Closed bas-rustenburg closed 7 years ago
Still seeing strange segmentation faults, now in code that isn't related to netCDF. Tests pass on the cluster. I'm going ahead with merging this.
See @Lnaden and the #yank channel on slack for info on how to work around the segfault by moving the openmm dev
package tests to a separate optional travis branch and using release openmm
for standard py27/35/36 tests.
More specifically, look at choderalab/openmmtools#275
Ah. Thanks, that is helpful.
This PR adds the BAR analysis and other related plotting tools to perform diagnostics and analyze results of protons simulations that use the netCDF reporter file formats.
Note that tests depending on the netCDF format have a tendency to segfault on linux, so they're not reliable enough to run on travis. I've verified that they pass locally, and this behavior has not been observed outside of the testing framework.