Plotter by HJafree · Pull Request #40 · Mu2e/CaloCalibration

HJafree · 2026-02-04T18:28:34Z

Updated code logic to re-try fits for failing sipms, added cut and count analysis for mc truth, created a separate file for 2d contour and updated the readme

…s and made a seperate doc for 2d contour plots

sophiemiddleton · 2026-02-04T21:52:55Z

Huma, I think this looks good. I think more documentation and comments would help to understand what each block of code is doing.

Perhaps we can encapsulate some of the functions into sub-classes or at most other function to avoid large blocks.

This is just style preference though - I can see a lot of great work here :)

Copilot

Pull request overview

This PR implements enhancements to the Mu2e Source Calibration Analysis framework, focusing on three main areas: MC truth analysis for validation, an intelligent fitting retry mechanism to handle convergence failures, and 2D likelihood contour visualization for parameter correlation studies.

Changes:

Added MC truth cut-and-count analysis module (mcinfo.cc) to extract peak fractions from Monte Carlo truth histograms
Implemented adaptive retry logic in the fitter with three levels: initial fit, covariance-guided refit, and randomized parameter search
Created standalone 2D contour plotting module (2dcontour.cc) for visualizing parameter correlations and fit quality
Extended plotter to support asymmetric errors, LYSO/CsI crystal separation, and additional diagnostic plots
Updated main executable to support MC mode and optional contour generation via command-line flags

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated 17 comments.

Show a summary per file

File	Description
SourceCalib/src/mcinfo.cc	New file implementing MC truth peak extraction and summary statistics
SourceCalib/src/SourceFitter.cc	Major refactor: added retry logic, asymmetric errors, improved error handling, message suppression
SourceCalib/src/2dcontour.cc	New file implementing 2D likelihood/chi2 contour scanning with auto-ranging
SourceCalib/src/SourcePlotter.cc	Enhanced plots: asymmetric error bars, CsI/LYSO splitting, odd/even comparisons, 45° projections
SourceCalib/src/MakeAnalysisTree_main.cc	Added MC mode block, improved flag parsing, convergence summary reporting
SourceCalib/inc/mcinfo.hh	Header for MC truth analysis class
SourceCalib/inc/SourceFitter.hh	Added static members for tracking retry statistics
SourceCalib/inc/SourcePlotter.hh	Added includes for new plot types
SourceCalib/inc/2dcontour.hh	Header for contour plotting function
SourceCalib/src/classes.h	Added includes for new modules
SourceCalib/src/classes_def.xml	Registered mcinfo class
SourceCalib/README.md	Comprehensive documentation update with usage examples

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-18T23:14:25Z

+//add to counter to calculae avg of all values
+covarAcc.count++;
+covarAcc.sumPeak     += fpeak;
+covarAcc.sumWidth    += fsigma;
+covarAcc.sumAlpha    += fcbalphaparam;
+covarAcc.sumBeta     += combetaparam;
+covarAcc.sumEvtFull  += fr_fullparam*nEvents;
+covarAcc.sumEvtFst   += fr_frstparam*nEvents;
+covarAcc.sumEvtScd   += fr_scndparam*nEvents;
+covarAcc.sumEvtBkg   += fr_bkgparam*nEvents;


The covariance accumulator is updated using uninitialized variables. Lines 224-231 use fpeak, fsigma, fcbalphaparam, combetaparam, fr_fullparam, fr_frstparam, fr_scndparam, and fr_bkgparam before they are assigned values. These parameters are only set much later (lines 355-364), after the fit has completed and the plot has been drawn. This means the accumulator is collecting garbage values on the first fit, which will corrupt the average values used for retry fits.

Suggested change

//add to counter to calculae avg of all values

covarAcc.count++;

covarAcc.sumPeak += fpeak;

covarAcc.sumWidth += fsigma;

covarAcc.sumAlpha += fcbalphaparam;

covarAcc.sumBeta += combetaparam;

covarAcc.sumEvtFull += fr_fullparam*nEvents;

covarAcc.sumEvtFst += fr_frstparam*nEvents;

covarAcc.sumEvtScd += fr_scndparam*nEvents;

covarAcc.sumEvtBkg += fr_bkgparam*nEvents;

//add to counter to calculate avg of all values

covarAcc.count++;

covarAcc.sumPeak += fpeak;

covarAcc.sumWidth += fsigma;

// For parameters that are not initialized in run_one_fit(), avoid

// reading them on the very first fit to prevent using undefined values.

if (covarAcc.count > 1) {

covarAcc.sumAlpha += fcbalphaparam;

covarAcc.sumBeta += combetaparam;

covarAcc.sumEvtFull += fr_fullparam*nEvents;

covarAcc.sumEvtFst += fr_frstparam*nEvents;

covarAcc.sumEvtScd += fr_scndparam*nEvents;

covarAcc.sumEvtBkg += fr_bkgparam*nEvents;

}

Copilot · 2026-02-18T23:14:26Z

+    truthtable->cd();
+    trueinfo->Write(); // Write the tree explicitly
+    truthtable->Write(); // Write any histograms created by FinalizeMCSummary
+    truthtable->Close();


Memory leak: The truthtable TFile pointer created on line 109 is closed but never deleted. Add delete truthtable; after truthtable->Close(); on line 163.

Suggested change

truthtable->Close();

truthtable->Close();

delete truthtable;

Copilot · 2026-02-18T23:14:27Z

+**1. Standard Analysis (Chi2 Fit)**
+Accumulate and fit events for crystals 0 to 674 on disk 0 using Chi-Square minimization:
+
+./build/al9-prof-e28-p056/CaloCalibration/bin/MakeAnalysisTree 0 674 "chi2" 0


The example command on line 88 is missing backticks for proper markdown code formatting, unlike the other examples. It should be wrapped in triple backticks like the other examples for consistency.

Copilot · 2026-02-18T23:14:27Z

+  // 1) First attempt: start from init values 
+  // -----------------------
+bool first_ok = run_one_fit(); //run with init guesses
+//add to counter to calculae avg of all values


Typo in comment: "calculae" should be "calculate".

Suggested change

//add to counter to calculae avg of all values

//add to counter to calculate avg of all values

Copilot · 2026-02-18T23:14:29Z

+  errbarhigh = mparam*(peakerrorhigh/fpeak);
+  errbarlo = mparam*(peakerrorlo/fpeak);


Potential division by zero: If fpeak is zero (which could happen in pathological fit failures), lines 369-370 will divide by zero when computing errbarhigh and errbarlo. Add a check to ensure fpeak != 0 before performing these divisions.

Suggested change

errbarhigh = mparam*(peakerrorhigh/fpeak);

errbarlo = mparam*(peakerrorlo/fpeak);

if (fpeak != 0.0) {

errbarhigh = mparam * (peakerrorhigh / fpeak);

errbarlo = mparam * (peakerrorlo / fpeak);

} else {

// Pathological case: peak height is zero; cannot compute scaled errors.

errbarhigh = 0.0;

errbarlo = 0.0;

}

Copilot · 2026-02-18T23:14:29Z

+    // Width of the first bin (in keV if the axis is in keV)
+	double binWidth = h_spec->GetXaxis()->GetBinWidth(500);  
+	std::cout << "Bin width = " << binWidth << " keV" << std::endl; 
+//main function that assignes the values to the params


Typo in comment: "assignes" should be "assigns".

Suggested change

//main function that assignes the values to the params

//main function that assigns the values to the params

Copilot · 2026-02-18T23:14:29Z

+        // Lambda: randomize all parameters
+        auto randomize_all_parameters = [&]() {
+            auto randomDouble = [](double min, double max) {
+                return min + (max - min) * ((double)rand() / RAND_MAX);


Using rand() for random number generation is not recommended for modern C++. The header <random> is included but not used. Consider using std::mt19937 and std::uniform_real_distribution for better randomness and thread safety, which is especially important since this code may be run in parallel contexts.

Suggested change

return min + (max - min) * ((double)rand() / RAND_MAX);

static thread_local std::mt19937 gen{std::random_device{}()};

std::uniform_real_distribution<double> dist(min, max);

return dist(gen);

Copilot · 2026-02-18T23:14:30Z

+    std::cout << " peak std dev " << Peaks_stddev << std::endl;
+    std::cout << " ------------ " << std::endl;
+
 //----create loop to count and print how many crys/sipms have pear error larger than 2 stdev----		


Typo in comment: "pear" should be "peak".

Suggested change

//----create loop to count and print how many crys/sipms have pear error larger than 2 stdev----

//----create loop to count and print how many crys/sipms have peak error larger than 2 stdev----

Copilot · 2026-02-18T23:14:30Z

+}();
+void SourceFitter::FitCrystal(TH1F* h_spec, TString opt, int crystalNo,  TTree *covar, Int_t &nEvents,Int_t &convergencestatus, Float_t &fpeak, Float_t &peakerrorhigh,Float_t &peakerrorlo, Float_t &fsigma,Float_t &widtherrorhigh,Float_t &widtherrorlo, Float_t &chiSq, Float_t &fstpeak, Float_t &scdpeak,Float_t &fcbalphaparam,Float_t &fcbndegparam,Float_t &comCnstparam, Float_t &combetaparam, Float_t &fr_fullparam, Float_t &fr_frstparam,Float_t &fr_scndparam,Float_t &crystalNoparam,Float_t &fr_bkgparam, Float_t &pval,Float_t &h_means,Float_t &h_stddevs, Float_t &unreducedchi2,Float_t &fval,Float_t &mparam,Float_t &etaparam,Int_t &ndof,bool contour, Float_t &errbarhigh, Float_t &errbarlo,Float_t &evtfullerrorhigh,Float_t &evtfullerrorlo,Float_t &eventsFull,Float_t &Esparam ){
+
  // set stlye optionsr


Typo in comment: "stlye optionsr" should be "style options".

Suggested change

// set stlye optionsr

// set style options

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

…input values from command line

…tter

Huma Jafree and others added 2 commits November 11, 2025 13:41

added new features and refitting scheme

c57016b

New logic update to re-try fits, added mc truth cut and count analysi…

142eb88

…s and made a seperate doc for 2d contour plots

sophiemiddleton requested a review from Copilot February 18, 2026 23:08

Copilot started reviewing on behalf of sophiemiddleton February 18, 2026 23:09 View session

Copilot AI reviewed Feb 18, 2026

View reviewed changes

Huma Jafree and others added 6 commits March 2, 2026 12:57

Merge remote-tracking branch 'origin/main' into plotter

d51a7c4

Update SourceCalib/src/mcinfo.cc

08daa87

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

updated chi2var to remove deprecated func + updated 2dcountour.cc to …

c88ac65

…input values from command line

added new conditions to pass filters

96a0ddf

Merge branch 'plotter' of github.com:HJafree/CaloCalibration into plo…

a4110ae

…tter

fixed bugs and cleaned file

d747dea

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Plotter#40

Plotter#40
HJafree wants to merge 8 commits intoMu2e:mainfrom
HJafree:plotter

HJafree commented Feb 4, 2026

Uh oh!

sophiemiddleton commented Feb 4, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Copilot AI Feb 18, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

	//add to counter to calculae avg of all values
	//add to counter to calculate avg of all values

		errbarhigh = mparam*(peakerrorhigh/fpeak);
		errbarlo = mparam*(peakerrorlo/fpeak);

-  errbarhigh = mparam*(peakerrorhigh/fpeak);
-  errbarlo = mparam*(peakerrorlo/fpeak);
+  if (fpeak != 0.0) {
+    errbarhigh = mparam * (peakerrorhigh / fpeak);
+    errbarlo   = mparam * (peakerrorlo   / fpeak);
+  } else {
+    // Pathological case: peak height is zero; cannot compute scaled errors.
+    errbarhigh = 0.0;
+    errbarlo   = 0.0;
+  }

	//main function that assignes the values to the params
	//main function that assigns the values to the params

-                return min + (max - min) * ((double)rand() / RAND_MAX);
+                static thread_local std::mt19937 gen{std::random_device{}()};
+                std::uniform_real_distribution<double> dist(min, max);
+                return dist(gen);

	//----create loop to count and print how many crys/sipms have pear error larger than 2 stdev----
	//----create loop to count and print how many crys/sipms have peak error larger than 2 stdev----

Conversation

HJafree commented Feb 4, 2026

Uh oh!

sophiemiddleton commented Feb 4, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 18, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants