Implementation of CMS_Z0J_13TeV - general cleanup

nnpdf_data/nnpdf_data/commondata/CMS_Z0J_13TEV/data.yaml

@@ -0,0 +1,84 @@
+data_central:
+- 1.8444
+- 1.7141
+- 1.5243
+- 1.329
+- 1.1897
+- 0.99489
+- 0.75792
+- 0.49533
+- 0.26635
+- 0.12569
+- 0.046018
+- 0.0091988
+- 0.0011748
+- 3.5385e-05
+- 37.237
+- 36.608
+- 34.621
+- 32.542
+- 30.324
+- 28.162
+- 25.957
+- 24.036
+- 22.098
+- 20.7
+- 18.4
+- 15.824
+- 13.912
+- 12.011
+- 10.329
+- 8.5419
+- 7.0432
+- 5.547
+- 4.2406
+- 3.0556
+- 1.9436
+- 1.0362
+- 0.40703
+- 0.14152
+- 0.059995
+- 0.032086
+- 0.018046
+- 0.0090947
+- 0.0040887
+- 0.0019895
+- 0.0010353
+- 0.0005749
+- 8.2356e-05
+- 0.81127
+- 0.75465
+- 0.67379
+- 0.59073
+- 0.52103
+- 0.42299
+- 0.32974
+- 0.24939
+- 0.17329
+- 0.10268
+- 0.046168
+- 0.011162
+- 0.0017995
+- 6.0169e-05
+- 0.090178
+- 0.087088
+- 0.082663
+- 0.07576
+- 0.065335
+- 0.054814
+- 0.046471
+- 0.035346
+- 0.020531
+- 0.0043687
+- 0.00010193
+- 0.0069927
+- 0.0068868
+- 0.0061386
+- 0.0060243
+- 0.0052844
+- 0.0051086
+- 0.0045253
+- 0.004045
+- 0.0021176
+- 0.00057801
+- 2.4683e-05

nnpdf_data/nnpdf_data/commondata/CMS_Z0J_13TEV/filter.py

@@ -0,0 +1,130 @@
+"""
+This file contains the piece of code needed to implement the CMS ZpT measurement
+at 13 TeV. Systematic uncertainties are implemented starting from the breakdown
+available on HepData. The correlation treatment follows the approach mentioned
+in the paper (see the lines immediately before Sect. 6): "The systematic
+and statistical uncertainties are obtained using the linear combination method 
+described in Ref. [77], considering as fully correlated the uncertainties in 
+the jet energy scale and resolution, the pileup, the background subtraction, 
+b tagging, and the integrated luminosity. Other uncertainties are considered 
+as uncorrelated." Note that correlations are kept not only across different 
+pT(ll) bins, but also across different m(ll) bins. Covariance matrices, albeit 
+only for covariances across pT(ll) bins, are available on HepData. These are
+disregarded, because they do not include any correlations in m(ll). Their 
+inspection confirms that indeed the aforementioned uncertainties are fully
+correlated. The other uncertainties are weakly correlated or anti-correlated. 
+"""
+
+import yaml, re
+
+def get_tables():
+    """
+    Get the Hepdata tables, given the tables and version specified in metadata 
+    """
+    prefix = "rawdata/HEPData-ins2079374"
+    with open("metadata.yaml", "r") as file:
+        metadata = yaml.safe_load(file)
+
+    version = metadata["hepdata"]["version"]
+    tables = metadata["implemented_observables"][0]["tables"]
+    hepdata_tables = []
+
+    for table in tables:
+        hepdata_tables.append(f"{prefix}-v{version}-pT_ll_mass_{table}.yaml")
+
+    return hepdata_tables
+
+def get_all():
+    """
+    Returns data, kinematics and uncertainties for dumping in the .yaml files
+    """
+    data_central = []
+    kinematics = []
+    uncertainties = []
+
+    hepdata_tables = get_tables()
+    for table in hepdata_tables:
+        if (table=="rawdata/HEPData-ins2079374-v1-pT_ll_mass_76-106.yaml"):
+            n = 4
+        else:
+            n=2
+        with open(table, 'r') as f:
+            input = yaml.safe_load(f)
+        ranges = list(map(int, re.search(r"mass_(\d+)-(\d+)", table).groups()))
+
+        min_mll = ranges[0]
+        mid_mll = 0.5 * (ranges[0] + ranges[1])
+        max_mll = ranges[1]
+
+        # Central values
+        data_values = input["dependent_variables"][0]["values"]
+        for data_value in data_values[n:]:
+            data_central.append(data_value["value"])
+
+        # Kinematic bins
+        kin_values = input["independent_variables"][0]["values"]
+        for kin_value in kin_values[n:]:
+            kin = {
+                'pT': {'min': kin_value['low'],
+                       'mid': 0.5 * (kin_value['low'] + kin_value['high']),
+                       'max': kin_value['high']},
+                'm_ll': {'min': min_mll, 'mid': mid_mll, 'max': max_mll},}
+            kinematics.append(kin)
+        # Uncertainties
+        for data_value in data_values[n:]:
+            errors = data_value["errors"]
+            uncertainty = {}
+            for error in errors:
+                uncertainty[error["label"]] = error["symerror"]
+                uncertainty.update(uncertainty)
+
+            uncertainties.append(uncertainty)
+
+    return (data_central, kinematics, uncertainties)
+
+def filter_CMS_Z0J_13TEV_PT():
+    """
+    Dumps data, kinematics, and uncertainties on .yaml files
+    """
+    central_values, kinematics, uncertainties = get_all()
+    # Central values
+    data_central_yaml = {"data_central": central_values}
+    # Kinematics
+    kinematics_yaml = {"bins": kinematics}
+    # Uncertainties
+    treatment = {"Data stat.": "ADD",
+                 "Unfolding stat.": "ADD",
+                 "Unfolding model": "ADD",
+                 "Int. luminosity": "MULT",
+                 "Lepton energy": "ADD",
+                 "Efficiency": "ADD",
+                 "Backgrounds": "MULT",
+                 "Jet energy": "MULT",
+                 "Others": "MULT"}
+    correlation = {"Data stat.": "UNCORR",
+                   "Unfolding stat.": "UNCORR",
+                   "Unfolding model": "UNCORR",
+                   "Int. luminosity": "CMSLUMI16",
+                   "Lepton energy": "UNCORR",
+                   "Efficiency": "UNCORR",
+                   "Backgrounds": "CORR",
+                   "Jet energy": "CORR",
+                   "Others": "CORR"}
+    definitions = {}
+    for key,value in uncertainties[0].items():
+        definition = {key :
+                      {"description": key + " unc. from HepData",
+                       "treatment": treatment[key],
+                       "type": correlation[key]}}
+        definitions.update(definition)
+    uncertainties_yaml = {"definitions": definitions,"bins": uncertainties}
+
+    with open("data.yaml", "w") as file:
+        yaml.dump(data_central_yaml, file, sort_keys=False)
+    with open("kinematics.yaml", "w") as file:
+        yaml.dump(kinematics_yaml, file, sort_keys=False)
+    with open("uncertainties.yaml", "w") as file:
+        yaml.dump(uncertainties_yaml, file, sort_keys=False)
+
+if __name__ == "__main__":
+    filter_CMS_Z0J_13TEV_PT()