MPI exceptions logging and more robust NetCDF closures #1084
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
This should be ready to be merged! |
jchodera
approved these changes
Sep 10, 2018
| # Open analysis file. | ||
| self._storage_analysis = self._open_dataset_robustly(self._storage_analysis_file_path, | ||
| mode, version=netcdf_format) | ||
| # TODO - AR: What's the purpose of set_auto_mask(False)? When we create the |
There was a problem hiding this comment.
We no longer need it since we're not using masking!
| @@ -373,13 +373,14 @@ def _mix_replicas(self): | |||
| # TODO: We may be able to refactor this to simply have different update schemes compute neighborhoods differently. | |||
| # TODO: Can we allow "plugin" addition of new update schemes that can be registered externally? | |||
| with mmtools.utils.time_it('Mixing of replicas'): | |||
| jump_and_mix = self._JumpAndMixPacket(self.n_replicas, self.n_states) | |||
| # Initialize statistics. This matrix is modified by the jump function and used when updating the logZ estimates. | |||
| replicas_log_P_k = np.zeros([self.n_replicas, self.n_states], np.float64) | |||
There was a problem hiding this comment.
This is much clearer now---thanks!
Yank/multistate/sams.py
Outdated
| for state_index in neighborhood: | ||
| u_k[state_index] = self._energy_thermodynamic_states[replica_index, state_index] | ||
| log_P_k[state_index] = self.log_weights[state_index] - u_k[state_index] | ||
| u_k = self._energy_thermodynamic_states[replica_index, state_index] |
There was a problem hiding this comment.
To match the equations, it would be clearer to write:
u_k = self._energy_thermodynamic_states[replica_index, :]
log_P_k[state_index] = - u_k[state_index] + self.log_weights[state_index] There was a problem hiding this comment.
Good idea! Will change.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I've made a few small modifications to fix the stack trace that is logged when an exception is raised in an MPI process and to make more robust the NetCDF handling. In particular,
AlchemicalPhase.__del__and makeyank.experiments.Experimentforce garbage collection after deleting the alchemical phase.Dataset.open/close()in the multi-state samplers.The
switch_phase_intervalis working much better for me after these modifications, which is hopeful.