PWGHF: Lc task: Remove the redundant Configurable and keep using the Filter

[zhangbiao-phy]

• remove the redundant Configurable "applyML"
• replace the "Partition" with "Filter"
• FixesPWGHF: Add ThnSparse of the ML scores vs Vars to the Lc task #5040

[zhangbiao-phy]

Hi @vkucera@DelloStritto, this is the new PR based on the previous comments. I keep using the Filter since there is no need to use Partition in the Lc task for the moment.

[vkucera]

Thanks @zhangbiao-phy , did you test the changes locally and can you confirm that you are getting the same results now as before #5040 ?

[alibuild]

Error while checking build/O2Physics/o2 for cade64e at 2024-03-19 11:42:

## sw/BUILD/O2Physics-latest/log
/sw/SOURCES/O2Physics/5232-slc7_x86-64/0/PWGHF/D2H/Tasks/taskLc.cxx:667:21: error: 'selectedLcCandidatesMcMl' was not declared in this scope; did you mean 'selectedLcCandidatesMlMc'?
ninja: build stopped: subcommand failed.

Full log here.

[zhangbiao-phy]

Thanks @zhangbiao-phy , did you test the changes locally and can you confirm that you are getting the same results now as before #5040 ?
@vkucera, yes, I have done the test. I have many more candidates from the #5040 w.r.t the keep using the "Filter". Then the candidates look reasonable now. Thanks for spotting this issue!

[vkucera]

Thanks @zhangbiao-phy , did you test the changes locally and can you confirm that you are getting the same results now as before #5040 ?
@vkucera, yes, I have done the test. I have many more candidates from the #5040 w.r.t the keep using the "Filter". Then the candidates look reasonable now. Thanks for spotting this issue!

Thanks @zhangbiao-phy for checking. However, this is not what I asked. Sorry if I was not clear. I want you to compare the histograms produced with your PR branch and histograms produced with master before #5040 was merged.
We need to validate this fix by confirming that we get the same results as before the bug was introduced.

[zhangbiao-phy]

Thanks @zhangbiao-phy , did you test the changes locally and can you confirm that you are getting the same results now as before #5040 ?
@vkucera, yes, I have done the test. I have many more candidates from the #5040 w.r.t the keep using the "Filter". Then the candidates look reasonable now. Thanks for spotting this issue!

Thanks @zhangbiao-phy for checking. However, this is not what I asked. Sorry if I was not clear. I want you to compare the histograms produced with your PR branch and histograms produced with master before #5040 was merged. We need to validate this fix by confirming that we get the same results as before the bug was introduced.

Hi @vkucera, ah, OK, then I finished this check. I get the same output before and after for the case without ML.

[vkucera]

Thanks @zhangbiao-phy , did you test the changes locally and can you confirm that you are getting the same results now as before #5040 ?
@vkucera, yes, I have done the test. I have many more candidates from the #5040 w.r.t the keep using the "Filter". Then the candidates look reasonable now. Thanks for spotting this issue!

Thanks @zhangbiao-phy for checking. However, this is not what I asked. Sorry if I was not clear. I want you to compare the histograms produced with your PR branch and histograms produced with master before #5040 was merged. We need to validate this fix by confirming that we get the same results as before the bug was introduced.

Hi @vkucera, ah, OK, then I finished this check. I get the same output before and after for the case without ML.

Great, thanks @zhangbiao-phy