slide2vec 6.3.2__tar.gz → 6.3.3__tar.gz

{slide2vec-6.3.2 → slide2vec-6.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: slide2vec
-Version: 6.3.2
+Version: 6.3.3
 Summary: Embedding of whole slide images with Foundation Models
 Author-email: Clément Grisi <clement.grisi@radboudumc.nl>
 License-Expression: Apache-2.0

{slide2vec-6.3.2 → slide2vec-6.3.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "slide2vec"
-version = "6.3.2"
+version = "6.3.3"
 description = "Embedding of whole slide images with Foundation Models"
 readme = "README.md"
 requires-python = ">=3.10"
@@ -180,7 +180,7 @@ no_implicit_reexport = true
 max-line-length = 160
 
 [tool.bumpver]
-current_version = "6.3.2"
+current_version = "6.3.3"
 version_pattern = "MAJOR.MINOR.PATCH"
 commit = false
 tag = false

{slide2vec-6.3.2 → slide2vec-6.3.3}/slide2vec/__init__.py

@@ -31,7 +31,7 @@ from slide2vec.encoders import (
 )
 
 
-__version__ = "6.3.2"
+__version__ = "6.3.3"
 
 __all__ = [
     "Model",

{slide2vec-6.3.2 → slide2vec-6.3.3}/slide2vec/inference.py

@@ -761,7 +761,10 @@ def run_pipeline(
     tiling_only: bool = False,
     execution: ExecutionOptions,
 ) -> RunResult:
-    if model.level == "patient":
+    # Tiling does not use patient ids or the patient pipeline: only a run that
+    # embeds patients needs them.
+    embeds_patients = model.level == "patient" and not tiling_only
+    if embeds_patients:
         patient_id_map = manifest.resolve_patient_id_map(slides=slides, manifest_path=manifest_path)
     else:
         patient_id_map = None
@@ -781,7 +784,7 @@ def run_pipeline(
         distributed_stage.validate_multi_gpu_execution(model, execution)
 
     resolved_preprocessing = tiling_pipeline.resolve_model_preprocessing(model, preprocessing)
-    if model.level == "patient" and not tiling_only:
+    if embeds_patients:
         patient_pipeline.reject_per_annotation_sampling(model, resolved_preprocessing)
     output_dir = Path(execution.output_dir)
     output_dir.mkdir(parents=True, exist_ok=True)

{slide2vec-6.3.2 → slide2vec-6.3.3}/slide2vec.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: slide2vec
-Version: 6.3.2
+Version: 6.3.3
 Summary: Embedding of whole slide images with Foundation Models
 Author-email: Clément Grisi <clement.grisi@radboudumc.nl>
 License-Expression: Apache-2.0

{slide2vec-6.3.2 → slide2vec-6.3.3}/slide2vec.egg-info/SOURCES.txt

@@ -157,6 +157,7 @@ tests/test_output_consistency.py
 tests/test_patch_size_metadata.py
 tests/test_patient_annotation_sampling.py
 tests/test_patient_manifest.py
+tests/test_patient_tiling_only.py
 tests/test_phaet.py
 tests/test_pooled_encoder_input.py
 tests/test_pooled_geometry.py

{slide2vec-6.3.2 → slide2vec-6.3.3}/tests/test_patient_manifest.py

@@ -60,7 +60,7 @@ def test_slide_manifest_preserves_leading_zero_sample_ids(tmp_path: Path):
     assert [slide.sample_id for slide in slides] == ["0007", "0008"]
 
 
-def test_patient_tiling_only_validates_ids_before_creating_output(tmp_path: Path):
+def test_patient_run_validates_ids_before_creating_output(tmp_path: Path):
     manifest = tmp_path / "manifest.csv"
     manifest.write_text(
         "sample_id,image_path,patient_id\n"
@@ -75,7 +75,6 @@ def test_patient_tiling_only_validates_ids_before_creating_output(tmp_path: Path
         run_pipeline(
             SimpleNamespace(level="patient"),
             manifest_path=manifest,
-            tiling_only=True,
             execution=SimpleNamespace(output_dir=output_dir, num_gpus=1),
         )
 

slide2vec-6.3.3/tests/test_patient_tiling_only.py

@@ -0,0 +1,86 @@
+"""Patient-level models and slides that carry no patient id.
+
+Tiling does not use patient ids, so a tiling-only run must not need them. An
+embedding run groups slides by patient, so it still needs one per slide. Ways this
+can fail:
+
+- a tiling-only run rejects slides without a patient id (for example hs2p
+  ``SlideSpec`` objects, which have no ``patient_id`` field);
+- a full run accepts such slides and groups or names patients wrongly;
+- a full run rejects them only after the slides were tiled or the model was loaded.
+"""
+
+from pathlib import Path
+
+import pandas as pd
+import pytest
+from hs2p import SlideSpec
+
+from slide2vec.api import ExecutionOptions, Model, Pipeline, PreprocessingConfig
+from tests.output_consistency_config import (
+    TILING_FILTER_PARAMS,
+    TILING_MASKS,
+    TILING_PARAMS,
+    TILING_SEG_PARAMS,
+)
+
+
+pytest.importorskip("openslide")
+
+FIXTURES = Path(__file__).parent / "fixtures" / "input"
+SAMPLE_IDS = ["slide-a", "slide-b"]
+
+
+def _slides() -> list[SlideSpec]:
+    return [
+        SlideSpec(
+            sample_id=sample_id,
+            image_path=FIXTURES / "test-wsi.tif",
+            mask_path=FIXTURES / "test-mask.tif",
+        )
+        for sample_id in SAMPLE_IDS
+    ]
+
+
+def _model_and_pipeline(output_dir: Path) -> tuple[Model, Pipeline]:
+    model = Model.from_preset("moozy", device="cpu")
+    pipeline = Pipeline(
+        model,
+        PreprocessingConfig(
+            backend="openslide",
+            mask_backend="openslide",
+            masks=TILING_MASKS,
+            segmentation=TILING_SEG_PARAMS,
+            filtering=TILING_FILTER_PARAMS,
+            preview={"save_mask_preview": False, "save_tiling_preview": False},
+            **TILING_PARAMS,
+        ),
+        execution=ExecutionOptions(output_dir=output_dir, num_gpus=1, num_preprocessing_workers=1),
+    )
+    return model, pipeline
+
+
+def test_tiling_only_patient_run_tiles_slides_without_patient_ids(tmp_path):
+    model, pipeline = _model_and_pipeline(tmp_path / "out")
+
+    result = pipeline.run(slides=_slides(), tiling_only=True)
+
+    rows = pd.read_csv(result.process_list_path, dtype=str, keep_default_na=False)
+    assert rows["sample_id"].tolist() == SAMPLE_IDS
+    assert rows["tiling_status"].tolist() == ["success"] * len(SAMPLE_IDS)
+    assert all(Path(path).is_file() for path in rows["coordinates_npz_path"])
+    assert result.tile_artifacts == []
+    assert result.slide_artifacts == []
+    assert result.patient_artifacts == []
+    assert model._backend is None  # tiling only: the model was never loaded
+
+
+def test_full_patient_run_rejects_slides_without_patient_ids_before_tiling(tmp_path):
+    output_dir = tmp_path / "out"
+    model, pipeline = _model_and_pipeline(output_dir)
+
+    with pytest.raises(ValueError, match="Patient-level models require a 'patient_id' for every slide"):
+        pipeline.run(slides=_slides())
+
+    assert model._backend is None  # nothing was loaded, so nothing was embedded
+    assert not output_dir.exists()  # nor tiled