[PWGLF] Minor adjustments and improvements to strangeness TOF PID

[ddobrigk]

This PR follows up on this work by:

• moves both V0 and cascade TOF PID tasks to only one to save resources (following the same rationale as applied to core services)
• removes completely all grouping from the resulting task to ensure maximum processing speed
• adds an 'instant gratification' getter: v0.tofLambdaCompatibility(nSigma); will check compatibility with the Lambda hypothesis if TOF is available and accepts if TOF is unavailable. This effectively leads only to background loss and no signal loss for any momentum. A similar getter exits for other species (tofAntiLambdaCompatibility, tofK0ShortCompatibility) and also for cascades (cascade.tofXiCompatibility and cascade.tofOmegaCompatibility).
• adds extra histograms to consider an effective e-loss correction for protons, since the calculation remains analytical without e-loss for the time being (which is extremely fast). However, notably, this is only relevant at the lowest momenta, and current calibration simply becomes more accepting at these values - and, remarkably, the "wrong species" bands are already rejected now.

In addition, an OO calibration is also provided now and can be used. It is most effective when combined with the new compatibility getters (no background loss) and especially for omegas, in which case one gets a factor 2x reduction of background at no cost of signal even within the subset of omegas that pass TPC PID. The reduction of background is negligible at lowest pT, but once it kicks in, it also goes to relatively high omega pT (due to decay kinematics). I'll prepare a presentation on this very soon and as soon as I manage, but in the meantime I'd use this PR to continue with light ions as well as to re-test Pb-Pb pass5. Tagging OO analysers who could also help with testing @spucillo @romainschotter @lhusova

[github-actions]

O2 linter results: ❌ 163 errors, ⚠️ 0 warnings, 🔕 0 disabled

[romainschotter]

Hi @ddobrigk ! Thanks a lot for the PR! It looks great!
I just have one comment: if we remove the lambdakzeropid task in this PR, can you please also remove the cascadepid task?
Thanks a lot!

[ddobrigk]

Hi @ddobrigk ! Thanks a lot for the PR! It looks great! I just have one comment: if we remove the lambdakzeropid task in this PR, can you please also remove the cascadepid task? Thanks a lot!

That's a very good point, one less task to compile :-) - removed! Thanks for having a look!