Updates and bug fixes

jet_tools/CompareClusters.py

@@ -14,15 +14,6 @@
 from jet_tools import TrueTag, InputTools, JetQuality, PlottingTools, RemovePileup
 import pandas as pd
 
-SCORE_COLS = ["QualityWidth", "QualityFraction",
-              "AveSignalMassRatio", "AveBGMassRatio",
-              "AveDistancePT", "AveDistancePhi", "AveDistanceRapidity",
-              "AvePercentFound",
-              "AveDistanceBG", "AveDistanceSignal", "AveSeperateJets",
-              "SeperateAveSignalMassRatio", "SeperateAveBGMassRatio",
-              "SeperateAveDistancePT", "SeperateAveDistancePhi",
-              "SeperateAveDistanceRapidity",
-              "SeperateAveDistanceBG", "SeperateAveDistanceSignal"]
 
 def get_best(eventWise, jet_class):
     """
@@ -660,7 +651,7 @@ def remove_scores(eventWise):
     """
     suffixes_to_remove = ["DistancePT", "DistancePhi", "DistanceRapidity",
                           "PercentFound", "BGMassRatio", "SignalMassRatio"]
-    suffixes_to_remove += SCORE_COLS
+    suffixes_to_remove += Constants.SCORE_COLS
     if isinstance(eventWise, str):
         eventWise = Components.EventWise.from_file(eventWise)
     remove_cols = [name for name in
@@ -783,7 +774,7 @@ def append_scores(eventWise, dijet_mass=None, end_time=None, duration=np.inf,
             new_hyperparameters[name + "_QualityWidth"] = best_width
             new_hyperparameters[name + "_QualityFraction"] = best_fraction
         # check if it has already been scored asside from quality scores
-        found = [name + '_' + col in eventWise.columns for col in SCORE_COLS
+        found = [name + '_' + col in eventWise.columns for col in Constants.SCORE_COLS
                  if "Quality" not in col]
         if not np.all(found) or overwrite:
             if not silent:
@@ -879,7 +870,7 @@ def tabulate_scores(eventWise_paths, variable_cols=None, score_cols=None):
     if variable_cols is None:
         variable_cols = sorted(FormJets.get_jet_input_params())
     if score_cols is None:
-        score_cols = SCORE_COLS
+        score_cols = Constants.SCORE_COLS
     # also record jet class, eventWise.svae_name and jet_name
     all_cols = ["jet_name", "jet_class", "eventWise_name"] + variable_cols + score_cols
     table = []
@@ -893,7 +884,7 @@ def tabulate_scores(eventWise_paths, variable_cols=None, score_cols=None):
         # there may be "score only" jets too;
         jet_names += [name.split('_', 1)[0]
                       for name in eventWise.hyperparameter_columns
-                      if "Jet_" in name and name.endswith(SCORE_COLS[0])]
+                      if "Jet_" in name and name.endswith(Constants.SCORE_COLS[0])]
         for name in sorted(set(jet_names)):
             row = [name, name.split("Jet", 1)[0], eventWise_name]
             for part in variable_cols + score_cols:
@@ -1691,7 +1682,6 @@ def project_2d(show_params, best_slice, all_cols, variable_cols, score_cols,
     mult_col = "AveSeperateJets"
     # if we are taking the best slice remove all other table rows
     if best_slice:
-        import ipdb; ipdb.set_trace()
         bg_values = table[bg_col].values
         sg_values = table[sg_col].values
         combined = np.sqrt(0.53 * np.square(bg_values) + np.square(sg_values))

jet_tools/MassPeaks.py

@@ -866,7 +866,7 @@ def get_data_for_correct_pairs(eventWise, jet_names, hard_event_pids,
     elif isinstance(eventWise, str):
         import os
         if Components.EventWise.pottential_file(eventWise):
-            Components.EventWise.from_file(eventWise)
+            eventWise = Components.EventWise.from_file(eventWise)
             return get_data_for_correct_pairs(eventWise, jet_names, hard_event_pids,
                                               additional_decay_pid, jet_pt_cut,
                                               signal_background)
@@ -879,26 +879,28 @@ def get_data_for_correct_pairs(eventWise, jet_names, hard_event_pids,
         else:
             raise ValueError("eventWise should be a path to an eventWise or directory")
     elif hasattr(eventWise, '__iter__'):
-        descendants_masses = {}
-        masses = []
+        descendants_mass = {}
+        masses = {}
         # assume it is a list of eventWises
         for item in eventWise:
             print(item)
             desc_mass, mass = get_data_for_correct_pairs(item, jet_names, hard_event_pids,
                                                          additional_decay_pid, jet_pt_cut,
                                                          signal_background)
             for key in desc_mass:
-                if key not in descendants_masses:
-                    descendants_masses[key] = desc_mass[key]
+                if key not in descendants_mass:
+                    descendants_mass[key] = desc_mass[key]
                 else:
-                    descendants_masses[key] = \
-                        ak.concatenate([descendants_masses[key], desc_mass[key]])
+                    descendants_mass[key] = \
+                        ak.concatenate([descendants_mass[key], desc_mass[key]])
             for key in mass:
                 if key not in masses:
                     masses[key] = mass[key]
                 else:
-                    masses[key] = ak.concatenate([masses[key], mass[key]])
-        return descendants_masses, masses
+                    for jet_key in mass[key]:
+                        existing = masses[key][jet_key]
+                        masses[key][jet_key] = ak.concatenate([existing, mass[key][jet_key]])
+        return descendants_mass, masses
     else:
         raise TypeError(f"eventWise type {type(eventWise)} not understood")
 
@@ -929,7 +931,7 @@ def plt_correct_pairs(eventWise, jet_names, hard_event_pids,
         ax_function = plot_correct_scatter_axis
     else:
         ax_function = plot_correct_means_axis
-    masses, descendants_mass = get_data_for_correct_pairs(eventWise, jet_names, hard_event_pids,
+    descendants_mass, masses = get_data_for_correct_pairs(eventWise, jet_names, hard_event_pids,
                                                           additional_decay_pid, jet_pt_cut,
                                                           signal_background)
     # work out the maximum counts for each pid in the data

jet_tools/ParallelFormJets.py

@@ -1209,9 +1209,19 @@ def run_list(eventWise_path, end_time, class_list, param_list, name_list):
 
 fix_CALEv2_1 = {'SeedGenerator': 'PtCenter',
                 'WeightExponent': 0.}
+
+fix_CALEv2_2 = {'SeedGenerator': 'PtCenter',
+                'WeightExponent': 1.}
+
 scan_CALEv2_1 = {'Sigma': [0.1, 0.5, 1., 1.5, 2., 2.5, 3.],
                  'Cutoff': [0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]}
 
+scan_CALEv2_2 = {'Sigma': [0.5, 1., 1.5, 2., 2.5, 3.],
+                 'Cutoff': [0.2, 0.3, 0.4, 0.5, 0.6, 0.7]}
+
+scan_CALEv2_3 = {'Sigma': [0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 1.],
+                 'Cutoff': [0.1, 0.15, 0.2, 0.25, 0.3]}
+
 def tabulate_fragments(eventWise_paths):
     """ Makes the assumption that jets with the same name are the same jet"""
     if isinstance(eventWise_paths, Components.EventWise):