Skip to content
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

Implement TH2Poly in DQM Services for HGCal DQM #41932

Open
wants to merge 10 commits into
base: master
Choose a base branch
from

Conversation

ywkao
Copy link

@ywkao ywkao commented Jun 12, 2023

PR description:

This PR introduces a new type of DQM MonitorElement, TH2Poly, for HGCal DQM in the future. This feature allows a display of polygonal histograms on the CMS DQM GUI. As a demonstration, a wafer map can be displayed like the screenshot here [1].

TH2Poly is a 2D histogram class inherited from TH2. Polygonal bins, defined by TGraph, can be loaded using the AddBin() method. After setting up the polygonal bins, a TH2Poly object can store information through Fill() or SetBinContent().

A workflow for creating polygonal histograms looks like this:
DQM Service -> DQM EDAnalyzer -> CMS DQM GUI

An implementation of TH2Poly in DQM Service and MonitorElement is necessary to display the polygonal histograms. It involves updates on two repositories: cmssw and dqmgui_prod. A pull request is created in the dqmgui_prod repository [2] with a relevant issue reported in this link [3], which is about setting up a CMS DQM GUI with the new feature.

PR validation:

The workflow and the implementation have been tested: (a) From this feature branch, monitor elements of TH2Poly can be stored in a DQM root file [4]. (b) The DQM root file can be uploaded to a CMS DQM GUI, which is built following the steps noted in this issue [3]. Polygonal maps can be displayed on the DQM GUI, as demonstrated in [1].

[1] https://ykao.web.cern.ch/ykao/raw_data_handling/hgcal_dqm_gui/screenshot_demo_th2poly_wafermap.png
[2] cms-DQM/dqmgui_prod#14
[3] cms-DQM/dqmgui_prod#13
[4] A DQM root file containing demo polygonal maps: /afs/cern.ch/work/y/ykao/public/example_HGCAL_DQM/DQM_V0001_HGCAL_R000123469.root

@cmsbuild
Copy link
Contributor

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-41932/35885

  • This PR adds an extra 44KB to repository

@cmsbuild
Copy link
Contributor

cmsbuild commented Jun 12, 2023

A new Pull Request was created by @ywkao for master.

It involves the following packages:

  • DQMServices/Core (dqm)
  • DataFormats/Histograms (dqm, core)

@smuzaffar, @Dr15Jones, @makortel, @nothingface0, @emanueleusai, @cmsbuild, @pmandrik, @syuvivida, @tjavaid, @micsucmed, @rvenditti can you please review it and eventually sign? Thanks.
@missirol, @barvic, @rovere this is something you requested to watch as well.
@perrotta, @dpiparo, @rappoccio you are the release manager for this.

cms-bot commands are listed here

Copy link
Contributor

@makortel makortel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think in addition the following places need to be amended accordingly

static TreeHelperBase* makeHelper(unsigned int iTypeIndex, TTree* iTree, std::string* iFullNameBufferPtr) {
switch (iTypeIndex) {
case kIntIndex:
return new IntTreeHelper(iTree, iFullNameBufferPtr);

m_treeReaders[kTProfileIndex].reset(new TreeObjectReader<TProfile>(MonitorElementData::Kind::TPROFILE, m_rescope));
m_treeReaders[kTProfile2DIndex].reset(
new TreeObjectReader<TProfile2D>(MonitorElementData::Kind::TPROFILE2D, m_rescope));
}

How about adding some tests to make sure all the I/O components work properly? (I don't remember if there are already any such tests that could be easily amended)

double lowY,
double highY,
FUNC onbooking = NOOP()) {
return bookME(name, MonitorElementData::Kind::TH2F, [=]() {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Shouldn't this be

Suggested change
return bookME(name, MonitorElementData::Kind::TH2F, [=]() {
return bookME(name, MonitorElementData::Kind::TH2Poly, [=]() {

?

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Indeed, this should be corrected. I will update it in a commit accordingly.

/// set polygon bin (TH2Poly)
void MonitorElement::addBin(TGraph *graph) {
auto access = this->accessMut();
static_cast<TH2Poly *>(accessRootObject(access, __PRETTY_FUNCTION__, 2))->AddBin(graph);
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What if the ROOT object is not TH2Poly? The other similar functions throw an exception via the incompatible() function.

Copy link
Author

@ywkao ywkao Jun 13, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Indeed, incompatible() function is necessary. This block of code will be modified as follows:

-    static_cast<TH2Poly *>(accessRootObject(access, __PRETTY_FUNCTION__, 2))->AddBin(graph);
+    if (kind() == Kind::TH2Poly) {
+      static_cast<TH2Poly *>(accessRootObject(access, __PRETTY_FUNCTION__, 2))->AddBin(graph);
+    } else {
+      incompatible(__PRETTY_FUNCTION__);
+    }

@@ -140,7 +140,8 @@ struct MonitorElementData {
TH2I = 0x23,
TH3F = 0x30,
TPROFILE = 0x40,
TPROFILE2D = 0x41
TPROFILE2D = 0x41,
TH2Poly = 0x60
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why such a large gap between the previous value (0x41)?

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I guess I kept thinking of hexagons for a wafer map and then put a number starting with six. We can assign 0x24 for TH2Poly if it is feasible.

+    TH2Poly = 0x24,
     TH3F = 0x30,
     TPROFILE = 0x40,
-    TPROFILE2D = 0x41,
-    TH2Poly = 0x60
+    TPROFILE2D = 0x41

@emanueleusai
Copy link
Member

type hgcal

@emanueleusai
Copy link
Member

please test

@cmsbuild
Copy link
Contributor

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-03d614/33112/summary.html
COMMIT: 5fa5e4e
CMSSW: CMSSW_13_2_X_2023-06-12-2300/el8_amd64_gcc11
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/41932/33112/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

  • You potentially added 15 lines to the logs
  • Reco comparison results: 59 differences found in the comparisons
  • DQMHistoTests: Total files compared: 48
  • DQMHistoTests: Total histograms compared: 3190910
  • DQMHistoTests: Total failures: 2477
  • DQMHistoTests: Total nulls: 0
  • DQMHistoTests: Total successes: 3188411
  • DQMHistoTests: Total skipped: 22
  • DQMHistoTests: Total Missing objects: 0
  • DQMHistoSizes: Histogram memory added: 0.0 KiB( 47 files compared)
  • Checked 207 log files, 159 edm output root files, 48 DQM output files
  • TriggerResults: no differences found

@cmsbuild
Copy link
Contributor

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-41932/35914

  • This PR adds an extra 40KB to repository

@cmsbuild
Copy link
Contributor

Pull request #41932 was updated. @smuzaffar, @Dr15Jones, @makortel, @nothingface0, @emanueleusai, @cmsbuild, @pmandrik, @syuvivida, @tjavaid, @micsucmed, @rvenditti can you please check and sign again.

@ywkao
Copy link
Author

ywkao commented Jun 13, 2023

I think in addition the following places need to be amended accordingly

static TreeHelperBase* makeHelper(unsigned int iTypeIndex, TTree* iTree, std::string* iFullNameBufferPtr) {
switch (iTypeIndex) {
case kIntIndex:
return new IntTreeHelper(iTree, iFullNameBufferPtr);

m_treeReaders[kTProfileIndex].reset(new TreeObjectReader<TProfile>(MonitorElementData::Kind::TPROFILE, m_rescope));
m_treeReaders[kTProfile2DIndex].reset(
new TreeObjectReader<TProfile2D>(MonitorElementData::Kind::TPROFILE2D, m_rescope));
}

How about adding some tests to make sure all the I/O components work properly? (I don't remember if there are already any such tests that could be easily amended)

Okay, I will modify the lines and test it. Would PR #37665 be a good reference for the implementation of TH2Poly?

Also, I would appreciate any instructions from experts on how to test all the I/O components.

@makortel
Copy link
Contributor

Would PR #37665 be a good reference for the implementation of TH2Poly?

That PR indeeds looks like a good example.

@IzaakWN
Copy link
Contributor

IzaakWN commented Dec 9, 2024

Hi @mmusich, @makortel, is there anything else needed from this PR? It would be very convenient for testing PR #45665 along with other HGCal tools if this PR can still make it into the next (pre-)release (CMSSW_15_0_0_pre1?). (Apologies for pushing.)

@mmusich
Copy link
Contributor

mmusich commented Dec 9, 2024

is there anything else needed from this PR?

it doesn't depend on me :)
@cms-sw/core-l2 and @cms-sw/dqm-l2 should review the PR and sign it if they are happy.
It looks like you might want to address the GUI PR cms-DQM/dqmgui_prod#14 as well, before pushing this in production.

// curr->update(it->getNbinsX() * it->getNbinsY(),
// getEmptyMetric(it->getTH2Poly()->GetArray(), it->getNbinsX() + 2, it->getNbinsY() + 2, 0),
// it->getNbinsX() * it->getNbinsY() * sizeof(int));
// break;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What about this code? Is something needed here? (I have no idea what exactly is being done in this function in general)

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I am not sure all the details either. I referred to PR#37665-files, the implementation of DQM support of TH1I and TH2I.

@mmusich could you help evaluate the necessity of having TH2Poly::GetArray() functionality? Maybe we need help from ROOT experts.

// double value2 = newMEtoEDMObject[i].object.GetBinContent(ibin);
// double total = value1 + value2;
// MEtoEdmObject[j].object.SetBinContent(ibin, total);
// }
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Does uncommenting of this block not compile? If that is the case (i.e. that motivates the specialization of the entire mergeProduct() function), I think we need to think a bit for a solution that wouldn't require duplicating the entire logic of the mergeProduct() function.

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, uncommenting this block of code can now compile without an error if a dummy structure in DataFormats/Histograms/test/metoedmformat_t.cc is properly handled:

+ void Add(const Dummy* iOther, float weight=1.) { m_i += iOther->m_i; }
+ int GetNcells() const { return 9; }
+ float GetBinContent(int bin) const { return -1.; }
+ void SetBinContent(int bin, float value) const { return; }
+ const TClass* IsA() const { return 0; }

We can also remove the ad-hoc mergeProduct() for TH2Poly and have minimal modification on L151:

-          MEtoEdmObject[j].object.Add(&newMEtoEDMObject[i].object);
+          MEtoEdmObject[j].object.Add(&newMEtoEDMObject[i].object, 1.);

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

After the modification of L151, I found that removing this ad-hoc TH2Poly addition code can also pass compilation and unit tests in DQMServices/Demo/test/runtests.sh. To ensure the complete functionality, I plan to:

  • Add polygonal bins test cases in DQMServices/Demo/test/TestDQMEDAnalyzer.cc.
  • Produce a few DQM root files for verification.

@antoniovagnerini
Copy link

is there anything else needed from this PR?

it doesn't depend on me :) @cms-sw/core-l2 and @cms-sw/dqm-l2 should review the PR and sign it if they are happy. It looks like you might want to address the GUI PR cms-DQM/dqmgui_prod#14 as well, before pushing this in production.

hi, since this PR touches the DQMStore which is the basic class that the clients use to manipulate/fill Histograms, we need to test it in our playback system DQM machines. The latter is unfortunately down for the programmed YETS, so I am afraid we will have to postpone this batch of test to the new year.

@smuzaffar
Copy link
Contributor

code-checks

@cmsbuild
Copy link
Contributor

-code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-41932/43100

Code check has found code style and quality issues which could be resolved by applying following patch(s)

@cmsbuild
Copy link
Contributor

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-41932/43111

@cmsbuild
Copy link
Contributor

Pull request #41932 was updated. @Dr15Jones, @antoniovagnerini, @cmsbuild, @makortel, @rseidita, @smuzaffar can you please check and sign again.

@antoniovagnerini
Copy link

please test

@cmsbuild
Copy link
Contributor

+1

Size: This PR adds an extra 20KB to repository
Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-03d614/43585/summary.html
COMMIT: 07470c1
CMSSW: CMSSW_15_0_X_2024-12-19-2300/el8_amd64_gcc12
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/41932/43585/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

  • You potentially removed 196 lines from the logs
  • Reco comparison results: 11 differences found in the comparisons
  • DQMHistoTests: Total files compared: 49
  • DQMHistoTests: Total histograms compared: 3818210
  • DQMHistoTests: Total failures: 444
  • DQMHistoTests: Total nulls: 0
  • DQMHistoTests: Total successes: 3817746
  • DQMHistoTests: Total skipped: 20
  • DQMHistoTests: Total Missing objects: 0
  • DQMHistoSizes: Histogram memory added: 0.0 KiB( 48 files compared)
  • Checked 214 log files, 184 edm output root files, 49 DQM output files
  • TriggerResults: found differences in 1 / 47 workflows

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
Status: Work in CMS
Development

Successfully merging this pull request may close these issues.