TrackdNdxDelphesBased Fix

[s6anloes]

BEGINRELEASENOTES

• Adapted and changed logic of accepted $\beta\gamma$ range in TrackdNdxDelphesBased

ENDRELEASENOTES

I recently noticed that the delphes accepted range for $\beta\gamma$ was lowered (whether intentionally or not, see delphes/delphes#199) to $10000$.
Thus, in the run_digi_reco.py script in FCC-Config for IDEA, the 10 GeV electrons would silently be out of the range and produce nclusters=0.

This PR aligns the upper limit of accepted $\beta\gamma$ in TrackdNdxDelphesBased with the range from delphes/TrkUtil.
Also, the logic is slighty changed, so that if the value is above the range, it will set $\beta\gamma$ to the max value as an approximation. This way 10 GeV electrons will still produce clusters.

Cheers,
Andreas

[jmcarcell]

The comment doesn't match, it is below 0.5

[s6anloes]

Sorry, I don't see the mismatch?
Could you clarify?

[jmcarcell]

I meant that it is always below 0.5 so it doesn't make sense to say it's below 0.5-10000

[jmcarcell]

Then once it is fixed in Delphes what do you plan to do? Go back to the current version?

[s6anloes]

Then once it is fixed in Delphes what do you plan to do? Go back to the current version?

Well, the approximation of setting to the max value of $\beta\gamma$ is above the range can stay I think.
We can adjust the upper limit to 20000 if it is updated in delphes again, but in any case, the 20000 value was the same as the 10000 one. It was only introduced to fix the TSpline behaviour, see delphes/delphes#157.
So I don't expect much difference in the actual cluster per length values.
I think it could stay like this, regardless of the delphes fix

[ueinhaus]

The fix looks good!
It's fair to set the value in case of βγ > 10000 to the max value of the delphes function, since this is only relevant for electrons and their distribution is flat there anyway. And this can indeed be expanded/adapted following changes in delphes, though we should in general aim to fully move from this parametrisation to full sim.
One minor thing: Would it make sense to change the current out-of-range warnings to info() or even debug()? In physics events there should be regularly particles outside of the given βγ range, leading potentially to many warnings with little inherent value?