feat: fixes towards v1.2 (#179)

* fix b-tagging threshold comparison * remove event offsetting by 1e-6 in histograms * add updated histogram references
iris-hep · Jul 20, 2023 · cce1afb · cce1afb
1 parent a03c3fd
commit cce1afb
Show file tree

Hide file tree

Showing 12 changed files with 30,157 additions and 30,159 deletions.
diff --git a/analyses/cms-open-data-ttbar/reference/histos_100_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_100_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_10_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_10_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_1_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_1_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_200_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_200_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_20_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_20_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_2_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_2_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_50_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_50_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_5_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_5_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/reference/histos_all_file_per_process.json b/analyses/cms-open-data-ttbar/reference/histos_all_file_per_process.json
diff --git a/analyses/cms-open-data-ttbar/ttbar_analysis_pipeline.ipynb b/analyses/cms-open-data-ttbar/ttbar_analysis_pipeline.ipynb
diff --git a/analyses/cms-open-data-ttbar/ttbar_analysis_pipeline.py b/analyses/cms-open-data-ttbar/ttbar_analysis_pipeline.py
@@ -212,7 +212,7 @@ def process(self, events):
             event_filters = event_filters & (ak.count(selected_jets.pt * pt_var_modifier, axis=1) >= 4)
             # at least one b-tagged jet ("tag" means score above threshold)
             B_TAG_THRESHOLD = 0.5
-            event_filters = event_filters & (ak.sum(selected_jets.btagCSVV2 >= B_TAG_THRESHOLD, axis=1) >= 1)
+            event_filters = event_filters & (ak.sum(selected_jets.btagCSVV2 > B_TAG_THRESHOLD, axis=1) >= 1)
 
             # apply event filters
             selected_events = events[event_filters]
@@ -223,7 +223,7 @@ def process(self, events):
             for region in ["4j1b", "4j2b"]:
                 # further filtering: 4j1b CR with single b-tag, 4j2b SR with two or more tags
                 if region == "4j1b":
-                    region_filter = ak.sum(selected_jets.btagCSVV2 >= B_TAG_THRESHOLD, axis=1) == 1
+                    region_filter = ak.sum(selected_jets.btagCSVV2 > B_TAG_THRESHOLD, axis=1) == 1
                     selected_jets_region = selected_jets[region_filter]
                     # use HT (scalar sum of jet pT) as observable
                     pt_var_modifier = (
@@ -333,12 +333,12 @@ def postprocess(self, accumulator):
 def get_query(source: ObjectStream) -> ObjectStream:
     """Query for event / column selection: >=4j >=1b, ==1 lep with pT>25 GeV, return relevant columns
     """
-    return source.Where(lambda e: e.Electron_pt.Where(lambda pt: pt > 25).Count() 
+    return source.Where(lambda e: e.Electron_pt.Where(lambda pt: pt > 25).Count()
                         + e.Muon_pt.Where(lambda pt: pt > 25).Count() == 1)\
                  .Where(lambda f: f.Jet_pt.Where(lambda pt: pt > 25).Count() >= 4)\
-                 .Where(lambda g: {"pt": g.Jet_pt, 
-                                   "btagCSVV2": g.Jet_btagCSVV2}.Zip().Where(lambda jet: 
-                                                                             jet.btagCSVV2 >= 0.5 
+                 .Where(lambda g: {"pt": g.Jet_pt,
+                                   "btagCSVV2": g.Jet_btagCSVV2}.Zip().Where(lambda jet:
+                                                                             jet.btagCSVV2 > 0.5
                                                                              and jet.pt > 25).Count() >= 1)\
                  .Select(lambda h: {"Electron_pt": h.Electron_pt,
                                     "Muon_pt": h.Muon_pt,

diff --git a/analyses/cms-open-data-ttbar/utils/__init__.py b/analyses/cms-open-data-ttbar/utils/__init__.py
@@ -93,8 +93,6 @@ def construct_fileset(n_files_max_per_sample, use_xcache=False, af_name=""):
 def save_histograms(all_histograms, fileset, filename):
     nominal_samples = [sample for sample in fileset.keys() if "nominal" in sample]
 
-    all_histograms += 1e-6  # add minimal event count to all bins to avoid crashes when processing a small number of samples
-
     pseudo_data = (all_histograms[:, :, "ttbar", "ME_var"] + all_histograms[:, :, "ttbar", "PS_var"]) / 2  + all_histograms[:, :, "wjets", "nominal"]
 
     with uproot.recreate(filename) as f: