Implemented correlation matrix in CMS 13 TeV Afb data set

Author: enocera

nnpdf_data/nnpdf_data/commondata/CMS_Z0_13TEV/data_AFB.yaml

@@ -1,64 +1,64 @@
 data_central:
-- -0.0088087
-- -0.033938
-- -0.021486
-- -0.01283
-- -0.0010866
-- 0.0017601
-- 0.014298
-- 0.023975
-- 0.02063
-- 0.04671
-- 0.04132
-- -0.065286
-- -0.098281
-- -0.064055
-- -0.035366
-- -0.0091856
-- 0.015125
-- 0.03051
-- 0.067186
-- 0.078415
-- 0.12568
-- 0.17502
-- -0.13264
-- -0.17966
-- -0.1151
-- -0.063701
-- -0.017779
-- 0.026632
-- 0.057933
-- 0.10609
-- 0.14562
-- 0.21103
-- 0.30288
-- -0.21994
-- -0.26467
-- -0.13147
-- 0.039362
-- 0.18504
-- 0.31841
-- 0.43577
-- -0.28089
-- -0.37435
-- -0.20281
-- 0.057507
-- 0.27497
-- 0.45481
-- 0.63044
-- -0.37894
-- -0.52607
-- -0.27475
-- 0.080962
-- 0.35646
-- 0.61701
-- 0.79779
-- -0.45325
-- 0.1066
-- 0.49703
-- -0.59247
-- 0.12488
-- 0.65322
-- -0.67106
-- 0.15607
-- 0.74539
+- -0.0086482
+- -0.033992
+- -0.021625
+- -0.012986
+- -0.0011154
+- 0.0017664
+- 0.01436
+- 0.024078
+- 0.020788
+- 0.046908
+- 0.04152
+- -0.064782
+- -0.098499
+- -0.064554
+- -0.035726
+- -0.0093437
+- 0.015155
+- 0.030693
+- 0.067548
+- 0.07893
+- 0.12633
+- 0.17564
+- -0.13178
+- -0.18003
+- -0.11602
+- -0.064338
+- -0.018044
+- 0.026687
+- 0.05825
+- 0.10674
+- 0.14654
+- 0.21216
+- 0.30394
+- -0.21858
+- -0.26521
+- -0.13269
+- 0.039435
+- 0.18622
+- 0.32014
+- 0.43738
+- -0.27888
+- -0.37527
+- -0.20469
+- 0.057624
+- 0.27664
+- 0.45724
+- 0.63255
+- -0.3762
+- -0.52712
+- -0.2774
+- 0.081106
+- 0.35895
+- 0.62059
+- 0.80039
+- -0.45736
+- 0.10689
+- 0.49997
+- -0.59639
+- 0.1251
+- 0.65723
+- -0.67574
+- 0.15638
+- 0.75026

nnpdf_data/nnpdf_data/commondata/CMS_Z0_13TEV/filter.py

@@ -7,6 +7,7 @@
 """
 
 import yaml
+from nnpdf_data.filter_utils.utils import cormat_to_covmat,covmat_to_artunc
 
 def get_tables():
     """
@@ -31,6 +32,9 @@ def get_all():
     data_central = []
     kinematics = []
     uncertainties = []
+    art_uncert = []
+    art_uncertainties = []
+    correlations = []
     hepdata_tables = get_tables()
 
     table=hepdata_tables[0]
@@ -58,36 +62,58 @@ def get_all():
                     'max': kin_values_mll_max[i]["value"]}}
         kinematics.append(kin)
 
-    # Uncertainties
+    # Artificial uncertainties (from correlation matrix)
+    # Errors
     for data_value in data_values:
         errors = data_value["errors"]
         uncertainty = {}
         for error in errors:
-            uncertainty[error["label"]] = error["symerror"]
-            uncertainty.update(uncertainty)
-        uncertainties.append(uncertainty)
+            uncertainties.append(error["symerror"])
+
+    # Correlation coefficients
+    table_corr = hepdata_tables[1]
+    with open(table_corr, 'r') as f:
+        input = yaml.safe_load(f)
+    corr_coeffs = input["dependent_variables"][0]["values"]
+    for corr_coeff in corr_coeffs:
+        correlations.append(corr_coeff["value"])
+
+    covmat = cormat_to_covmat(uncertainties,correlations)
+    art_uncertainties = covmat_to_artunc(len(uncertainties),covmat,1)
+    
+    # Remap uncertainties
+    for uncertainties in art_uncertainties:
+        j = 1
+        artificial_unc={}
+        for uncertainty in uncertainties:
+            key = "artificial uncertainty " + str(j)
+            art_unc = {key: uncertainty}
+            j = j + 1
+            
+            artificial_unc.update(art_unc)
 
-    return(data_central, kinematics, uncertainties)
+        art_uncert.append(artificial_unc)
+            
+    return(data_central, kinematics, art_uncert)
 
 def filter_CMS_Z0_13TEV_PT():
     """
     Dumps data, kinematics, and uncertainties on .yaml files
     """
-    #central_values, kinematics, uncertainties = get_all()
     central_values, kinematics, uncertainties = get_all()
     # Central values
     data_central_yaml = {"data_central": central_values}
     # Kinematics
     kinematics_yaml = {"bins": kinematics}
     # Uncertainties
-    treatment = {"A4 uncertainty": "ADD"}
-    correlation = {"A4 uncertainty": "UNCORR"}
+    treatment = {"artificial uncertainty": "ADD"}
+    correlation = {"artificial uncertainty": "CORR"}
     definitions = {}
     for key,value in uncertainties[0].items():
         definition = {key :
                       {"description": key,
-                       "treatment": treatment[key],
-                       "type": correlation[key]}}
+                       "treatment": "ADD",
+                       "type": "CORR"}}
         definitions.update(definition)
     uncertainties_yaml = {"definitions": definitions,"bins": uncertainties}
 

nnpdf_data/nnpdf_data/commondata/CMS_Z0_13TEV/metadata.yaml

@@ -8,7 +8,7 @@ iNSPIRE:
   url: https://inspirehep.net/literature/2818125
 hepdata:
   url: https://www.hepdata.net/record/ins2818125
-  version: 2
+  version: 1
 
 nnpdf_metadata:
   nnpdf31_process: "DY NC"