Wasn't running because the hist service was not being configured correctly
Fixes ATR-28759
Mark Sutton (26ef2c24) at 22 Mar 22:01
Run the TIDA Tier0 monitoring correctly
... and 92947 more commits
This line should be removed now, as the default Roi width is now used for the jetSuper instance.
OK, I had already commented oin the rest before I saw this, so I will leave them there, but feel free to resolve them if they cesase to be relevant. I will ressolve this already in anticipation.
This is fine here at the moment, but at some point we should move this out of the tracking cofiguration to somewnhere more central in the steering configuration since it sets the default RoiDescriptor z width for all RoiDescriptors, not just the tracking.
Again, really we should have this controlled by a flag in the code for the specific signatures, otherwise we have some signatures behaving in one way, and others in a different way, and it would not be obvious why.
This is almost completely opaque code, can it be simplified or at least commented on what it is trying to do and how ?
Are these "modes" ever autogenerated in the calling functions ? or are they known a-priori at the point of access to the signatureTrigTrackingFlags()
call ?
Similarly, these signature specific settings should be moved to their relevant functions
we want to avoid this sort of ad hoc if
setting, so it should be set to 12 here, and then set to 9 in the specific bjet function on line 315
Thanks TJ
So it looks as if the CI tests ran successfully this time, so we should be good to go.
Yes, the location is sadly, somewhat obscure - the original intention was that the default setting was not ID Trigger related at all, and we really wanted it to be more Core software based, as for the previous hardcoded default RoiDescriptor width which was in the RoiDescriptor code itself.
But as that is not an algorithm, we needed some Core, non-tracking related algorithm to set it - it doesn't really affect the signatures, it only affects the default value for the RoiDescriptor constructor if a zedWidth is not provided to the RoiDescriptor constructor, exactly as before. So really it should be thought of as no different from any other default steering parameter provided in the code, so I don't think that it should have any specific comment in the signature code, any more than other default parameters.
If we move it out of the TrigConfigSettings, as it should not really have been there anyhow, it might make things more obvious, and we can add appropriate comments then.
OK, I see what you mean. The global setting should not be set in the signature based ID Trigger config settings as they are for signature specific settings, so it instead has a mechanism to set it once and only once for all Rois in the HLT Steering separate from the algorithm configuration in
https://gitlab.cern.ch/atlas/athena/blob/24.0/Trigger/TrigSteer/HLTSeeding/src/HLTSeeding.cxx#L51
so it is a parameter of the HLTSeeding, and at the moment, once set for a run, it cannot be changed again, to avoid any unintended changes throughout a run. If we ever decided that we needed to update the global Roi width during a run, we would then only need to modify the setting code for this as appropriate.
Where it is actually set is from
which, for the time being, is somewhat unpleasantly being configured from here
Cheers Mark
This now uses the same global 180 mm used across the rest of the trigger, so if we change that, it will change for the superRoi and for all the other Rois in the event. The option is still available to update the JetSuper width individually if required so the same functionality is there, but it now uses the same Roi width as everywhere else, rather than setting the width separately to the same as used by the rest of the trigger.
Many thanks Tim.