[tutorials][ML] Add RDataLoader tutorial for XGBoost

tutorials/CMakeLists.txt

@@ -78,6 +78,7 @@ if(MSVC AND NOT win_broken_tests)
   list(APPEND dataframe_veto machine_learning/ml_dataloader_filters_vectors.py)
   list(APPEND dataframe_veto machine_learning/ml_dataloader_Higgs_Classification.py)
   list(APPEND dataframe_veto machine_learning/ml_dataloader_resampling.py)
+  list(APPEND dataframe_veto machine_learning/ml_dataloader_XGBoost.py)
   # df036* and df037* seem to trigger OS errors when trying to delete the
   # test files created in the tutorials. It is unclear why.
   list(APPEND dataframe_veto analysis/dataframe/df036_missingBranches.C)
@@ -132,6 +133,7 @@ if (NOT dataframe)
     list(APPEND dataframe_veto machine_learning/ml_dataloader_filters_vectors.py)
     list(APPEND dataframe_veto machine_learning/ml_dataloader_resampling.py)
     list(APPEND dataframe_veto machine_learning/ml_dataloader_Higgs_Classification.py)
+    list(APPEND dataframe_veto machine_learning/ml_dataloader_XGBoost.py)
     # RooFit tutorials depending on RDataFrame
     list(APPEND dataframe_veto
         roofit/roofit/rf408_RDataFrameToRooFit.C

tutorials/machine_learning/index.md

@@ -137,5 +137,6 @@
 | ml_dataloader_PyTorch.py | Loading batches of events from a ROOT dataset into a basic PyTorch workflow. |
 | ml_dataloader_TensorFlow.py | Loading batches of events from a ROOT dataset into a basic TensorFlow workflow. |
 | ml_dataloader_Higgs_Classification.py | Loading batches of events from different files for a data-normalization workflow. |
-| ml_dataloader_resampling.py | Loading batches of events from an imbalanced ROOT dataset and balancing them. | 
+| ml_dataloader_resampling.py | Loading batches of events from an imbalanced ROOT dataset and balancing them. |
+| ml_dataloader_XGBoost.py | Training XGBoost directly from remote ROOT files. |
 

tutorials/machine_learning/ml_dataloader_XGBoost.py

@@ -0,0 +1,96 @@
+## \file
+## \ingroup tutorial_ml
+## \notebook -nodraw
+## This tutorial demonstrates XGBoost training using RDataLoader to load data
+## directly from ROOT files without intermediate preparation steps.
+## This is an alternative implementation to tutorial tmva101_Training.py,
+## which stores intermediate files after train/test splitting
+## and uses numpy arrays to load data.
+##
+## \macro_code
+## \macro_output
+##
+## \date July 2026
+## \author Silia Taider
+
+import ROOT
+
+variables = ["Muon_pt_1", "Muon_pt_2", "Electron_pt_1", "Electron_pt_2"]
+
+
+def filter_events(df):
+    """Reduce initial dataset to only events which shall be used for training"""
+    return df.Filter("nElectron>=2 && nMuon>=2", "At least two electrons and two muons")
+
+
+def define_variables(df):
+    """Define the variables which shall be used for training"""
+    return (
+        df
+        .Define("Muon_pt_1", "Muon_pt[0]")
+        .Define("Muon_pt_2", "Muon_pt[1]")
+        .Define("Electron_pt_1", "Electron_pt[0]")
+        .Define("Electron_pt_2", "Electron_pt[1]")
+    )
+
+
+def prepare_rdf(filename, label_value):
+    """Load, filter, define variables, and add label column"""
+    filepath = "root://eospublic.cern.ch//eos/root-eos/cms_opendata_2012_nanoaod/" + filename
+    df = ROOT.RDataFrame("Events", filepath)
+    df = filter_events(df)
+    df = define_variables(df)
+    df = df.Define("label", f"{label_value}.0")
+    return df
+
+
+def load_data():
+    """Load signal and background data"""
+    rdf_sig = prepare_rdf("SMHiggsToZZTo4L.root", 1)
+    rdf_bkg = prepare_rdf("ZZTo2e2mu.root", 0)
+
+    # Compute class-balancing weights
+    num_sig = rdf_sig.Count().GetValue()
+    num_bkg = rdf_bkg.Count().GetValue()
+    num_all = num_sig + num_bkg
+
+    rdf_sig = rdf_sig.Define("weight", f"{num_all}.0/{num_sig}.0")
+    rdf_bkg = rdf_bkg.Define("weight", f"{num_all}.0/{num_bkg}.0")
+
+    loader = ROOT.Experimental.ML.RDataLoader(
+        [rdf_sig, rdf_bkg],
+        columns=variables + ["label", "weight"],
+        target="label",
+        weights="weight",
+        batch_size=num_all,  # Load all data in one batch
+        drop_remainder=False,
+        set_seed=42,
+    )
+
+    # Split into training and testing sets
+    train, test = loader.train_test_split(test_size=0.5)
+
+    X_train, y_train, w_train = next(iter(train.as_numpy()))
+    X_test, y_test, w_test = next(iter(test.as_numpy()))
+
+    # Flatten target and weights from (n,1) to (n,) as expected by XGBoost
+    return (X_train, y_train.ravel(), w_train.ravel(), X_test, y_test.ravel(), w_test.ravel())
+
+
+if __name__ == "__main__":
+    from sklearn.metrics import roc_auc_score
+    from xgboost import XGBClassifier
+
+    X_train, y_train, w_train, X_test, y_test, w_test = load_data()
+
+    print(f"Training events: {X_train.shape[0]}")
+    print(f"Testing events:  {X_test.shape[0]}")
+
+    bdt = XGBClassifier(max_depth=3, n_estimators=500)
+    bdt.fit(X_train, y_train, sample_weight=w_train)
+
+    # Evaluate on test set
+    y_proba = bdt.predict_proba(X_test)[:, 1]
+    auc = roc_auc_score(y_test, y_proba)
+
+    print(f"Training done. ROC AUC: {auc:.4f}")