pmda
pmda copied to clipboard
MDAnalysis
Parallel algorithms for MDAnalysis
## Aim Turn ParallelAnalysisBase into a custom dask collection (https://docs.dask.org/en/latest/custom-collections.html). ### Current syntax ```python u = mda.Universe(TPR, XTC) ow = u.select_atoms("name OW") D = pmda.density.DensityAnalysis(ow, delta=1.0) # Option one (...
Fix #131 Still on-going, showing some possible simplifications after `Universe` can be serialized. PR Checklist ------------ - [ ] Tests? - [ ] Docs? - [ ] CHANGELOG updated? -...
### Expected behaviour pmda can deal with any universe/trajectory that can be built with MDAnalysis ### Actual behaviour Fails with *TypeError* "TypeError: can't pickle generator objects" (see Stack trace below)...
The PMDA version of the new h-bond analysis code - raises ValueError - check `u.bonds` The serial version - raises NoDataError - checks u._topology.bonds in PR https://github.com/MDAnalysis/mdanalysis/pull/2572 This came out...
Fixes #76 Changes made in this Pull Request: - add scheduler tests - make it work for all schedulers PR Checklist ------------ - [x] Tests? - [ ] Docs? -...
When initializing a parallel analysis we can hand it several atomgroups. The assumption is that all atomgroups belong to the same universe and can be *unpickled* in `_dask_helper`. This assumption...
The memory reader will not work currently because we are recreating a universe for each process. For it to work we need to special handle it. As a first step...
It's currently not possible to use updating atomgroups because we use the indices of an atomgroup to rebuild it for a thread. We don't even show a warning right now...
Parallelize obtaining a timeseries of dihedral angles, e.g. - backbone angles (phi, psi, omega) - user-defined angles This will be useful for Ramachandran maps and dihedral PCA and other dimensionality...
Very quick issue with varying things that came up during PR #66; apologies for the messy report. See PR #81 for initial (failing) tests. # n_jobs LeafletFinder with `n_jobs ==...