Mark slow tests to improve test execution time

docs/contributing.md

@@ -64,6 +64,12 @@ pytest
 
 in the root of the repository. Continuous integration will automatically run the tests on all pull requests.
 
+Some tests are marked as `slow` and are skipped by default on the CI. To run only the fast tests, you can run:
+
+```bash
+pytest -m "not slow"
+```
+
 ### Continuous integration
 
 Continuous integration will automatically run the tests on all pull requests and test against the minimum and maximum supported Python version.

pyproject.toml

@@ -103,7 +103,7 @@ addopts = [
 ]
 # These are all markers coming from xarray, dask or anndata. Added here to silence warnings.
 markers = [
-    "slow: marks tests as slow (deselect with '-m \"not slow\"')",
+    "slow: marks tests as slow to run",
     "gpu: run test on GPU using CuPY.",
     "skip_with_pyarrow_strings: skipwhen pyarrow string conversion is turned on",
 ]

tests/core/operations/test_rasterize.py

@@ -157,6 +157,7 @@ def test_rasterize_labels_value_key_specified():
     assert values == {True, False}, values
 
 
+@pytest.mark.slow
 def test_rasterize_points_shapes_with_string_index(points, shapes):
     sdata = SpatialData.init_from_elements({"points_0": points["points_0"], "circles": shapes["circles"]})
 
@@ -213,6 +214,7 @@ def _rasterize_shapes_prepare_data() -> tuple[SpatialData, GeoDataFrame, str]:
     return SpatialData.init_from_elements({element_name: gdf[["geometry"]], "table": adata}), gdf, element_name
 
 
+@pytest.mark.slow
 def test_rasterize_shapes():
     sdata, gdf, element_name = _rasterize_shapes_prepare_data()
 
@@ -322,6 +324,7 @@ def _rasterize(element: GeoDataFrame, **kwargs) -> SpatialImage:
     assert res[10, 37] == 2
 
 
+@pytest.mark.slow
 def test_rasterize_points():
     data = {
         "x": [0, 1, 0, 1, 2, 3, 3, 5.1],

tests/core/operations/test_vectorize.py

@@ -21,6 +21,7 @@
 
 
 # conversion from labels
+@pytest.mark.slow
 @pytest.mark.parametrize("is_multiscale", [False, True])
 def test_labels_2d_to_circles(is_multiscale: bool) -> None:
     key = "blobs" + ("_multiscale" if is_multiscale else "") + "_labels"

tests/dataloader/test_datasets.py

@@ -6,6 +6,7 @@
 
 
 class TestImageTilesDataset:
+    @pytest.mark.slow
     @pytest.mark.parametrize("image_element", ["blobs_image", "blobs_multiscale_image"])
     @pytest.mark.parametrize(
         "regions_element",
@@ -26,6 +27,7 @@ def test_validation(self, sdata_blobs, image_element: str, regions_element: str,
             return_annotations="instance_id" if table else None,
         )
 
+    @pytest.mark.slow
     @pytest.mark.parametrize(
         "regions_element",
         ["blobs_circles", "blobs_polygons", "blobs_multipolygons", "blobs_labels", "blobs_multiscale_labels"],

tests/io/test_format.py

@@ -217,6 +217,7 @@ def test_tables_v1_to_v2(self, table_multiple_annotations):
             table_read_v2 = read_zarr(f2)
             assert_spatial_data_objects_are_identical(table_multiple_annotations, table_read_v2)
 
+    @pytest.mark.slow
     def test_container_v1_to_v2(self, full_sdata):
         with tempfile.TemporaryDirectory() as tmpdir:
             f1 = Path(tmpdir) / "data1.zarr"

tests/io/test_readwrite.py

@@ -230,6 +230,7 @@ def _workaround1_dask_backed(
         del sdata[new_name]
         sdata.delete_element_from_disk(new_name)
 
+    @pytest.mark.slow
     @pytest.mark.parametrize("dask_backed", [True, False])
     @pytest.mark.parametrize("workaround", [1, 2])
     def test_incremental_io_on_disk(
@@ -546,6 +547,7 @@ def test_bug_rechunking_after_queried_raster():
         queried.write(f)
 
 
+@pytest.mark.slow
 @pytest.mark.parametrize("sdata_container_format", SDATA_FORMATS)
 def test_self_contained(full_sdata: SpatialData, sdata_container_format: SpatialDataContainerFormatType) -> None:
     # data only in-memory, so the SpatialData object and all its elements are self-contained

tests/models/test_models.py

@@ -832,6 +832,7 @@ def test_warning_on_large_chunks():
         assert "Detected chunks larger than:" in str(w[-1].message)
 
 
+@pytest.mark.slow
 def test_categories_on_partitioned_dataframe(sdata_blobs: SpatialData):
     df = sdata_blobs["blobs_points"].compute()
     df["genes"] = RNG.choice([f"gene_{i}" for i in range(200)], len(df))