Fix bug in PVAssignment: missing abs for dxy

[arizzi]

This fixes the bug found by @VinInn of missing abs in PV association

Same as #8721

@davidlange6 @slava77 can we include this in 741 in the shadow of the PF nan fix?

[cmsbuild]

A new Pull Request was created by @arizzi for CMSSW_7_4_X.

Fix bug in PVAssignment: missing abs for dxy

It involves the following packages:

CommonTools/RecoAlgos

@cmsbuild, @cvuosalo, @nclopezo, @slava77 can you please review it and eventually sign? Thanks.
@ahinzmann, @jdolen, @rappoccio, @abbiendi, @jhgoh this is something you requested to watch as well.
You can sign-off by replying to this message having '+1' in the first line of your reply.
You can reject by replying to this message having '-1' in the first line of your reply.
If you are a L2 or a release manager you can ask for tests by saying 'please test' in the first line of a comment.
@Degano you are the release manager for this.
You can merge this pull request by typing 'merge' in the first line of your comment.

[slava77]

@cmsbuild please test

[slava77]

@cmsbuild please test
trying one more time

[cmsbuild]

The tests are being triggered in jenkins.

[cmsbuild]

+1
Tested at: fcd5dab
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-8722/4202/summary.html

[arizzi]

Here some more number testing also the original version of this PR (i.e. including "OtherDZ")

ZH->nunubb
no preselection
82.5%(current PR) 83.0% (buggy) 83.3%(original version of this PR) vs 71% (old PV assignment)

high pt preselection (analysis like)
89.4% (current) 90.0% (buggy) 90.3% (original PR) vs 81% (old PV ranking)

VBF Hbb
leading jet preselection (analysis like)
95.8% (current) 95.7% (buggy) 95.6% (original PR) vs 92.3 (old PV ranking)

so we can stay with current version because it is considered safer and then fine tune in next released (for example using the unassigned category for stuff that would be "OtherDZ" but with unreasonably high I.P.)

[slava77]

there was an operator error on my end in running the tests, restarted to run correctly now

[slava77]

+1

for #8722 17da848
tested additionally locally in CMSSW_7_4_0 /test area sign537/

In the tests running, differences in monitored quantities are most visible in min-bias-like events (several similar vertices): fairly rarely the leading vertex changes the downstream effects are then in PF neutrals (slight momentum change) and then more downstream plots.
There are no useful plots to attach here.

Looking at subleading vertices: the order changes more frequently.

I looked manually at larger changes in the score values (stored in value maps) and the ones with changes of more than a few units ("GeV**2") are from cases where there are tracks or clusters of tracks fairly far away (at least 5 cm away from the vertex that had this track counted in the vertex score earlier).
This was a low stat check of just 2 events (out of 100 running, among the cases with vertex order changes). Still, the behavior is as expected: skipping of OtherDz tracks in score computation just throws away unrelated tracks.

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next CMSSW_7_4_X IBs unless changes (tests are also fine). This pull request requires discussion in the ORP meeting before it's merged. @davidlange6, @Degano, @ktf, @smuzaffar

[davidlange6]

+1