Updated notebooks/experimental storage #984
Conversation
Minimal changes made to fit new usage. I've rewritten bits like the Engine to be taken directly from the paths module to make sure the monkey_patch_all passed through. Only openpathsampling.visualise is imported separately in notebook 3, but this doesn't actually have anything to do with storage as it's just visualising the data
Codecov Report
@@ Coverage Diff @@
## master #984 +/- ##
==========================================
- Coverage 80.59% 80.58% -0.01%
==========================================
Files 138 138
Lines 14671 14671
==========================================
- Hits 11824 11823 -1
- Misses 2847 2848 +1
Continue to review full report at Codecov.
|
I fixed this this morning, literally just a couple hours before your commits. I haven't even opened the PR yet. I'll be combining it with some additional testing for SimStore. Here's the commit: dwhswenson@6ef1284 We'll keep this on hold until that can get merged in (I expect to open the PR on that tomorrow; probably to be merged by end of weekend). |
|
With #985 merged in, could you try against that? It should now work without needing to build the |
update to most recent
|
I've rerun with the lastest master branch pulled in, can confirm that old style access to CVs is functional again. |
Updated to use storage.cvs['cv_name'] Also limited number of cpu cores in md_run options '-nt 4' as errors caused if it uses >=16 cores because domains too large to decompose
Not sure how you want to manage it, but I updated the gromacs tps example to use the experimental SimStore. It works. As currently CVs are under 'storable_functions' I make a dict then select for choosing named cv's. Other than that it's been dropped in.
Obviously don't accept this pull request. I'm doing it to highlight the commit, which may be useful to you.