Skip to content

TrigTauHypo: fix TrigTauTrackRoiUpdater when no FTF Core track is found

Bertrand Martin Dit Latour requested to merge martindl/athena:R22_TauTrkUpdatorFix into master

Hello,

When running the fast track finder for HLT taus, we update the ROI (eta, phi, z and corresponding widths) between the first and second steps of the FTF. As discussed in ATR-24395, it appears that the ROI is not updated in case no track is found at the 1st stage of the FTF. In Run2, we were updating the width of the ROI even when no 1st stage track was found, handing this over to the second stage, so that there is a chance to find tracks in the second stage as the eta/phi window is larger. This MR restores the Run2 behaviour.

The code was also cleaned up.

Apparently, runTrigART.py reports no change in trigger counts. This must be due to another problem (ATR-24371): tau candidates for which no track is found in the ROI are currently discarded, while they should be kept and reconstructed as 0-prong taus. Once ATR-24371 is fixed, I expect this MR to change the counts of 'perf' and 'RNN' chains.

Tagging @ademaria , @iriu , @carquin .

Cheers, Bertrand

Merge request reports