Skip to content

[ATR-21186] Better roi for egamma

Tim Martin requested to merge tamartin/athena:betterROIForEgamma into master

Updates ROI Tools used by egamma slice

  • Step 1 uses L1 ROI
  • Step 2 forms a new ROI around the clusters found in Step1
    • Either a SuperROI is formed around all clusters
    • Or, the highest Et cluster is chosen. This is the current configured version, due to a problem with using the super ROI.
  • Steps 3, 4 (soon, 5), use the previous ROI tool and hence keep within the ROI defined at the start of Step2.

The ROI is currently set to eta +- 0.05 and phi +- 0.1 for both electron and photon. These numbers can be updated in the future.

Should a single L1 ROI be active for electron and photon chains entering the common Step3, the electron ROI will be used rather than the photon one. Note that this subtly cross-correlates the chains. As such, we should probably stick with the same eta and phi extent in both.

A currently unused option is added to the previous ROI tool. This will let it write out a copy of the ROIs which it chooses, rather than simply re-using the existing ElementLinks. This may be used in the future to provide disambiguation based on ROI collection SG keys.

The initial CI will fail as reference files are not yet updated.

Edited by Tim Martin

Merge request reports