Skip to content

Conversation

@dwhswenson
Copy link
Member

OpenPathSampling 1.5 is a significant release, including many new features and enhancements. Highlights of this release are the implementation of Spring Shooting and the development of new "hooks" where users can patch in custom code during the simulation process, enabling much greater flexibility.

Note that the hooks are still considered beta code, and therefore some aspects of that API may change;

DEPRECATION: As of OPS 1.5, MDTrajTopology should be imported from openpathsampling.engines, not from openpathsampling.engines.openmm.

New features and enhancements

  • Spring Shooting mover/move scheme (#850 @sroet)
  • "Hooks" to enable more simulation flexibility (#755, #911 @hejung)
  • Gracious Kill Hook (#914 @hejung)
  • Warn if CVDefinedVolume CVs return iterables (#1030)
  • Shooting point selector with arbitrary bias (#1024)
  • Add global random number generator; support in shooting pt selectors and pathmovers (#998 @sroet, #1008 @sroet #1015)
  • Move n_dofs to engine; add engine.has_constraints() (#979)

Experimental (beta) features

Bugs fixed

  • Make histogramming more robust (#1029 @sroet)
  • Fix problem with repeated ensemble names in MISTIS (#988)
  • Fix MISTIS minus ensemble not forbidding other states (#983)
  • Fix problem with Histogram.bin_range not getting set (#972)

Miscellaneous improvements

dwhswenson and others added 30 commits January 29, 2018 12:49
This will enable analysis, even with parallel hooks
dwhswenson and others added 23 commits June 18, 2021 12:56
Warn if CVDefinedVolume CVs return iterables
"Hooks" for ShootFromSnapshotsSimulation (i.e., committor)
…dified storage hook to also suit pathsampling
…ties to ensure a smooth user experience; PathsimulatorOutputHook now uses regex matching to allow for some leeway in the timing output
This will make it easier to implement other approaches to running
path sampling. For example, this should make it possible to:

* clean up run_until_decorrelated
* allow running until a certain number of accepted trajectories
* use algorithms that use a replica reservoir
@codecov
Copy link

codecov bot commented Jul 5, 2021

Codecov Report

Merging #1031 (82b5284) into stable (8c46f25) will increase coverage by 1.26%.
The diff coverage is 91.66%.

Impacted file tree graph

@@            Coverage Diff             @@
##           stable    #1031      +/-   ##
==========================================
+ Coverage   80.25%   81.51%   +1.26%     
==========================================
  Files         136      140       +4     
  Lines       14449    15399     +950     
==========================================
+ Hits        11596    12553     +957     
+ Misses       2853     2846       -7     
Impacted Files Coverage Δ
openpathsampling/collectivevariable.py 75.32% <ø> (+9.78%) ⬆️
openpathsampling/step_visualizer_2D.py 17.28% <0.00%> (ø)
openpathsampling/engines/openmm/topology.py 33.33% <33.33%> (-20.38%) ⬇️
openpathsampling/engines/topology.py 61.53% <57.62%> (-38.47%) ⬇️
openpathsampling/pathsimulators/path_sampling.py 72.26% <69.38%> (-4.86%) ⬇️
openpathsampling/pathsimulators/path_simulator.py 92.30% <92.30%> (ø)
openpathsampling/numerics/histogram.py 83.98% <92.85%> (+0.48%) ⬆️
openpathsampling/__init__.py 100.00% <100.00%> (ø)
openpathsampling/beta/hooks.py 100.00% <100.00%> (ø)
openpathsampling/collectivevariables/__init__.py 100.00% <100.00%> (ø)
... and 103 more

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 8cb8f3f...82b5284. Read the comment docs.

@dwhswenson
Copy link
Member Author

My automatic release note writing code has some issues with large releases, which was compounded by a change GitHub made which made manually creating these parts of these release notes a bit more of a headache. I'm hoping that contributors can double-check that all their contributions to this are included. From the commit list, that's @sroet and @hejung (and especially @sroet). If there are any PRs you've contributed to this release that got skipped in my manual process, please let me know. (Of course, release notes can also be edited later to correct things, so no worries if something slips through.)

I will leave this open for at least 24 hours, merging no earlier than Tue 06 Jul 19:00 GMT (15:00 my local). On merging, this should cut a new release (immediately on PyPI, expect a few hours before that makes its way to conda-forge).

@sroet
Copy link
Member

sroet commented Jul 6, 2021

LGTM, all PR's from me that have been updated since Jan 30 (last OPS release) and were merged into master are included.

One another note, we might want to add a place to track 2.0 release notes (so that things like #977 are not missed/forgotten)

@hejung
Copy link
Contributor

hejung commented Jul 6, 2021

LGTM! I dont recall any finished PRs from me except for the hooks stuff.
(Only the unfinished gromacs MDP parsing PR is still open from my side, which I will hopefully get to finish in the coming days but not in time for 1.5)

@dwhswenson
Copy link
Member Author

Thanks! I'll merge this later today, once the promised comment time window runs out (just in case anyone else has comments -- not likely).

One another note, we might want to add a place to track 2.0 release notes (so that things like #977 are not missed/forgotten)

We have a 2.0 label for that purpose. If I forget to label anything as 2.0 that should be, please let me know!

(Only the unfinished gromacs MDP parsing PR is still open from my side, which I will hopefully get to finish in the coming days but not in time for 1.5)

No worries -- that one will be great to have, but isn't something I've promised elsewhere with a specific version number. Merging in hooks (promised for 1.5) a was big weight off my shoulders. The things I've promised for 1.6 are in the 1.6 milestone (in particular #1001 and #1026), but of course there's room for other contributions to go in as well!

@dwhswenson dwhswenson merged commit d6d97a2 into stable Jul 6, 2021
@dwhswenson dwhswenson deleted the release-1.5.0 branch July 22, 2021 16:24
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants