rename TileMetadata to OutlierDetector

Author: jopo666

README.md

@@ -96,15 +96,15 @@ let's try it out!
 
 
 ```python
-from histoprep.utils import TileMetadata
+from histoprep.utils import OutlierDetector
 
 # Let's wrap the tile metadata with a helper class.
-metadata = TileMetadata(tile_metadata)
+detector = OutlierDetector(tile_metadata)
 # Cluster tiles based on image metrics.
-clusters = metadata.cluster_kmeans(num_clusters=4, random_state=666)
+clusters = detector.cluster_kmeans(num_clusters=4, random_state=666)
 # Visualise first cluster.
 reader.get_annotated_thumbnail(
-    image=reader.read_level(-1), coordinates=metadata.coordinates[clusters == 0]
+    image=reader.read_level(-1), coordinates=detector.coordinates[clusters == 0]
 )
 ```
 ![Tiles in cluster 0](images/thumbnail_blue.jpeg)

histoprep/utils/_process.py

@@ -1,6 +1,6 @@
 """Helper class for preprocessing tile images."""
 
-__all__ = ["TileMetadata"]
+__all__ = ["OutlierDetector"]
 
 from typing import Optional, Union
 
@@ -26,7 +26,7 @@
 XYWH_COLUMNS = ["x", "y", "w", "h"]
 
 
-class TileMetadata:
+class OutlierDetector:
     """Class for exploring tile metadata and detecting outliers."""
 
     def __init__(self, dataframe: pl.DataFrame) -> None:

scripts/usage.ipynb

@@ -9,7 +9,7 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "slide_with_ink: 100%|██████████| 390/390 [00:01<00:00, 271.79it/s]\n"
+      "slide_with_ink: 100%|██████████| 390/390 [00:01<00:00, 259.65it/s]\n"
      ]
     }
    ],
@@ -75,21 +75,21 @@
     }
    ],
    "source": [
-    "from histoprep.utils import TileMetadata\n",
+    "from histoprep.utils import OutlierDetector\n",
     "\n",
     "# Let's wrap the tile metadata with a helper class.\n",
-    "metadata = TileMetadata(tile_metadata)\n",
+    "detector = OutlierDetector(tile_metadata)\n",
     "# Cluster tiles based on image metrics.\n",
-    "clusters = metadata.cluster_kmeans(num_clusters=4, random_state=666)\n",
+    "clusters = detector.cluster_kmeans(num_clusters=4, random_state=666)\n",
     "# Visualise first cluster.\n",
     "reader.get_annotated_thumbnail(\n",
-    "    image=reader.read_level(-1), coordinates=metadata.coordinates[clusters == 0]\n",
+    "    image=reader.read_level(-1), coordinates=detector.coordinates[clusters == 0]\n",
     ")"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -98,19 +98,19 @@
     "# Copy images.\n",
     "shutil.copy(\n",
     "    \"./train_tiles/slide_with_ink/thumbnail.jpeg\",\n",
-    "    \"../docs/images/thumbnail.jpeg\",\n",
+    "    \"../images/thumbnail.jpeg\",\n",
     ")\n",
     "shutil.copy(\n",
     "    \"./train_tiles/slide_with_ink/thumbnail_tiles.jpeg\",\n",
-    "    \"../docs/images/thumbnail_tiles.jpeg\",\n",
+    "    \"../images/thumbnail_tiles.jpeg\",\n",
     ")\n",
     "shutil.copy(\n",
     "    \"./train_tiles/slide_with_ink/thumbnail_tissue.jpeg\",\n",
-    "    \"../docs/images/thumbnail_tissue.jpeg\",\n",
+    "    \"../images/thumbnail_tissue.jpeg\",\n",
     ")\n",
     "reader.get_annotated_thumbnail(\n",
-    "    image=reader.read_level(-1), coordinates=metadata.coordinates[clusters == 0]\n",
-    ").save(\"../docs/images/thumbnail_blue.jpeg\")\n",
+    "    image=reader.read_level(-1), coordinates=detector.coordinates[clusters == 0]\n",
+    ").save(\"../images/thumbnail_blue.jpeg\")\n",
     "shutil.rmtree(\"./train_tiles\")"
    ]
   }

tests/metadata_test.py

@@ -3,7 +3,7 @@
 import pytest
 
 from histoprep import SlideReader
-from histoprep.utils import TileMetadata
+from histoprep.utils import OutlierDetector
 
 from ._utils import SLIDE_PATH_JPEG, TMP_DIRECTORY, clean_temporary_directory
 
@@ -24,7 +24,7 @@ def generate_metadata(*, clean_tmp: bool = True, **kwargs) -> pl.DataFrame:
 
 
 def test_metadata_properties() -> None:
-    metadata = TileMetadata(generate_metadata())
+    metadata = OutlierDetector(generate_metadata())
     assert isinstance(metadata.dataframe, pl.DataFrame)
     assert isinstance(metadata.dataframe_without_metrics, pl.DataFrame)
     assert metadata.dataframe_without_metrics.columns == [
@@ -39,50 +39,50 @@ def test_metadata_properties() -> None:
         (0.8448732156862745, 0.7013530588235295, 0.7794474117647058),
         (0.13158384313725494, 0.1708792549019608, 0.13072776470588235),
     )
-    assert str(metadata) == "TileMetadata(num_images=100, num_outliers=0)"
+    assert str(metadata) == "OutlierDetector(num_images=100, num_outliers=0)"
 
 
 def test_metadata_index_columns() -> None:
-    metadata = TileMetadata(generate_metadata())
+    metadata = OutlierDetector(generate_metadata())
     assert isinstance(metadata["path"], np.ndarray)
     assert isinstance(metadata["red_mean"], np.ndarray)
 
 
 def test_metadata_from_parquet() -> None:
     generate_metadata(clean_tmp=False)
-    metadata = TileMetadata.from_parquet(TMP_DIRECTORY / "slide" / "*.parquet")
+    metadata = OutlierDetector.from_parquet(TMP_DIRECTORY / "slide" / "*.parquet")
     assert len(metadata.metric_columns) == 64
     clean_temporary_directory()
 
 
 def test_metadata_from_csv() -> None:
     generate_metadata(clean_tmp=False, use_csv=True)
-    metadata = TileMetadata.from_csv(TMP_DIRECTORY / "slide" / "*.csv")
+    metadata = OutlierDetector.from_csv(TMP_DIRECTORY / "slide" / "*.csv")
     assert len(metadata.metric_columns) == 64
     clean_temporary_directory()
 
 
 def test_metadata_plot_histogram() -> None:
-    metadata = TileMetadata(generate_metadata(clean_tmp=False))
+    metadata = OutlierDetector(generate_metadata(clean_tmp=False))
     metadata.plot_histogram("red_mean", num_images=0)
     metadata.plot_histogram("red_mean", num_images=12)
     clean_temporary_directory()
 
 
 def test_metadata_plot_histogram_fail() -> None:
-    metadata = TileMetadata(generate_metadata())
+    metadata = OutlierDetector(generate_metadata())
     with pytest.raises(ValueError, match="Difference between min=0.0 and max=0.0"):
         metadata.plot_histogram("black_pixels", num_images=0)
 
 
 def test_metadata_no_metrics_fail() -> None:
     dataframe = generate_metadata()
     with pytest.raises(ValueError, match="Metadata does not contain any metrics"):
-        TileMetadata(dataframe["x", "y", "w", "h", "path"])
+        OutlierDetector(dataframe["x", "y", "w", "h", "path"])
 
 
 def test_metadata_plot_collage() -> None:
-    metadata = TileMetadata(generate_metadata(clean_tmp=False))
+    metadata = OutlierDetector(generate_metadata(clean_tmp=False))
     assert metadata.random_image_collage(~metadata.outliers, num_rows=4).size == (
         1024,
         256,
@@ -97,7 +97,7 @@ def test_metadata_plot_collage() -> None:
 
 
 def test_metadata_add_outliers() -> None:
-    metadata = TileMetadata(generate_metadata())
+    metadata = OutlierDetector(generate_metadata())
     metadata.add_outliers(metadata["background"] > 0.5, desc="too high background")
     assert metadata.outliers.sum() == 27
     assert len(metadata.outlier_selections) == 1
@@ -106,6 +106,6 @@ def test_metadata_add_outliers() -> None:
 
 
 def test_metadata_cluster() -> None:
-    metadata = TileMetadata(generate_metadata())
+    metadata = OutlierDetector(generate_metadata())
     clusters = metadata.cluster_kmeans(10)
     assert len(clusters) == len(metadata)