Match electrons to photon conversion tracks

[bainbrid]

This PR is based on code in Nancy's ConversionSelector RR, #84.

• the main developments are in a new ConversionInfo class, which is used in the ElectronMerger module,
• pat::Electrons are matched to reco::Conversions from the gsfTracksOpenConversions collection,
• flags are stored in the electronBPark table to indicate if each electron is matched to a conversion, which may satisfy various WPs (open, loose, tight),
• several "raw" variables are also stored in the electronBPark table if a boolean configurable is set, which may be used to apply additional/tuned selections to the conversion candidates.

Please check carefully on the logic and comment if necessary!

@gkaratha @amartelli @nancymarinelli @ottolau @crovelli

[bainbrid]

Needs a little more work, but feel free to start taking a look ...

[bainbrid]

Ok, so this now ready for review @gkaratha @nancymarinelli.

Most of the code is here and here and it is used here.

This configurable (current default = True) controls whether we add several variables to the electron table or not. Otherwise, we just store simple flags that indicate if a match between an electron and a conversion was found, with various working points defined for the latter ("open", "loose", "tight"). These are based on code and presentations from Nancy and must be checked! Also please check the logic, which is based on Nancy's selection code here and matching code here.

Thanks!

[bainbrid]

all ok @gkaratha @nancymarinelli ?

[nancymarinelli]

seems ok as far as I can tell

…

On 09/11/20 23:42, bainbrid wrote: all ok @gkaratha @nancymarinelli ?

--

________________________________________ Nancy Marinelli Research Professor University of Notre Dame, IN, US CERN, Bdg 40/3-A01, 1211 Geneva SWITZERLAND Phone +41-22-76-70809 fax +41-22-76-78940

[gkaratha]

Hi @bainbrid ,
I think is fine as well
best,
george

[bainbrid]

Shall I merge? Do you want to test it first? (ie. run with it?) I have tested on MC and data, but only in a limited fashion using the provided example cfg file.

[bainbrid]

BTW, the default config will add to the tree:

• various flags that indicate if matched to a conversion
• as well as the extra variables shown here, which can be switched off using this configurable boolean.

[bainbrid]

Potential issues to address in a new PR:

• Check mass variable to cut on? Remove mass cut from Tight selection?
• Add deltaCotgTheta cut?
• Add radius cut? Relax radius cut to 0. < R < 12.