merging #1022 (cc268e6) into master (8910ac3) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1022   +/-   ##
=======================================
  Coverage   81.17%   81.17%           
=======================================
  Files         139      139           
  Lines       15143    15143           
=======================================
  Hits        12292    12292           
  Misses       2851     2851           

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 8910ac3...cc268e6. Read the comment docs.

Do you have any list of functionality that you want to have tested via this notebook? (So I might exchange this notebook with proper unit-tests if I run into this when I have some more time)

I would be very much in favor of converting this to proper unit tests. Also something I just haven't had time to do (it's less of a headache to just re-run when I wake up to a failing test.)

It looks to me like the notebook basically tests the following:

1. PyEMMA CVs give correct results (which might be better tested again MDTraj results on the same thing -- I think that, under the hood, PyEMMA is using MDTraj, but this can be used to ensure that it works correctly.
2. PyEMMA CV results are properly stored and can be reloaded from storage
3. PyEMMA CVs can be applied to iterables, returning an iterable (might also be good to check that snapshots return scalars)
4. PyEMMA CVs (the CV object itself) can be stored and reloaded

I suspect that actually putting that stuff into the real unit tests will make a decent bump in test coverage, because I suspect that a lot of the stuff tested here is what never gets covered anywhere else. (Of course, it's also stuff that will be removed in 2.0.)