Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.17%. Comparing base (8292193) to head (2801508).
⚠️ Report is 597 commits behind head on master.

@@           Coverage Diff           @@
##           master    #1005   +/-   ##
=======================================
  Coverage   81.17%   81.17%           
=======================================
  Files         139      139           
  Lines       15143    15143           
=======================================
  Hits        12292    12292           
  Misses       2851     2851           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

@sroet, @hejung, @gyorgy-hantal (and anyone else who has some old CVs from netcdfplus and is comfortable trying an experimental branch): Could you try this branch out and see if the from_netcdfplus_cv methods are working correctly? They work on the things I have tested, but those are based on the way that I create CVs; just checking to see if there are use cases I'm missing.

The following script is a rough template of I hope you can try, especially for any CV that might have an unusual setup:

import openpathsampling as paths
from openpathsampling.experimental.storage.collective_variables import MDTrajFunctionCV

storage = paths.Storage("old_file.nc", mode='r')
old = storage.cvs['phi']  # was originally defined as a paths.MDTrajFunctionCV

# create the new CV using the classmethod `from_netcdfplus_cv`
new = MDTrajFunctionCV.from_netcdfplus_cv(old)

snap = storage.snapshots[-1]  # take an arbitrary snapshot from storage
assert old(snap) == new(snap)

So far, the CV types in the following table have been implemented in the new storage. (Only FunctionCV changed its name, because SimStore naming changed a little to avoid some domain-specific things in netcdfplus.) If you have a different CV type, let me know and I can prioritize setting up those, as well -- I assume these are the most-used CV types.