Add Precision Photon Hypo

This MR is an important step towards photon chain completion. It adds a new precisionPhotonHypo algorithm and tool with its structure based on precisionCaloHypo in order to be used in a final HLT photon chain. It relates to ATR-18929 and ATR-19912.

In addition it fixes unit tests fails in ATR-20048

added Egamma Trigger TriggerMenu master labels

This merge request affects 1 package:

• Trigger/TrigHypothesis/TrigEgammaHypo

added review-pending-level-1 label

CI Result FAILURE

	Athena	AthSimulation
externals
cmake
make
required tests
optional tests

Full details available at NICOS MR-24951-2019-07-18-05-39
Athena: number of compilation errors 0, warnings 0
AthSimulation: number of compilation errors 0, warnings 0
For experts only: Jenkins output [CI-MERGE-REQUEST-CC7 1245]

Hi @fernando The error in the unit test seems to be related to your changes. I am going to change label until you fix this. -Ruchi (L1)

added review-user-action-required label and removed review-pending-level-1 label

This is reported in https://its.cern.ch/jira/browse/ATR-20048

I think it's just because the hypo unit tests uses chain name without L1 name appended. Should be easy to fix.

In the ticket Peter claims he fixed this for all tests, but I think he only did for Jet hypo.

Hi @rbielski Its weird because I didn't add any new chain, nor changed anything in the Ls2_v1.py menu! and no configuration for this new step is added. Why then the unit test would fail because of missing L1 items?.

Indeed I don't have any chain in LS2_v1.py without L1 item:

TriggerFlags.EgammaSlice.signatures = [
        # ElectronChains----------
        ChainProp(name='HLT_e3_etcut1step_L1EM3', groups=SingleElectronGroup),
        ChainProp(name='HLT_e3_etcut_L1EM3', groups=SingleElectronGroup),
        ChainProp(name='HLT_e5_etcut_L1EM3', groups=SingleElectronGroup),
        ChainProp(name='HLT_e7_etcut_L1EM3', stream=[PhysicsStream, 'express'], groups=SingleElectronGroup),
        # currently disabled, seems to be a problem with the precision calo step
        #ChainProp(name='HLT_2e3_etcut_L12EM3', stream=[PhysicsStream], groups=MultiElectronGroup),
        #ChainProp(name='HLT_e3_etcut_e7_etcut_L12EM3', stream=[PhysicsStream], groups=MultiElectronGroup),
        # enabling only one step
        #ChainProp(name='HLT_2e3_etcut1step_L12EM3', stream=[PhysicsStream], groups=MultiElectronGroup),
        #ChainProp(name='HLT_e3_etcut1step_e7_etcut1step_L12EM3', stream=[PhysicsStream], groups=MultiElectronGroup),

        #chain with asymmetric number of steps not yet working
        #ChainProp(name='HLT_e3_etcut1step_e7_etcut_L12EM3', stream=[PhysicsStream], groups=MultiElectronGroup), 

        # PhotonChains------------
        ChainProp(name='HLT_g5_etcut_L1EM3', groups=SinglePhotonGroup),
    ]

Unit tests are failing locally too :-(

Because your unit tests don't use the menu. They just test a hard-coded name:

https://gitlab.cern.ch/atlas/athena/blob/master/Trigger/TrigHypothesis/TrigEgammaHypo/CMakeLists.txt#L67-74

https://gitlab.cern.ch/atlas/athena/blob/master/Trigger/TrigHypothesis/TrigEgammaHypo/python/TrigL2ElectronHypoTool.py#L71-83

Thanks a lot Rafal! Ill fix it right away. Then the UT was already broken. I think, we need to rethink the UT for these packages on trigegamma side.

Cheers!

Fer

added 1 commit

• 8a812dd1 - Fixed TrigEgammaHypo unit tests

Compare with previous version

Am pushing again. Unit tests needed to be migrated. I added all L1 names to the test chains there and moved from:

decodedDict = decoder.analyseShortName(conf, [], "") # no L1 info

to

decodedDict = decoder.getChainDict(conf)

Commented out the asymmetric chains as there is an issue with the multiplicity and the L1 item, and is not important for this particular unit test

This merge request affects 1 package:

• Trigger/TrigHypothesis/TrigEgammaHypo

added review-pending-level-1 label and removed review-user-action-required label

CI Result SUCCESS

	Athena	AthSimulation
externals
cmake
make
required tests
optional tests

Full details available at NICOS MR-24951-2019-07-19-03-20
Athena: number of compilation errors 0, warnings 0
AthSimulation: number of compilation errors 0, warnings 0
For experts only: Jenkins output [CI-MERGE-REQUEST-CC7 1288]

added review-approved label and removed review-pending-level-1 label

added review-pending-level-2 label and removed review-approved label

It all looks ok. Passing on to L2 for another check. (Clicked on wrong label by mistake) Ruchi (L1)

mentioned in merge request !24482 (merged)