improve friendliness of histogram pool (#1854)

* WIP: start refactor, not compiling

* rename file and get compiling

To Do
- check that changning how bin edges are passed doesn't seg fault
- add more creation helpers
- check if we are in a subdirectory? share Process handle so we can open
histogram directory for the user upon creation?

* pass callback for directory creation to pool, start tests

* more docs and updated tests

* WIP: attempt to get categorical histograms working

* pool with categorical axes working

- need to add more creation options
- more tests of those different histogram axes
- documentation
- tests on weights

* add ability to store sumw2 when creating

* make sure weighting and 2-D categories works

* start some more documentation on histograms

* clang-tidy the histogram pool developments

* Apply clang-format

* add categorical-axis support to python config histograms

- move interpretatin of pyton config parameters into HistogramPool
- add ninth and final 2D create method
- add x_label to 1D categorical histograms back for uniformity

* Apply clang-tidy

* add documentation to insert, avoid single-char variable names

* Apply clang-format

* convert category histograms in DQM to new interface

* undo accidental clang-format

* fixup category bin indexing

* pass weighted through python config as well

* more docs on example usage and create methods

* Apply clang-format

* add warning about prior naming scheme

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

Author: tomeichlersmith

DQM/include/DQM/DarkBremInteraction.h

@@ -32,14 +32,6 @@ class DarkBremInteraction : public framework::Producer {
  public:
   DarkBremInteraction(const std::string& n, framework::Process& p)
       : framework::Producer(n, p) {}
-  /**
-   * update the labels of some categorial histograms
-   *
-   * This is helpful for downstream viewers of the histograms
-   * so that ROOT will display the bins properly.
-   */
-  virtual void onProcessStart() override;
-
   /**
    * extract the kinematics of the dark brem interaction from the SimParticles
    *
@@ -58,12 +50,6 @@ class DarkBremInteraction : public framework::Producer {
   void configure(framework::config::Parameters& parameters) override;
 
  private:
-  /**
-   * Set the labels of the histogram of the input name with the input labels
-   */
-  void setHistLabels(const std::string& name,
-                     const std::vector<std::string>& labels);
-
   /**
    * the list of known materials assigning them to material ID numbers
    *

DQM/include/DQM/EcalClusterAnalyzer.h

@@ -34,9 +34,6 @@ class EcalClusterAnalyzer : public framework::Analyzer {
   ~EcalClusterAnalyzer() override = default;
   void configure(framework::config::Parameters& ps) override;
   void analyze(const framework::Event& event) override;
-  void onProcessStart() override;
-  void setHistLabels(const std::string& name,
-                     const std::vector<std::string>& labels);
 
  private:
   /// Use the number of simulated electrons

DQM/include/DQM/HcalVetoResults.h

@@ -32,11 +32,6 @@ class HcalVetoResults : public framework::Analyzer {
    */
   virtual void configure(framework::config::Parameters& ps);
 
-  /**
-   * Things to do only on process start
-   */
-  virtual void onProcessStart();
-
   /**
    * Fills histograms
    */

DQM/include/DQM/PhotoNuclearDQM.h

@@ -75,15 +75,6 @@ class PhotoNuclearDQM : public framework::Analyzer {
    */
   void analyze(const framework::Event &event) override;
 
-  /// @brief  Helper function to label categorical histos
-  /// @param name : Name of the histo for the labels to set
-  /// @param labels : Labels on the X axis
-  void setHistLabels(const std::string &name,
-                     const std::vector<std::string> &labels);
-
-  /// Method executed before processing of events begins.
-  void onProcessStart() override;
-
  private:
   /** Method used to classify events. Note: Assumes that daughters is sorted by
    * kinetic energy. */

DQM/include/DQM/SampleValidation.h

@@ -19,8 +19,6 @@ class SampleValidation : public framework::Analyzer {
   virtual void configure(framework::config::Parameters& ps) override;
   virtual void analyze(const framework::Event& event) override;
   int pdgidLabel(const int pdgid);
-  /// Method executed before processing of events begins.
-  void onProcessStart() override;
 
  private:
   std::string target_scoring_plane_passname_;

DQM/include/DQM/TrkDeDxMassEstFeatures.h

@@ -31,9 +31,6 @@ class TrkDeDxMassEstFeatures : public framework::Analyzer {
    */
   virtual void analyze(const framework::Event& event) override;
 
-  /// Method executed before processing of events begins.
-  void onProcessStart() override;
-
  private:
   /// Collection Name for mass estimate object
   std::string mass_estimate_name_;

DQM/python/dqm.py

@@ -200,7 +200,8 @@ def __init__(self,name="HcalVetoResults") :
         self.build1DHistogram('num_valid_hits',
                 'Total number of valid HCAL hits', 500, -0.5, 499.5)
         self.build1DHistogram('max_section',
-                'Maximal PE hit section', 5, -0.5, 4.5)
+                'Maximal PE hit section',
+                ['Back', 'Top', 'Bottom', 'Right', 'Left'])
         self.build1DHistogram('max_pos_z',
                 'Maximal PE hit postion Z [mm]', 6000, 200., 6200)
         self.build1DHistogram('veto_pass',
@@ -646,10 +647,11 @@ def __init__(self) :
         # elements are hydrogen and carbon (for trigger pads) and tungsten target
         self.build1DHistogram('dark_brem_element',
             'Element in which Dark Brem Occurred',
-            12, 0, 12)
+            ["did not happen", "H 1", "C 6", "O 8", "Na 11", "Si 14",
+             "Ca 20", "Cu 29", "Y 39", "Lu 71", "W 74", "unlisted"])
         self.build1DHistogram('dark_brem_material',
             'Material in which Dark Brem Occurred',
-            8, 0, 8)
+            ["Unknown", "C", "PCB", "Glue", "Si", "Al", "W / LYSO", "PVT"])
 
 
 class HCalRawDigi(ldmxcfg.Analyzer) :
@@ -736,17 +738,56 @@ def __init__(self,name='PN', count_light_ions=True) :
 
         self.sim_particles_passname = ''
 
+        event_type_labels = [
+            "Nothing hard",   # 0
+            "1 n",            # 1
+            "2 n",            # 2
+            "#geq 3 n",       # 3
+            "1 #pi",          # 4
+            "2 #pi",          # 5
+            "1 #pi_{0}",      # 6
+            "1 #pi A",        # 7
+            "1 #pi 2 A",      # 8
+            "2 #pi A",        # 9
+            "1 #pi_{0} A",    # 10
+            "1 #pi_{0} 2 A",  # 11
+            "#pi_{0} #pi A",  # 12
+            "1 p",            # 13
+            "2 p",            # 14
+            "pn",             # 15
+            "K^{0}_{L} X",    # 16
+            "K X",            # 17
+            "K^{0}_{S} X",    # 18
+            "exotics",        # 19
+            "multi-body",     # 20
+        ]
+        event_type_compact_labels = [
+            "1 n",      # 0
+            "K#pm X",   # 1
+            "1 K^{0}",  # 2
+            "2 n",      # 3
+            "Soft",     # 4
+            "Other",    # 5
+        ]
 
         self.count_light_ions=count_light_ions
-        self.build1DHistogram("event_type"         , "", 24, -1, 23)
-        self.build1DHistogram("event_type_500mev"  , "", 24, -1, 23)
-        self.build1DHistogram("event_type_2000mev" , "", 24, -1, 23)
-        self.build1DHistogram("event_type_compact"         , "", 8, -1, 7)
-        self.build1DHistogram("event_type_compact_500mev"  , "", 8, -1, 7)
-        self.build1DHistogram("event_type_compact_2000mev" , "", 8, -1, 7)
-        self.build1DHistogram("1n_event_type"              , "", 7,  -1, 6)
-        self.build1DHistogram("pn_vertex_volume"           , "", 13,  -0.5, 12.5)
-        self.build1DHistogram("pn_interaction_material"    , "", 10,  -0.5, 9.5)
+        self.build1DHistogram("event_type"         , "", event_type_labels)
+        self.build1DHistogram("event_type_500mev"  , "", event_type_labels)
+        self.build1DHistogram("event_type_2000mev" , "", event_type_labels)
+        self.build1DHistogram("event_type_compact"         , "", event_type_compact_labels)
+        self.build1DHistogram("event_type_compact_500mev"  , "", event_type_compact_labels)
+        self.build1DHistogram("event_type_compact_2000mev" , "", event_type_compact_labels)
+        self.build1DHistogram("1n_event_type"              , "", [
+            "nn", "pn", "#pi^{+}n", "#pi^{0}n", "other"
+        ])
+        self.build1DHistogram("pn_vertex_volume"           , "", [
+            "Didn't happen", "Else", "W Cooling", "C Cooling", "PCB",
+            "CarbonBasePlate", "Absorber", "Sensor", "Glue", "Motherboard"
+        ])
+        self.build1DHistogram("pn_interaction_material"    , "", [
+            "Didn't happen", "Else", "Si", "W", "FR4", "Steel", "Epoxy",
+            "PVT", "Glue", "Air"
+        ])
         self.build1DHistogram("pn_particle_mult"   , "Photo-nuclear Multiplicity", 200, 0, 200)
         self.build1DHistogram("pn_neutron_mult"    , "Photo-nuclear Neutron Multiplicity", 200,0, 200)
         self.build1DHistogram("pn_gamma_energy"    , "#gamma Energy [MeV]", 100, 0, 10000)
@@ -859,7 +900,7 @@ def __init__(self,name='TrkDeDxMassEstFeatures') :
         self.build1DHistogram("mass_estimate_very_low_p_pion", "Mass Estimate for pions [MeV]", 20, 0., 200.)
         self.build1DHistogram("mass_estimate_very_low_p_kaon", "Mass Estimate for kaons [MeV]", 60, 200., 800.)
         self.build1DHistogram("mass_estimate_very_low_p_proton", "Mass Estimate for proton [MeV]", 40, 800., 1200.)
-        self.build1DHistogram("track_type", "Track Type", 3, 0, 3)
+        self.build1DHistogram("track_type", "Track Type", ['Other', 'Tagger', 'Recoil'])
 
 
 class TrigScintSimDQM(ldmxcfg.Analyzer) :
@@ -998,21 +1039,43 @@ def __init__(self, name='SampleValidation') :
         self.sim_particles_passname = ''
         self.target_scoring_plane_passname = ''
 
-        # primary histograms
-        self.build1DHistogram("primaries_pdgid", "ID of primary particles", 20, 0, 20)
+        pdgid_bin_labels = [
+            "e^{+}",                   # 0
+            "e^{-}",                   # 1
+            "#mu^{+}",                 # 2
+            "#mu^{-}",                 # 3
+            "#gamma",                  # 4
+            "p^{+}",                   # 5
+            "n^{0}",                   # 6
+            "#pi^{+}",                 # 7
+            "#pi^{-}",                 # 8
+            "#pi^{0}",                 # 9
+            "K^{+}",                   # 10
+            "K^{-}",                   # 11
+            "k_{L}",                   # 12
+            "k_{S}",                   # 13
+            "light-N",                 # 14
+            "heavy-N",                 # 15
+            "#Lambda / #Sigma / #Xi",  # 16
+            "A'",                      # 17
+            "else",
+        ]
+
+        #primary histograms
+        self.build1DHistogram("primaries_pdgid", "ID of primary particles", pdgid_bin_labels)
         self.build1DHistogram("primaries_energy", "Energy of primary particles [MeV]", 90, 0, 9000) # range applicable for 4 GeV beam
         self.build2DHistogram("beam_smear", "x [mm]", 30, -150, 150, "y [mm]", 30, -150, 150)
-        self.build1DHistogram("primarydaughters_pdgid", "ID of primary daughters", 20, 0, 20)
+        self.build1DHistogram("primarydaughters_pdgid", "ID of primary daughters", pdgid_bin_labels)
         self.build1DHistogram("daughterphoton_energy", "Energy spectrum of all photons from primary [MeV]", 170, 0, 8500)
 
-        # primary daughter of interest (brem / dark brem) histograms
-        self.build1DHistogram("harddaughters_pdgid", "ID of primary daughters", 20, 0, 20)
+        #primary daughter of interest(brem / dark brem) histograms
+        self.build1DHistogram("harddaughters_pdgid", "ID of primary daughters", pdgid_bin_labels)
         self.build1DHistogram("harddaughters_startZ", "Start z position of hard primary daughter [mm]", 100, -500, 500)
         self.build1DHistogram("harddaughters_endZ", "End z position of hard primary daughter [mm]", 100, -500, 500)
         self.build1DHistogram("harddaughters_energy", "Energy spectrum of hard primary daughter [MeV]", 130, 2000, 8500)
 
-        # daughters of hard brem histograms
-        self.build1DHistogram("hardbremdaughters_pdgid", "ID of hard brem daughters", 20, 0, 20)
+        #daughters of hard brem histograms
+        self.build1DHistogram("hardbremdaughters_pdgid", "ID of hard brem daughters", pdgid_bin_labels)
         self.build1DHistogram("hardbremdaughters_startZ", "Start z position of hard brem daughters  [mm]", 200, -1000, 1000)
         self.build1DHistogram("hardbremdaughters_endZ", "End z position of hard brem daughters  [mm]", 70, -1000, 6000)
         self.build1DHistogram("hardbremdaughters_energy", "Energy of hard brem daughters  [MeV]", 170, 0, 8500)
@@ -1053,7 +1116,7 @@ def __init__(self,name='EcalClusterAnalyzer') :
         self.ecal_sim_hit_coll = "EcalSimHits"
         self.ecal_sim_hit_pass = "" #use whatever pass is available
 
-        # Pass name for ecal digis and rec hits
+        #Pass name for ecal digis and rec hits
         self.rec_hit_coll_name = 'EcalRecHits'
         self.rec_hit_pass_name = ''
 
@@ -1066,9 +1129,10 @@ def __init__(self,name='EcalClusterAnalyzer') :
         self.build1DHistogram("number_of_clusters_first_layer", "Number of CLUE clusters on the first layer", 5, -0.5, 4.5)
         self.build1DHistogram("number_of_clusters_per_layer", "Number of CLUE clusters per layer", 5, -0.5, 4.5)
         self.build1DHistogram("number_of_clusters", "Total number of CLUE clusters", 51, -0.5, 50.5)
-        self.build1DHistogram("correctly_predicted_events", "Correct Cluster Count", 3, 0., 3.)
+        self.build1DHistogram("correctly_predicted_events", "Correct Cluster Count",
+                              ["Underpredicted", "Correct", "Overpredicted"])
 
-        # Need to mod for more than two electrons
+        #Need to mod for more than two electrons
         self.build1DHistogram("ancestors", "Ancestors of particles", 4, 0., 4.)
 
         self.build1DHistogram("same_ancestor", "Percentage of hits in cluster coming from the electron that produced most hits", 21, 0., 105.)

DQM/src/DQM/DarkBremInteraction.cxx

@@ -34,28 +34,6 @@ static double quadsum(const std::initializer_list<double>& list) {
   return sqrt(sum);
 }
 
-void DarkBremInteraction::setHistLabels(
-    const std::string& name, const std::vector<std::string>& labels) {
-  /**
-   * We could probably move this into Framework since it is a relatively
-   * common task. I could even imagine a way of constructing a StrCategory
-   * histogram.
-   */
-  auto h{histograms_.get(name)};
-  for (std::size_t ibin{1}; ibin <= labels.size(); ibin++) {
-    h->GetXaxis()->SetBinLabel(ibin, labels[ibin - 1].c_str());
-  }
-}
-
-void DarkBremInteraction::onProcessStart() {
-  setHistLabels("dark_brem_material",
-                {"Unknown", "C", "PCB", "Glue", "Si", "Al", "W / LYSO", "PVT"});
-
-  setHistLabels("dark_brem_element",
-                {"did not happen", "H 1", "C 6", "O 8", "Na 11", "Si 14",
-                 "Ca 20", "Cu 29", "Y 39", "Lu 71", "W 74", "unlisted"});
-}
-
 void DarkBremInteraction::produce(framework::Event& event) {
   histograms_.setWeight(event.getEventHeader().getWeight());
   const auto& particle_map{

DQM/src/DQM/EcalClusterAnalyzer.cxx

@@ -281,20 +281,6 @@ void EcalClusterAnalyzer::analyze(const framework::Event& event) {
       100. * (ecal_rec_hits.size() - clustered_hits) / ecal_rec_hits.size());
 }
 
-void EcalClusterAnalyzer::setHistLabels(
-    const std::string& name, const std::vector<std::string>& labels) {
-  auto histo{histograms_.get(name)};
-  for (std::size_t ibin{1}; ibin <= labels.size(); ibin++) {
-    histo->GetXaxis()->SetBinLabel(ibin, labels[ibin - 1].c_str());
-  }
-}
-
-void EcalClusterAnalyzer::onProcessStart() {
-  setHistLabels("correctly_predicted_events",
-                {"Underpredicted", "Correct", "Overpredicted"});
-
-}  // end of onProcessStart
-
 }  // namespace dqm
 
 DECLARE_ANALYZER(dqm::EcalClusterAnalyzer)

DQM/src/DQM/HcalVetoResults.cxx

@@ -10,25 +10,6 @@ void HcalVetoResults::configure(framework::config::Parameters &ps) {
   hcal_veto_passname_ = ps.get<std::string>("hcal_veto_passname");
 }
 
-void HcalVetoResults::onProcessStart() {
-  std::vector<TH1 *> hists_hca_lsector = {histograms_.get("max_section")};
-
-  // enum HcalSection { BACK = 0, TOP = 1, BOTTOM = 2, RIGHT = 3, LEFT = 4 };
-  std::vector<std::string> labels_hca_lsector = {"HCAL BACK",    // 1
-                                                 "HCAL TOP",     // 2
-                                                 "HCAL BOTTOM",  // 3
-                                                 "HCAL RIGHT",   // 4
-                                                 "HCAL LEFT",    // 5
-                                                 ""};
-
-  for (int ilabel{1}; ilabel < labels_hca_lsector.size(); ++ilabel) {
-    for (auto &hist : hists_hca_lsector) {
-      hist->GetXaxis()->SetBinLabel(ilabel,
-                                    labels_hca_lsector[ilabel - 1].c_str());
-    }
-  }
-}
-
 void HcalVetoResults::analyze(const framework::Event &event) {
   // Get the veto object
   auto hcal_veto{