fix: [Bug]: Pipeline error: Column 1 named vector expected length 4...

packages/graphrag-vectors/graphrag_vectors/lancedb.py

@@ -102,6 +102,15 @@ def load_documents(self, documents: list[VectorStoreDocument]) -> None:
         if not ids:
             return
 
+        actual_vector_size = len(vectors[0])
+        if actual_vector_size != self.vector_size:
+            msg = (
+                f"Embedding dimension {actual_vector_size} does not match "
+                f"configured vector_size {self.vector_size} for index '{self.index_name}'. "
+                f"Please update vector_size in your vector_store configuration."
+            )
+            raise ValueError(msg)
+
         flat_vector = np.concatenate(vectors).astype(np.float32)
         flat_array = pa.array(flat_vector, type=pa.float32())
         vector_column = pa.FixedSizeListArray.from_arrays(flat_array, self.vector_size)

tests/integration/vector_stores/test_lancedb.py

@@ -547,6 +547,29 @@ def test_user_defined_date_field_exploded(self):
         finally:
             shutil.rmtree(temp_dir)
 
+    def test_load_documents_vector_dimension_mismatch(self):
+        """Test that load_documents raises ValueError when vector dimension does not match vector_size."""
+        temp_dir = tempfile.mkdtemp()
+        try:
+            vector_store = LanceDBVectorStore(
+                db_uri=temp_dir, index_name="dim_mismatch", vector_size=10
+            )
+            vector_store.connect()
+            vector_store.create_index()
+
+            documents = [
+                VectorStoreDocument(id="1", vector=[0.1, 0.2, 0.3, 0.4, 0.5]),
+                VectorStoreDocument(id="2", vector=[0.2, 0.3, 0.4, 0.5, 0.6]),
+            ]
+
+            with pytest.raises(
+                ValueError,
+                match="Embedding dimension 5 does not match configured vector_size 10",
+            ):
+                vector_store.load_documents(documents)
+        finally:
+            shutil.rmtree(temp_dir)
+
     def test_vector_store_customization(self, sample_documents):
         """Test vector store customization with LanceDB."""
         temp_dir = tempfile.mkdtemp()