[PWGHF] Make default sgSelector parameters configurable #14923
Conversation
Rrantu
commented
Feb 11, 2026
Rrantu
commented
- The default values of NDtColl, MinNBCs, MinNTracks and MaxNTracksFix in utilsUpcHf.h were not optimal and have been changed to Configurable parameters.
- Reduced the number of bins in the UPC QA histograms in taskLc.cxx to optimize memory usage.
Rrantu
requested review from
NicoleBastid,
alibuild,
apalasciano,
deepathoms,
fcatalan92,
fcolamar,
fgrosa,
gluparel,
hahassan7,
jpxrk,
mfaggin,
mmazzilli,
singhra1994,
stefanopolitano,
vkucera and
zhangbiao-phy
as code owners
github-actions
bot
changed the title
|
Hi @Rrantu , when these parameters were being added, my understanding was that these were already optimised constants taken from the UD. Can you please explain the motivation of your PR? |
Rrantu
changed the title
Hi @vkucera, the original default values were not fully consistent with the UD configurations. I have now made these parameters configurable based on the settings used in the UD analysis. |
| /// \brief Configurable group for UPC gap determination thresholds | ||
| struct HfUpcGapThresholds : o2::framework::ConfigurableGroup { | ||
| std::string prefix = "upc"; // JSON group name | ||
| o2::framework::Configurable<int> nDtColl{"nDtColl", static_cast<int>(defaults::NDtColl), "Number of standard deviations to consider in BC range"}; |
There was a problem hiding this comment.
Why is the casting necessary?
There was a problem hiding this comment.
Because the defaults::NDtColl is constexpr int and cannot bind to Configurable constructor directly.
There was a problem hiding this comment.
The issue is that the constructor takes an rvalue reference which cannot bind to NDtColl which is an lvalue. You can cast it to an rvalue reference with std::forward instead of static_cast.
There was a problem hiding this comment.
When I try std::forward(defaults::NDtColl), it gives a compilation error. It seems that NDtColl is a constexpr int, and std::forward doesn’t work with constexpr values. Could you please advise how to handle this?
2 participants
