Small Shashlik reco speed improvement #4924
Merged
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Outdated suggestions cannot be applied.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.
@davidlange6 is this what you meant?
Functionally the same but the order of checks has been changed to improve speed. There is a minor side effect in that fewer debug printouts will occur when the previously-second "if" evaluates to false, so that the previously-first "if" (which would produce the printout) is skipped.
All plots are ttbar at 140 pileup in Shashlik; "before" plots include #4909.
Total event times before and after
This equates to a drop from 6.1+/-1.7 minutes to 5.9+/-1.8 minutes per event (mean+/-sqrt(variance)). Doesn't sound that impressive when stated like that, but they're the same events not random data.
User time by module before
User time by module after
Some modules appear to slightly increase in time (e.g. particleFlowTmp, particleFlowEGamma) but others significantly decrease (particleFlowBlock). I'm not sure how accurate the timing is so I suspect this is just noise.
Updated igprof file at /afs/cern.ch/user/g/grimes/public/SLHC16_patch1Preview/igreport_perf_ShashlikRecoAfterLangeSuggestion_15evt.res
@davidlange6, @lgray