Runtime pT input + low pT occupancies (rebase PR39 and PR54)

[ariostas]

This is a rebase of PRs #39 and #54. They both deal with pt cutoffs, so I grouped them together.

One thing we have to think is whether adding the notebook would be okay for CMSSW since it's a relatively large file (2.4MB). I think maybe we should move it to another SegmentLinking repo or at least strip out the outputs.

[slava77]

This is a rebase of PRs #38 and #54.

I guess it should be #39

Does this already have conflicts with other rebases? I'm curious when we should start staging them out (merging to the devel_rebased ) branch

[ariostas]

I guess it should be #39

Oops, fixed it.

Does this already have conflicts with other rebases? I'm curious when we should start staging them out (merging to the devel_rebased ) branch

I checked and there are no conflicts between them.

[GNiendorf]

One thing we have to think is whether adding the notebook would be okay for CMSSW since it's a relatively large file (2.4MB). I think maybe we should move it to another SegmentLinking repo or at least strip out the outputs.

Yeah I agree with moving it to another repo, no reason to include it in CMSSW. Same with the DNN training notebook, but maybe other people have different opinions.

[slava77]

One thing we have to think is whether adding the notebook would be okay for CMSSW since it's a relatively large file (2.4MB). I think maybe we should move it to another SegmentLinking repo or at least strip out the outputs.

Yeah I agree with moving it to another repo, no reason to include it in CMSSW. Same with the DNN training notebook, but maybe other people have different opinions.

I checked on MM, it sounds like there is some support to have notebooks for documentation. ( I mentioned ~< 1MB size)

[ariostas]

I checked on MM, it sounds like there is some support to have notebooks for documentation. ( I mentioned ~< 1MB size)

I tried using ipynbcompress to reduce the resolution of the images to get it down to ~300KB. The images are pretty blurry, but you can still make out the important features. Alternatively, we could just remove all images.

https://gist.github.com/ariostas/91c12257271a72cbd2245a0a9d99043b

[ariostas]

/run all

[github-actions]

The PR was built and ran successfully in standalone mode. Here are some of the comparison plots.

The full set of validation and comparison plots can be found here.

Here is a timing comparison:

   Evt    Hits       MD       LS      T3       T5       pLS       pT5      pT3      TC       Reset    Event     Short             Rate
   avg     40.6    384.4    191.7    128.3    135.1    500.5    117.7    245.1    102.0      2.9    1848.2    1307.2+/- 365.6     491.0   explicit[s=4] (target branch)
   avg     40.5    388.4    183.1    127.2    136.0    546.7    118.5    246.3    108.4      3.0    1898.1    1310.9+/- 365.8     501.2   explicit[s=4] (this PR)

[github-actions]

The PR was built and ran successfully with CMSSW. Here are some plots.

OOTB All Tracks

The full set of validation and comparison plots can be found here.