Fixed np.NaN to np.nan

bucketed_scene_flow_eval/eval/base_per_frame_sceneflow_eval.py

@@ -255,8 +255,8 @@ def _category_to_average_stats(
             counts = np.array([v.count for v in values])
             # Average of the epes weighted by the counts
 
-            weighted_split_avg_speed = np.NaN
-            weighted_average_epe = np.NaN
+            weighted_split_avg_speed = np.nan
+            weighted_average_epe = np.nan
             if counts.sum() > 0:
                 valid_counts_mask = counts > 0
 

tests/datasets/nuscenes/nuscenes_tests.py_bak

@@ -0,0 +1,35 @@
+from pathlib import Path
+
+import pytest
+
+from bucketed_scene_flow_eval.datasets.nuscenes import NuScenesRawSequenceLoader
+
+
+@pytest.fixture
+def nuscenes_loader() -> NuScenesRawSequenceLoader:
+    return NuScenesRawSequenceLoader(
+        sequence_dir=Path("/tmp/nuscenes"),
+        version="v1.0-mini",
+        split="mini_train",
+        verbose=False,
+    )
+
+
+def test_nuscenes_loader_basic_load_and_len_check(nuscenes_loader: NuScenesRawSequenceLoader):
+    assert len(nuscenes_loader) > 0, f"no sequences found in {nuscenes_loader}"
+    expected_lens= [236, 236, 236, 233, 223, 239, 231, 231]
+    assert len(nuscenes_loader) == len(
+        expected_lens
+    ), f"expected {len(expected_lens)} sequences, got {len(nuscenes_loader)}"
+
+    num_loop_iterations = 0
+    for sequence_id, expected_len in zip(nuscenes_loader.get_sequence_ids(), expected_lens):
+        num_loop_iterations += 1
+        nusc_seq = nuscenes_loader.load_sequence(sequence_id)
+        assert (
+            len(nusc_seq) == expected_len
+        ), f"expected {expected_len} frames, got {len(nusc_seq)} for {sequence_id}"
+
+    assert num_loop_iterations == len(
+        expected_lens
+    ), f"expected {len(expected_lens)} loop iterations, got {num_loop_iterations}"