diff --git a/pynuml/process/hitgraph.py b/pynuml/process/hitgraph.py
index 3abc0e4..a36e378 100644
--- a/pynuml/process/hitgraph.py
+++ b/pynuml/process/hitgraph.py
@@ -58,7 +58,7 @@ def columns(self) -> dict[str, list[str]]:
             else:
                 groups['event_table'] = keys
         if self.label_position:
-            groups["hit_table"].extend(["x_position", "y_position", "z_position"])
+            groups["edep_table"] = []
         return groups
 
     @property
@@ -98,14 +98,13 @@ def __call__(self, evt: 'pynuml.io.Event') -> tuple[str, Any]:
             
             # charge-weighted average of 3D position
             if self.label_position:
-                edeps = edeps.drop("g4_id", axis="columns")
-                edeps["x_position"] = edeps.x_position * edeps.energy
-                edeps["y_position"] = edeps.y_position * edeps.energy
-                edeps["z_position"] = edeps.z_position * edeps.energy
+                edeps = edeps[["hit_id", "energy", "x_position", "y_position", "z_position"]]
+                for col in ["x_position", "y_position", "z_position"]:
+                    edeps.loc[:, col] *= edeps.energy
                 edeps = edeps.groupby("hit_id").sum()
-                edeps["x_position"] = edeps.x_position / edeps.energy
-                edeps["y_position"] = edeps.y_position / edeps.energy
-                edeps["z_position"] = edeps.z_position / edeps.energy
+                for col in ["x_position", "y_position", "z_position"]:
+                    edeps.loc[:, col] /= edeps.energy
+                edeps = edeps.drop("energy", axis="columns")
                 hits = edeps.merge(hits, on="hit_id", how="right")
 
             hits['filter_label'] = ~hits[energy_col].isnull()
diff --git a/tests/test_process.py b/tests/test_process.py
index f340606..b7fcc08 100644
--- a/tests/test_process.py
+++ b/tests/test_process.py
@@ -20,3 +20,25 @@ def test_process_uboone():
             continue
         plot.plot(data, target='semantic', how='true', filter='show')
         plot.plot(data, target='instance', how='true', filter='true')
+
+def test_process_dune_nutau():
+    """Test graph processing with DUNE beam nutau dataset"""
+    f = pynuml.io.File("/raid/nugraph/dune-nutau/NeutrinoML_r00140_s00000_ts814031.h5")
+    processor = pynuml.process.HitGraphProducer(
+        file=f,
+        semantic_labeller=pynuml.labels.StandardLabels(),
+        event_labeller=pynuml.labels.FlavorLabels(),
+        label_position=True)
+    plot = pynuml.plot.GraphPlot(
+        planes=["u", "v", "y"],
+        classes=pynuml.labels.StandardLabels().labels[:-1])
+    f.read_data(0, 100)
+    evts = f.build_evt()
+    for evt in evts:
+        _, data = processor(evt)
+        if not data:
+            continue
+        plot.plot(data, target="filter", how="true", filter="show")
+        plot.plot(data, target='semantic', how='true', filter='show')
+        plot.plot(data, target='instance', how='true', filter='true')
+        plot.plot(data, target="semantic", how="true", filter="show", xyz=True)