-
Notifications
You must be signed in to change notification settings - Fork 613
[PWGHF] Centrality changes in Lc correlator, Correlation task #11724
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
|
O2 linter results: ❌ 0 errors, |
|
Hi @zeptotera , please notice the automatic formatting PR. You don't have to run clang-format at all. There is also the option to use pre-commit hooks. Both options are documented. |
|
Please fix the errors. |
fgrosa
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @zeptotera, thanks for the implementation! See my comments below
867bb03 to
8858bd6
Compare
|
@zeptotera Please do not mark your PR as ready for review if it has conflicts or errors. |
2c95540 to
0b5e797
Compare
| if (isMultiplicityDependent) { | ||
| registry.add("hMassLcVsPtvsMult", "Lc candidates;inv. mass (p K #pi) (GeV/#it{c}^{2});entries", {HistType::kTH3F, {{axisMassLc}, {axisPtLc}, {axisCent}}}); | ||
| } else { | ||
| registry.add("hMassLcVsPt", "Lc candidates;inv. mass (p K #pi) (GeV/#it{c}^{2});entries", {HistType::kTH2F, {{axisMassLc}, {axisPtLc}}}); | ||
| } |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't see any relation to multiplicity.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why do you need hMassLcVsPt?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Okay, We have changed the name of the histogram to have hMassLcVsPtvsCent
hMassLcVsPt is filled so that other people who want the histogram without centrality can still access it and the merge creates a safe working environment for existing Analyzers, This was added based on Antonio’s suggestion.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What is unsafe about projecting hMassLcVsPtvsCent along the centrality axis?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Projecting along the centrality axis works fine for us too. We only kept the separate hMassLcVsPt upon Antonio’s suggestion, so that others using the same task without centrality wouldn’t need to change anything. If you think it's better to just keep one and project, we’re happy to update it — please suggest.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I suggest a single histogram.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @vkucera I am not sure what is the recommended process in HF when adding new things to an exciting analysis task that is being used by others. During the discussion with coordinators, we agreed that while extending this task to perform a multiplicity based study, Tushar adds flags and additional histograms in a way that has minimal impact on current analyzers. Changing a 2D histogram to a 3D will disrupt others, which we would like to minimize. If you still insist on changing it, I would appreciate if you could share a reasoning to do it as well.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @deepathoms @apalasciano , here is how I see the two scenarios:
- split conditional histograms
- pros
- no change for current users
- cons
- bigger code
- more code branching at many places -> less maintainable, more error-prone
- lot of code duplication -> less maintainable, more error-prone
- analysis-dependent output
- train duplication for different analyses
- pros
- no split histograms
- pros
- smaller code -> more maintainable, safer
- no branching nor duplication -> more maintainable, safer
- Same output can be used for all analyses.
- cons
- changing histogram name in the postprocessing code?
- pros
IMO, the drawbacks of the split scenario and the benefits of the no-split scenario are far more important than protecting users from changing a histogram name in their local scripts. One can even keep the current histogram name.
Concerning the projection, I don't see how it changes anything for the users. You said
Changing a 2D histogram to a 3D will disrupt others
but I don't understand what you mean here.
- There is no change of a 2D histogram into a 3D histogram. There is a change from a
kTHnSparseFhistogram withNaxes to akTHnSparseFhistogram withN + 1axes. - The signature of the
THnSparse::Projectionmethod does not depend on the number of object axes.
In general, I don't agree that keeping the framework interface and the output intact should be a top priority. The framework is evolving based on the analysis needs and we should be able to collaborate on the evolution of the common code, not fragment it and protect it from changes.
We have introduced breaking changes even in the core code many times before. We communicated them to the community in an announcement and explained the consequences. It's just about communication.
| if (isMultiplicityDependent) { | ||
| registry.add("hCorrel2DVsPtSidebandsWithCentrality", stringLcHadron + stringSideband + stringDeltaPhi + stringDeltaEta + stringPtLc + stringPtHadron + "entries", {HistType::kTHnSparseF, {{axisDeltaPhi}, {axisDeltaEta}, {axisPtCorr}, {axisPtHadron}, {axisPoolBin}, {axisCentFT0M}}}); | ||
| } else { | ||
| registry.add("hCorrel2DVsPtSidebands", stringLcHadron + stringSideband + stringDeltaPhi + stringDeltaEta + stringPtLc + stringPtHadron + "entries", {HistType::kTHnSparseF, {{axisDeltaPhi}, {axisDeltaEta}, {axisPtCorr}, {axisPtHadron}, {axisPoolBin}}}); | ||
| } |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Same comment about two histograms
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
hCorrel2DvsPtSidebands is filled so that other people who want the histogram without centrality can still access it and the merge creates a safe working environment for existing Analyzers, This was added based on Antonio’s suggestion.
|
Hi @fcatalan92 @fcolamar @gluparel @deepathoms @mmazzilli @mfaggin @hahassan7 @jpxrk @apalasciano @zhangbiao-phy @NicoleBastid @vkucera I have addressed and implemented all the requested changes in the PR. A few of the questions raised seem to be beyond the scope of this PR. We would appreciate it if the PR could be approved at the earliest so we can proceed further. Thanks! |
|
If you wish your PR to be reviewed, please use the button "Ready for review". |
fgrosa
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok for what concerns my comments!
|
Hi @fcatalan92 @fcolamar @gluparel @deepathoms @mmazzilli @mfaggin @hahassan7 @jpxrk @apalasciano @zhangbiao-phy @NicoleBastid @vkucera, We’ve updated the tasks according to Vit’s suggestions. Could you please review and approve the PR? |
vkucera
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @zeptotera , thanks for removing the split histograms. Please see my comments on a few remaining issues.
|
Hi @vkucera, We have made changes according to your latest suggestions, Can you please review and Approve the PR? |
vkucera
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks @zeptotera for addressing my comments.
Updated the LcHadronPair table to include centrality