Average structure not being distributed to all ranks in parallel #1096
Comments
drroe
added a commit
to drroe/cpptraj
that referenced
this issue
Nov 5, 2024
ranks for future use. Addresses Amber-MD#1096.
drroe
added a commit
that referenced
this issue
Nov 5, 2024
…' and 'average' (#1113) * Add test for handling generated data set vs read-in dataset * Test that a generated set can be used as a read-in set once processing is complete. * Add broadcast function * Ensure set is properly broadcast if it was previously generated but we are using it in a new trajectory processing run * Enable parallel dataset handling test * Add broadcast to the coords frames data set. Make sure we use long int when broadcasting dataset size. * Add data broadcast to frames coords set * No need for a broadcast, all data is on the disk * Add broadcast * Add broadcast for string set * Add Sync and Bcast to tensor dataset * Add bcast * Add broadcast for vector data * Add bcast * Ensure 'average' broadcasts the final average coords set to non-master ranks for future use. Addresses #1096. * Version 6.29.7. Revision bump for fixed broadcasting of data created during trajectory processing in a subsequent trajectory processing run for rotate and average commands. --------- Co-authored-by: Daniel R. Roe <[email protected]>
|
Should be addressed by #1113. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
See #1083.
The text was updated successfully, but these errors were encountered: