Add an exception in DecayTreeFit when constraint cannot be applied

Closes #408 (closed) . Needs DaVinci!759 (merged)

added DPA-WP3 label

requested review from @jzhuo

mentioned in merge request DaVinci!759 (merged)

changed the description

added 1 commit

• 06a8ea80 - Fixed formatting

Compare with previous version

added RTA label

I git checkout this and did a test, and then I saw 800+ warnings...

But I think it's expected, the user should never add invalid constraints to the tree, actually before using the DTF, we should firstly call ParticleRangeFilter to filter unnecessary particles.

WDYT?

approved this merge request

Thanks for testing :) Yes, the problem is that DecayTreeFit is a class, not a Gaudi tool. There's no access to message level. So either one throws an exception or one prints a warning for each candidate.

Maybe the former is better. Just imagine the trigger prints that warning at 10 MHz.

marked this merge request as draft

added 2 commits

• 869fd54b - Throw an exception
• 0d7f196e - Throw an exception

Compare with previous version

changed title from Draft: Add a warning in DecayTreeFit when constraint cannot be applied to Draft: Add n exception in DecayTreeFit when constraint cannot be applied

marked this merge request as ready

changed title from Add n exception in DecayTreeFit when constraint cannot be applied to Add an exception in DecayTreeFit when constraint cannot be applied

This is now ready. Now the execution stops if the particle to be constrained is not in the decay chain. That may cause troubles if some users throw two different chains at DecayTreeFitterAlg. For instance if I merge J/psi and psi(2S) candidates and then want to constrain the dimuon in both cases. Before this MR I could do

    DTF = DTFAlg(
        Input=dimuons, MassConstraints=["J/psi(1S)", "psi(2S)"])

That would not work any longer.

It's a choice between protecting users from stupid mistakes and allowing them to be smart.

/ci-test DaVinci!759 (merged)

added ci-test-triggered label

• [2022-10-01 15:08] Validation started with lhcb-master-mr#5810

mentioned in issue Moore#482 (closed)

resolved all threads

merged

mentioned in commit f3b0626f

added Tuples cleanup labels

mentioned in issue Moore#485 (closed)

mentioned in issue Moore#487 (closed)