Developments for boosted category

configs/HH4b/HH4b_parton_matching_config.py

@@ -27,6 +27,7 @@
 from configs.HH4b_common.config_files.configurator_tools import (
     SPANET_TRAINING_DEFAULT_COLUMNS,
     SPANET_TRAINING_DEFAULT_COLUMNS_BTWP,
+    DEFAULT_FATJET_COLUMNS,
     create_DNN_columns_list,
     define_categories,
     define_single_category,
@@ -40,8 +41,10 @@
 )
 from configs.HH4b_common.dnn_input_variables import (
     bkg_morphing_dnn_input_variables,
+    bkg_morphing_boosted_dnn_input_variables,
     # bkg_morphing_dnn_input_variables_altOrder,
     sig_bkg_dnn_input_variables,
+    sig_bkg_boosted_dnn_input_variables,
 )
 from configs.HH4b_common.params.CustomWeights import SF_btag_fixed_multiple_wp
 
@@ -83,6 +86,7 @@
     SCORE=bool(config_options_dict["sig_bkg_dnn"]),
     RUN2=config_options_dict["run2"],
     SPANET=bool(config_options_dict["spanet"]),
+    BOOSTED=config_options_dict["boosted"],
 )
 # print(variables_dict)
 
@@ -130,11 +134,12 @@
     spanet=config_options_dict["spanet"],
     run2=config_options_dict["run2"],
     vr1=config_options_dict["vr1"],
+    boosted=config_options_dict["boosted"],
 )
 # AKA if no model is applied
 # print(onnx_model_dict)
-if all([model == "" for model in onnx_model_dict.values()]):
-    print("Didn't find any onnx model. Will choose region for SPANet training")
+if all([model == "" for model in onnx_model_dict.values()]) and not (config_options_dict["boosted"]):
+    print("Didn't find any onnx model and not running boosted analysis. Will choose region for SPANet training")
     categories_dict = define_single_category("4b_region")
 
 # print("categories_dict", categories_dict)
@@ -165,11 +170,19 @@
 
 assert not (config_options_dict["random_pt"] and config_options_dict["run2"])
 if config_options_dict["dnn_variables"]:
-    total_input_variables = (
-        sig_bkg_dnn_input_variables
-        | bkg_morphing_dnn_input_variables
-        | {"year": ["events", "year"]}
-    )
+    if config_options_dict["boosted"]:
+        print("Using boosted DNN input variables")
+        total_input_variables = (
+            sig_bkg_boosted_dnn_input_variables
+            | bkg_morphing_boosted_dnn_input_variables
+            | {"year": ["events", "year"]}
+        )
+    else:
+        total_input_variables = (
+            sig_bkg_dnn_input_variables
+            | bkg_morphing_dnn_input_variables
+            | {"year": ["events", "year"]}
+        )
     if config_options_dict["spanet"]:
         total_input_variables |= {
             "Delta_pairing_probabilities": ["events", "Delta_pairing_probabilities"],
@@ -194,7 +207,7 @@
     column_listRun2 = create_DNN_columns_list(
         True, not config_options_dict["save_chunk"], total_input_variables, btag=False
     )
-elif all([model == "" for model in onnx_model_dict.values()]):
+elif all([model == "" for model in onnx_model_dict.values()]) and not (config_options_dict["boosted"]):
     if "wp" in config_options_dict["spanet_input_name_list"][-1]:
         print("Taking btag Working Points")
         column_list = get_columns_list(SPANET_TRAINING_DEFAULT_COLUMNS_BTWP, not config_options_dict["save_chunk"])
@@ -235,6 +248,9 @@
             ]
         }
     )
+if config_options_dict["boosted"]:
+    column_list += get_columns_list(DEFAULT_FATJET_COLUMNS, not config_options_dict["save_chunk"])
+
 
 
 bysample_bycategory_column_dict = {}
@@ -288,6 +304,11 @@
 
 # print("bysample_bycategory_weight_dict", bysample_bycategory_weight_dict)
 
+if config_options_dict["boosted"]:
+    skimming_cut_list = cuts.skimming_cut_list_boosted
+else:
+    skimming_cut_list = cuts.skimming_cut_list
+
 cfg = Configurator(
     parameters=parameters,
     datasets={
@@ -314,7 +335,7 @@
     },
     workflow=HH4bbQuarkMatchingProcessor,
     workflow_options=config_options_dict,
-    skim=cuts.skimming_cut_list,
+    skim=skimming_cut_list,
     preselections=preselection,
     categories=categories_dict,
     weights_classes=common_weights

configs/HH4b/HH4b_save_skimmed.py

@@ -30,21 +30,26 @@
     # save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/tharte/HH4b/ntuples/DATA_JetMET_JMENano_skimmed",
 #    save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/tharte/HH4b/ntuples/DATA_JetMET_JMENano_F_skimmed",
     # save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/tharte/HH4b/ntuples/DATA_JetMET_ParkingHH_2023_D_skimmed",
-    save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/tharte/HH4b/testing/DATA_JetMET_JMENano_C_skimmed",
+    # save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/tharte/HH4b/testing/DATA_JetMET_JMENano_C_skimmed",
+    # save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/bevila_t/PostDoc/HH4b/skimmed_files/ggHH_boosted_skimmed",
+    # save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/bevila_t/PostDoc/HH4b/skimmed_files/ttbar_boosted_skimmed",
+    save_skimmed_files="root://t3dcachedb03.psi.ch:1094//pnfs/psi.ch/cms/trivcat/store/user/bevila_t/PostDoc/HH4b/skimmed_files/vbf_boosted_skimmed",
     parameters=parameters,
     datasets={
         "jsons": [
             f"{localdir}/../HH4b_common/datasets/signal_ggF_HH4b.json",
             f"{localdir}/../HH4b_common/datasets/signal_ggF_HH4b_spanet_redirector.json",
+            f"{localdir}/../HH4b_common/datasets/background_TTtoX_redirector.json",
             f"{localdir}/../HH4b_common/datasets/DATA_JetMET_skimmed.json",
             f"{localdir}/../HH4b_common/datasets/QCD.json",
             f"{localdir}/../HH4b_common/datasets/DATA_ParkingHH.json",
             f"{localdir}/../HH4b_common/datasets/DATA_JetMET_redirector.json",
+            f"{localdir}/../HH4b_common/datasets/signal_VBF_HH4b_redirector.json",
         ],
         "filter": {
             "samples": (
                 [
-                    "DATA_JetMET_JMENano_C",
+                    # "DATA_JetMET_JMENano_C",
                     # "DATA_JetMET_JMENano_D",
                     # "DATA_JetMET_JMENano_E",
                     # "DATA_JetMET_JMENano_F",
@@ -55,7 +60,12 @@
                     # "DATA_ParkingHH_2023_Cv4",
                     # "DATA_ParkingHH_2023_Dv1",
                     # "DATA_ParkingHH_2023_Dv2",
-                    # "GluGlutoHHto4B_spanet"
+                    # "GluGlutoHHto4B_spanet",
+                    # "TTto4Q",
+                    # "TTtoLNu2Q",
+                    # "TTto2L2Nu",
+                    "VBF_HHto4B",
+                    "VBF_HHto4B_BSM"
                 ]
             ),
             "samples_exclude": [],
@@ -66,7 +76,8 @@
     workflow=HH4bbQuarkMatchingProcessorDummy,
     workflow_options={
     },
-    skim=cuts.skimming_cut_list,
+    # skim=cuts.skimming_cut_list,
+    skim=cuts.skimming_cut_list_boosted,
     # skim=[
     #     get_HLTsel(primaryDatasets=["JetMET"]),
     #     # get_HLTsel(primaryDatasets=["ParkingHH"]),

configs/HH4b/workflow_dummy.py

@@ -1,6 +1,8 @@
 import awkward as ak
 
 from pocket_coffea.workflows.base import BaseProcessorABC
+from configs.HH4b_common.workflow_common import HH4bCommonProcessor
+
 class HH4bbQuarkMatchingProcessorDummy(HH4bCommonProcessor):
 
     def count_objects(self, variation):

configs/HH4b_common/config_files/boosted_method_cutflow.py

@@ -0,0 +1,26 @@
+from configs.HH4b_common.dnn_input_variables import (
+    bkg_morphing_boosted_dnn_input_variables,
+    sig_bkg_boosted_dnn_input_variables,
+)
+
+from configs.HH4b_common.config_files.default_config import default_onnx_model_dict as onnx_model_dict
+
+from configs.HH4b_common.config_files.default_config import default_config_options_dict as config_options_dict
+
+
+onnx_model_dict |= {
+    "bkg_morphing_dnn": "/work/bevila_t/PostDoc/HH4b/Output/ML_trainings/bkg_reweigting_boosted/first_test/best_models/average_model_from_onnx.onnx", 
+}
+
+config_options_dict |= {
+    "dnn_variables": True,
+    "sig_bkg_dnn_input_variables": sig_bkg_boosted_dnn_input_variables,
+    "bkg_morphing_dnn_input_variables": bkg_morphing_boosted_dnn_input_variables,
+    "run2": False,
+    "fifth_jet": "pt",
+    "pad_value": -999.0,
+    "add_jet_spanet": False,
+    "boosted": True,
+    "boosted_presel": True,
+    "split_qcd": True,
+}| onnx_model_dict | {  "spanet": False,}