Amplitude and Data Reader Revisions#394
Conversation
|
Test status for this pull request: SUCCESS Summary: /work/halld/pull_request_test/halld_sim^shepherd_at_revisions/tests/summary.txt Build log: /work/halld/pull_request_test/halld_sim^shepherd_at_revisions/make_shepherd_at_revisions.log |
kevScheuer
left a comment
kevScheuer
left a comment
There was a problem hiding this comment.
Left a comment on adjusting the comment in the bootstrap reader, but everything else looks good!
| // | ||
| // 6 args: inFileName inTreeName numParticles {fourMomentumPrefix|friendFileName} {randSeed|friendTreeName} weightBranchName | ||
| // - If arg[4] is integer: first case of fourMomentumPrefix, randSeed | ||
| // - Otherwise: second case of friendFileName, friendTreeName | ||
| // | ||
| // 7 args: inFileName inTreeName numParticles friendFileName friendTreeName weightBranchName fourMomentumPrefix | ||
| // - Adds friend tree and custom prefix |
There was a problem hiding this comment.
Can you align the comments with the changes? Something along the lines of
| // 6 args: inFileName inTreeName numParticles friendFileName friendTreeName weightBranchName | |
| // | |
| // 7 args: inFileName inTreeName numParticles friendFileName friendTreeName weightBranchName {randSeed|fourMomentumPrefix} | |
| // - If args[6] is an integer, it's interpreted as a random seed | |
| // - otherwise it's a four momentum prefix |
3 participants
This adds a few amplitudes with GPU classes that are being used for b1 fitting. It does not include changes to Vec_ps class because the branch being used in analysis has diverged from the master.
Significant performance enhancements were obtained by caching all relevant sines and cosines for DeltaAngles. This reduces the amplitude calculation down to basic floating point operations.
In other places I cleaned up printing to the screen to use AmpTools report class. This avoids a repeated messages when using MPI and allows the user to adjust the level of verbosity.