Open tvami opened 3 weeks ago
I mean, the times are ranging from tenths to hundreds of microseconds so I feel like the time changing is just a product of the performance of the runner that we happen to land on.
Sure that makes sense, let's just round them and get the diffs to zero with that solution!
It seems this wasnt enough and we need to go more course https://github.com/LDMX-Software/ldmx-sw/actions/runs/10746102897/job/29806990784#step:6:22
< [ SeedFinderProcessor ] 1 : AVG Time/Event: 0.0980 ms
---
> [ SeedFinderProcessor ] 1 : AVG Time/Event: 0.0931 ms
10284c10284
< [ SeedRecoil ] 1 : AVG Time/Event: 0.1364 ms
---
> [ SeedRecoil ] 1 : AVG Time/Event: 0.1349 ms
10296c10296
< [ Tagger_TrackFinder ] 1 : AVG Time/Event: 0.3939 ms
---
> [ Tagger_TrackFinder ] 1 : AVG Time/Event: 0.3829 ms
10298,10301c10298,10301
< [ Tagger_TrackFinder ] 1 : setup Avg Time/Event = 0.0021 ms
< [ Tagger_TrackFinder ] 1 : hits Avg Time/Event = 0.0441 ms
< [ Tagger_TrackFinder ] 1 : seeds Avg Time/Event = 0.0032 ms
< [ Tagger_TrackFinder ] 1 : cf_setup Avg Time/Event = 0.0043 ms
---
> [ Tagger_TrackFinder ] 1 : setup Avg Time/Event = 0.0020 ms
> [ Tagger_TrackFinder ] 1 : hits Avg Time/Event = 0.0434 ms
> [ Tagger_TrackFinder ] 1 : seeds Avg Time/Event = 0.0030 ms
> [ Tagger_TrackFinder ] 1 : cf_setup Avg Time/Event = 0.0041 ms
10303c10303
< [ Tagger_TrackFinder ] 1 : result_loop Avg Time/Event = 0.3382 ms
---
> [ Tagger_TrackFinder ] 1 : result_loop Avg Time/Event = 0.3283 ms
10305c10305
< [ Recoil_TrackFinder ] 1 : AVG Time/Event: 0.3056 ms
---
> [ Recoil_TrackFinder ] 1 : AVG Time/Event: 0.3026 ms
10308c10308
< [ Recoil_TrackFinder ] 1 : hits Avg Time/Event = 0.0451 ms
---
> [ Recoil_TrackFinder ] 1 : hits Avg Time/Event = 0.0437 ms
10310,10312c10310,10312
< [ Recoil_TrackFinder ] 1 : cf_setup Avg Time/Event = 0.0020 ms
< [ Recoil_TrackFinder ] 1 : ckf_run Avg Time/Event = 0.0001 ms
< [ Recoil_TrackFinder ] 1 : result_loop Avg Time/Event = 0.[25](https://github.com/LDMX-Software/ldmx-sw/actions/runs/10746102897/job/29806990784#step:6:28)49 ms
---
> [ Recoil_TrackFinder ] 1 : cf_setup Avg Time/Event = 0.0021 ms
> [ Recoil_TrackFinder ] 1 : ckf_run Avg Time/Event = 0.0002 ms
> [ Recoil_TrackFinder ] 1 : result_loop Avg Time/Event = 0.25[31](https://github.com/LDMX-Software/ldmx-sw/actions/runs/10746102897/job/29806990784#step:6:34) ms
Is your feature request related to a problem? Please describe. Logs always fail now because of the character count different
Describe the solution you'd like
We should round these numbers to a fixed number of digits. But also it's interesting that rerunning ACTS on the same events takes different times.
Describe alternatives you've considered
Figure out why those numbers are different, is that expected @bloodyyugo
Additional context
PR test here https://github.com/LDMX-Software/ldmx-sw/actions/runs/10496314488/job/29077738861?pr=1401