From 26f0b748a8d4042cf53c94fcdc106729f8a2ee92 Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 11:40:07 -0500
Subject: [PATCH 01/15] rm unused args, fix #185

---
 wsinfer/modellib/data.py | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/wsinfer/modellib/data.py b/wsinfer/modellib/data.py
index c5393bd..4c0c7d5 100644
--- a/wsinfer/modellib/data.py
+++ b/wsinfer/modellib/data.py
@@ -60,10 +60,6 @@ class WholeSlideImagePatches(torch.utils.data.Dataset):
         Path to whole slide image file.
     patch_path : str, Path
         Path to npy file with coordinates of input image.
-    um_px : float
-        Scale of the resulting patches. For example, 0.5 for ~20x magnification.
-    patch_size : int
-        The size of patches in pixels.
     transform : callable, optional
         A callable to modify a retrieved patch. The callable must accept a
         PIL.Image.Image instance and return a torch.Tensor.
@@ -73,14 +69,10 @@ def __init__(
         self,
         wsi_path: str | Path,
         patch_path: str | Path,
-        um_px: float,
-        patch_size: int,
         transform: Callable[[Image.Image], torch.Tensor] | None = None,
     ):
         self.wsi_path = wsi_path
         self.patch_path = patch_path
-        self.um_px = float(um_px)
-        self.patch_size = int(patch_size)
         self.transform = transform
 
         assert Path(wsi_path).exists(), "wsi path not found"

From 51029b29ae0f549feb9c04d2eaf08ea9f0c766b7 Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 11:58:49 -0500
Subject: [PATCH 02/15] account for empty hierarchy or patches

fix #195
---
 wsinfer/patchlib/__init__.py | 28 +++++++++++++++++-----------
 wsinfer/patchlib/patch.py    |  6 ++++--
 2 files changed, 21 insertions(+), 13 deletions(-)

diff --git a/wsinfer/patchlib/__init__.py b/wsinfer/patchlib/__init__.py
index 4aa0c25..ba73018 100644
--- a/wsinfer/patchlib/__init__.py
+++ b/wsinfer/patchlib/__init__.py
@@ -158,9 +158,11 @@ def segment_and_patch_one_slide(
         slide.dimensions[0] / thumb.size[0],
         slide.dimensions[1] / thumb.size[1],
     )
-    polygon, contours, hierarchy = get_multipolygon_from_binary_arr(
-        arr.astype("uint8") * 255, scale=scale
-    )
+    _res = get_multipolygon_from_binary_arr(arr.astype("uint8") * 255, scale=scale)
+    if _res is None:
+        logger.warning(f"No tissue was found in slide {slide_path}")
+        return None
+    polygon, contours, hierarchy = _res
 
     # Get the coordinates of patches inside the tissue polygon.
     slide_width, slide_height = slide.dimensions
@@ -176,15 +178,19 @@ def segment_and_patch_one_slide(
     )
     logger.info(f"Found {len(coords)} patches within tissue")
 
-    # Save coordinates to HDF5.
+    # Save coordinates to HDF5, if at least one patch was found.
     patch_path.parent.mkdir(exist_ok=True, parents=True)
-    save_hdf5(
-        path=patch_path,
-        coords=coords,
-        patch_size=patch_size,
-        patch_spacing_um_px=patch_spacing_um_px,
-        compression="gzip",
-    )
+    if coords.size > 0:
+        logger.info(f"Writing patches to {patch_path}")
+        save_hdf5(
+            path=patch_path,
+            coords=coords,
+            patch_size=patch_size,
+            patch_spacing_um_px=patch_spacing_um_px,
+            compression="gzip",
+        )
+    else:
+        logger.warning(f"No patches found for slide {slide_path}")
 
     # Save thumbnail with drawn contours.
     logger.info(f"Writing tissue thumbnail with contours to disk: {mask_path}")
diff --git a/wsinfer/patchlib/patch.py b/wsinfer/patchlib/patch.py
index dcacee2..4cfe038 100644
--- a/wsinfer/patchlib/patch.py
+++ b/wsinfer/patchlib/patch.py
@@ -32,7 +32,7 @@ def temporary_recursion_limit(limit: int) -> Iterator[None]:
 
 def get_multipolygon_from_binary_arr(
     arr: npt.NDArray[np.int_], scale: tuple[float, float] | None = None
-) -> tuple[MultiPolygon, Sequence[npt.NDArray[np.int_]], npt.NDArray[np.int_]]:
+) -> tuple[MultiPolygon, Sequence[npt.NDArray[np.int_]], npt.NDArray[np.int_]] | None:
     """Create a Shapely Polygon from a binary array.
 
     Parameters
@@ -54,8 +54,10 @@ def get_multipolygon_from_binary_arr(
     """
     # Find contours and hierarchy
     contours: Sequence[npt.NDArray]
-    hierarchy: npt.NDArray
+    hierarchy: npt.NDArray | None
     contours, hierarchy = cv.findContours(arr, cv.RETR_CCOMP, cv.CHAIN_APPROX_SIMPLE)
+    if hierarchy is None:
+        return None
     hierarchy = hierarchy.squeeze(0)
 
     logger.info(f"Detected {len(contours)} contours")

From 4dc0117cd13928ac27882159c34203269ef28434 Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 12:04:17 -0500
Subject: [PATCH 03/15] please mypy

---
 wsinfer/patchlib/patch.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/wsinfer/patchlib/patch.py b/wsinfer/patchlib/patch.py
index 4cfe038..5d2687d 100644
--- a/wsinfer/patchlib/patch.py
+++ b/wsinfer/patchlib/patch.py
@@ -100,6 +100,8 @@ def merge_polygons(polygon: MultiPolygon, idx: int, add: bool) -> MultiPolygon:
                 polygon = polygon.difference(new_poly)
 
         # Check if current polygon has a child
+        if hierarchy is None:
+            raise NotImplementedError()
         child_idx = hierarchy[idx][2]
         if child_idx >= 0:
             # Call this function recursively, negate `add` parameter

From 54f40d91185e8a5425926161521329eae473de9e Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 12:04:28 -0500
Subject: [PATCH 04/15] make error message more helpful if patch dir does not
 exist

---
 wsinfer/modellib/run_inference.py | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/wsinfer/modellib/run_inference.py b/wsinfer/modellib/run_inference.py
index 4c0aec4..1046a04 100644
--- a/wsinfer/modellib/run_inference.py
+++ b/wsinfer/modellib/run_inference.py
@@ -83,7 +83,12 @@ def run_inference(
     # Check patches directory.
     patch_dir = results_dir / "patches"
     if not patch_dir.exists():
-        raise errors.PatchDirectoryNotFound("Results dir must include 'patches' dir")
+        raise errors.PatchDirectoryNotFound(
+            "The 'patches' directory was not found in results directory. This can"
+            " happen for a few reasons: 1) no tissue was detected in the slides,"
+            " 2) the physical spacing (MPP) could not be read from any of the slides"
+            ", or 3) something else... Please read the logs above for potential errors."
+        )
     # Create the patch paths based on the whole slide image paths. In effect, only
     # create patch paths if the whole slide image patch exists.
     patch_paths = [patch_dir / p.with_suffix(".h5").name for p in wsi_paths]

From 78086a80cc10f898fa50f78d4ff0bec5adafe59f Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 12:14:11 -0500
Subject: [PATCH 05/15] rm unused args in wsi patches dataset

fixes #185
---
 wsinfer/modellib/run_inference.py | 2 --
 1 file changed, 2 deletions(-)

diff --git a/wsinfer/modellib/run_inference.py b/wsinfer/modellib/run_inference.py
index 1046a04..7ba8aa5 100644
--- a/wsinfer/modellib/run_inference.py
+++ b/wsinfer/modellib/run_inference.py
@@ -150,8 +150,6 @@ def run_inference(
             dset = WholeSlideImagePatches(
                 wsi_path=wsi_path,
                 patch_path=patch_path,
-                um_px=model_info.config.spacing_um_px,
-                patch_size=model_info.config.patch_size_pixels,
                 transform=transform,
             )
         except Exception:

From db8c4c717517e0f21814064a56fc015cdd452e86 Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 12:14:36 -0500
Subject: [PATCH 06/15] check that there are patches in the loaded file

fixes #195
---
 wsinfer/modellib/data.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/wsinfer/modellib/data.py b/wsinfer/modellib/data.py
index 4c0c7d5..b0d9ed7 100644
--- a/wsinfer/modellib/data.py
+++ b/wsinfer/modellib/data.py
@@ -79,7 +79,8 @@ def __init__(
         assert Path(patch_path).exists(), "patch path not found"
 
         self.patches = _read_patch_coords(self.patch_path)
-
+        if self.patches.size == 0:
+            raise ValueError(f"No patches were found in {self.patch_path}")
         assert self.patches.ndim == 2, "expected 2D array of patch coordinates"
         # x, y, width, height
         assert self.patches.shape[1] == 4, "expected second dimension to have len 4"

From 6e81da708c76d5bbe9158309ff50dd9796a8b6fa Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 12:16:12 -0500
Subject: [PATCH 07/15] ensure thumbnail is RGB

fixes #216
---
 wsinfer/patchlib/__init__.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/wsinfer/patchlib/__init__.py b/wsinfer/patchlib/__init__.py
index ba73018..38776fd 100644
--- a/wsinfer/patchlib/__init__.py
+++ b/wsinfer/patchlib/__init__.py
@@ -121,7 +121,9 @@ def segment_and_patch_one_slide(
     if len(thumbsize) != 2:
         raise ValueError(f"Length of 'thumbsize' must be 2 but got {len(thumbsize)}")
     thumb: Image.Image = slide.get_thumbnail(thumbsize)
-    # TODO: allow the min hole size and min object size to be set in physical units.
+    if thumb.mode != "RGB":
+        logger.warning(f"Converting mode of thumbnail from {thumb.mode} to RGB")
+        thumb = thumb.convert("RGB")
 
     # thumb has ~12 MPP.
     thumb_mpp = (mpp * (np.array(slide.dimensions) / thumb.size)).mean()

From 064ba854265a667af7ec36c2b803e227541a3b5b Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 13:15:43 -0500
Subject: [PATCH 08/15] appease mypy

---
 tests/test_all.py | 26 ++++++++++++++------------
 1 file changed, 14 insertions(+), 12 deletions(-)

diff --git a/tests/test_all.py b/tests/test_all.py
index 551b9b2..f79eb21 100644
--- a/tests/test_all.py
+++ b/tests/test_all.py
@@ -6,6 +6,7 @@
 import sys
 import time
 from pathlib import Path
+from unittest.mock import patch as mock_patch, MagicMock
 
 import geojson as geojsonlib
 import h5py
@@ -24,6 +25,7 @@
 from wsinfer.wsi import HAS_OPENSLIDE
 from wsinfer.wsi import HAS_TIFFSLIDE
 
+
 @pytest.fixture
 def tiff_image(tmp_path: Path) -> Path:
     x = np.empty((4096, 4096, 3), dtype="uint8")
@@ -82,7 +84,7 @@ def test_cli_run_with_registered_models(
     backend: str,
     tiff_image: Path,
     tmp_path: Path,
-):
+) -> None:
     """A regression test of the command 'wsinfer run'."""
 
     reference_csv = Path(__file__).parent / "reference" / model / "purple.csv"
@@ -151,7 +153,7 @@ def test_cli_run_with_registered_models(
 
     for geojson_row in d["features"]:
         assert geojson_row["type"] == "Feature"
-        isinstance(geojson_row["id"] , str)
+        isinstance(geojson_row["id"], str)
         assert geojson_row["geometry"]["type"] == "Polygon"
     res = []
     for i, prob_col in enumerate(prob_cols):
@@ -178,7 +180,7 @@ def test_cli_run_with_registered_models(
         assert [df_coords] == geojson_row["geometry"]["coordinates"]
 
 
-def test_cli_run_with_local_model(tmp_path: Path, tiff_image: Path):
+def test_cli_run_with_local_model(tmp_path: Path, tiff_image: Path) -> None:
     model = "breast-tumor-resnet34.tcga-brca"
     reference_csv = Path(__file__).parent / "reference" / model / "purple.csv"
     if not reference_csv.exists():
@@ -246,7 +248,7 @@ def test_cli_run_with_local_model(tmp_path: Path, tiff_image: Path):
         ), f"Column {prob_col} not allclose at atol=1e-07"
 
 
-def test_cli_run_no_model_or_config(tmp_path: Path):
+def test_cli_run_no_model_or_config(tmp_path: Path) -> None:
     """Test that --model or (--config and --model-path) is required."""
     wsi_dir = tmp_path / "slides"
     wsi_dir.mkdir()
@@ -265,7 +267,7 @@ def test_cli_run_no_model_or_config(tmp_path: Path):
     assert "one of --model or (--config and --model-path) is required" in result.output
 
 
-def test_cli_run_model_and_config(tmp_path: Path):
+def test_cli_run_model_and_config(tmp_path: Path) -> None:
     """Test that (model and weights) or config is required."""
     wsi_dir = tmp_path / "slides"
     wsi_dir.mkdir()
@@ -298,7 +300,7 @@ def test_cli_run_model_and_config(tmp_path: Path):
 
 
 @pytest.mark.xfail
-def test_convert_to_sbu():
+def test_convert_to_sbu() -> None:
     # TODO: create a synthetic output and then convert it. Check that it is valid.
     assert False
 
@@ -330,7 +332,7 @@ def test_patch_cli(
     backend: str,
     tmp_path: Path,
     tiff_image: Path,
-):
+) -> None:
     """Test of 'wsinfer patch'."""
     orig_slide_size = 4096
     orig_slide_spacing = 0.25
@@ -380,7 +382,7 @@ def test_patch_cli(
 
 
 # FIXME: parametrize this test across our models.
-def test_jit_compile():
+def test_jit_compile() -> None:
     w = get_registered_model("breast-tumor-resnet34.tcga-brca")
     model = get_pretrained_torch_module(w)
 
@@ -411,7 +413,7 @@ def test_jit_compile():
         )
 
 
-def test_issue_89():
+def test_issue_89() -> None:
     """Do not fail if 'git' is not installed."""
     model_obj = get_registered_model("breast-tumor-resnet34.tcga-brca")
     d = _get_info_for_save(model_obj)
@@ -433,7 +435,7 @@ def test_issue_89():
         os.environ["PATH"] = orig_path  # reset path
 
 
-def test_issue_94(tmp_path: Path, tiff_image: Path):
+def test_issue_94(tmp_path: Path, tiff_image: Path) -> None:
     """Gracefully handle unreadable slides."""
 
     # We have a valid tiff in 'tiff_image.parent'. We put in an unreadable file too.
@@ -461,7 +463,7 @@ def test_issue_94(tmp_path: Path, tiff_image: Path):
     assert not results_dir.joinpath("model-outputs-csv").joinpath("bad.csv").exists()
 
 
-def test_issue_97(tmp_path: Path, tiff_image: Path):
+def test_issue_97(tmp_path: Path, tiff_image: Path) -> None:
     """Write a run_metadata file per run."""
 
     runner = CliRunner()
@@ -502,7 +504,7 @@ def test_issue_97(tmp_path: Path, tiff_image: Path):
     assert len(metas) == 2
 
 
-def test_issue_125(tmp_path: Path):
+def test_issue_125(tmp_path: Path) -> None:
     """Test that path in model config can be saved when a pathlib.Path object."""
 
     w = get_registered_model("breast-tumor-resnet34.tcga-brca")

From 06accf76591eae264b191351b43b9e27adaac514 Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 13:16:10 -0500
Subject: [PATCH 09/15] test that openslide and tiffslide pad regions

fixes #203
---
 tests/test_all.py | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/tests/test_all.py b/tests/test_all.py
index f79eb21..fd93b24 100644
--- a/tests/test_all.py
+++ b/tests/test_all.py
@@ -512,3 +512,24 @@ def test_issue_125(tmp_path: Path) -> None:
     info = _get_info_for_save(w)
     with open(tmp_path / "foo.json", "w") as f:
         json.dump(info, f)
+
+
+def test_issue_203(tiff_image: Path) -> None:
+    """Test that openslide and tiffslide pad an image if an out-of-bounds region
+    is requested.
+    """
+    import openslide
+    import tiffslide
+
+    with tiffslide.TiffSlide(tiff_image) as tslide:
+        w, h = tslide.dimensions
+        img = tslide.read_region((w, h), level=0, size=(256, 256))
+        assert img.size == (256, 256)
+        assert np.allclose(np.array(img), 0)
+    del tslide, img
+
+    with openslide.OpenSlide(tiff_image) as oslide:
+        w, h = oslide.dimensions
+        img = oslide.read_region((w, h), level=0, size=(256, 256))
+        assert img.size == (256, 256)
+        assert np.allclose(np.array(img), 0)

From 73dd4eaaa7a78c34575544576c7e21fdbed6433a Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 13:16:40 -0500
Subject: [PATCH 10/15] rm unused import

---
 tests/test_all.py | 1 -
 1 file changed, 1 deletion(-)

diff --git a/tests/test_all.py b/tests/test_all.py
index fd93b24..8d175be 100644
--- a/tests/test_all.py
+++ b/tests/test_all.py
@@ -6,7 +6,6 @@
 import sys
 import time
 from pathlib import Path
-from unittest.mock import patch as mock_patch, MagicMock
 
 import geojson as geojsonlib
 import h5py

From 95e3b0609ea42d5b078150d23bc71624dac9e1bd Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 13:25:09 -0500
Subject: [PATCH 11/15] handle symlinked slides dirs

fixes #214
---
 tests/test_all.py                    | 25 +++++++++++++++++
 wsinfer/cli/convert_csv_to_sbubmi.py | 10 +++----
 wsinfer/cli/infer.py                 | 11 +++-----
 wsinfer/cli/patch.py                 | 40 ++++++++++++++--------------
 wsinfer/patchlib/__init__.py         |  4 +--
 5 files changed, 54 insertions(+), 36 deletions(-)

diff --git a/tests/test_all.py b/tests/test_all.py
index 8d175be..a6c47d0 100644
--- a/tests/test_all.py
+++ b/tests/test_all.py
@@ -532,3 +532,28 @@ def test_issue_203(tiff_image: Path) -> None:
         img = oslide.read_region((w, h), level=0, size=(256, 256))
         assert img.size == (256, 256)
         assert np.allclose(np.array(img), 0)
+
+
+def test_issue_214(tmp_path: Path, tiff_image: Path) -> None:
+    """Test that symlinked slides don't mess things up."""
+    link = tmp_path / "forlinks" / "arbitrary-link-name.tiff"
+    link.parent.mkdir(parents=True)
+    link.symlink_to(tiff_image)
+
+    runner = CliRunner()
+    results_dir = tmp_path / "inference"
+    result = runner.invoke(
+        cli,
+        [
+            "run",
+            "--wsi-dir",
+            str(link.parent),
+            "--results-dir",
+            str(results_dir),
+            "--model",
+            "breast-tumor-resnet34.tcga-brca",
+        ],
+    )
+    assert result.exit_code == 0
+    assert (results_dir / "patches" / link.with_suffix(".h5").name).exists()
+    assert (results_dir / "model-outputs-csv" / link.with_suffix(".csv").name).exists()
diff --git a/wsinfer/cli/convert_csv_to_sbubmi.py b/wsinfer/cli/convert_csv_to_sbubmi.py
index d55df94..091aaef 100644
--- a/wsinfer/cli/convert_csv_to_sbubmi.py
+++ b/wsinfer/cli/convert_csv_to_sbubmi.py
@@ -249,21 +249,17 @@ def get_color(row: pd.Series) -> tuple[float, float, float]:
 @click.command()
 @click.argument(
     "results_dir",
-    type=click.Path(
-        exists=True, file_okay=False, dir_okay=True, path_type=Path, resolve_path=True
-    ),
+    type=click.Path(exists=True, file_okay=False, dir_okay=True, path_type=Path),
 )
 @click.argument(
     "output",
-    type=click.Path(exists=False, path_type=Path, resolve_path=True),
+    type=click.Path(exists=False, path_type=Path),
 )
 @click.option(
     "--wsi-dir",
     required=True,
     help="Directory with whole slide images.",
-    type=click.Path(
-        exists=True, file_okay=False, dir_okay=True, path_type=Path, resolve_path=True
-    ),
+    type=click.Path(exists=True, file_okay=False, dir_okay=True, path_type=Path),
 )
 @click.option("--execution-id", required=True, help="Unique execution ID for this run.")
 @click.option("--study-id", required=True, help="Study ID, like TCGA-BRCA.")
diff --git a/wsinfer/cli/infer.py b/wsinfer/cli/infer.py
index 639452b..5c03c2e 100644
--- a/wsinfer/cli/infer.py
+++ b/wsinfer/cli/infer.py
@@ -188,7 +188,7 @@ def get_stdout(args: list[str]) -> str:
 @click.option(
     "-i",
     "--wsi-dir",
-    type=click.Path(exists=True, file_okay=False, path_type=Path, resolve_path=True),
+    type=click.Path(exists=True, file_okay=False, path_type=Path),
     required=True,
     help="Directory containing whole slide images. This directory can *only* contain"
     " whole slide images.",
@@ -196,7 +196,7 @@ def get_stdout(args: list[str]) -> str:
 @click.option(
     "-o",
     "--results-dir",
-    type=click.Path(file_okay=False, path_type=Path, resolve_path=True),
+    type=click.Path(file_okay=False, path_type=Path),
     required=True,
     help="Directory to store results. If directory exists, will skip"
     " whole slides for which outputs exist.",
@@ -212,7 +212,7 @@ def get_stdout(args: list[str]) -> str:
 @click.option(
     "-c",
     "--config",
-    type=click.Path(exists=True, dir_okay=False, path_type=Path, resolve_path=True),
+    type=click.Path(exists=True, dir_okay=False, path_type=Path),
     help=(
         "Path to configuration for the trained model. Use this option if the"
         " model weights are not registered in wsinfer. Mutually exclusive with"
@@ -222,7 +222,7 @@ def get_stdout(args: list[str]) -> str:
 @click.option(
     "-p",
     "--model-path",
-    type=click.Path(exists=True, dir_okay=False, path_type=Path, resolve_path=True),
+    type=click.Path(exists=True, dir_okay=False, path_type=Path),
     help=(
         "Path to the pretrained model. Use only when --config is passed. Mutually "
         "exclusive with --model."
@@ -349,9 +349,6 @@ def run(
             "--config and --model-path must both be set if one is set."
         )
 
-    wsi_dir = wsi_dir.resolve()
-    results_dir = results_dir.resolve()
-
     if not wsi_dir.exists():
         raise FileNotFoundError(f"Whole slide image directory not found: {wsi_dir}")
 
diff --git a/wsinfer/cli/patch.py b/wsinfer/cli/patch.py
index 56d9983..d914bf3 100644
--- a/wsinfer/cli/patch.py
+++ b/wsinfer/cli/patch.py
@@ -11,7 +11,7 @@
 @click.option(
     "-i",
     "--wsi-dir",
-    type=click.Path(exists=True, file_okay=False, path_type=Path, resolve_path=True),
+    type=click.Path(exists=True, file_okay=False, path_type=Path),
     required=True,
     help="Directory containing whole slide images. This directory can *only* contain"
     " whole slide images.",
@@ -19,7 +19,7 @@
 @click.option(
     "-o",
     "--results-dir",
-    type=click.Path(file_okay=False, path_type=Path, resolve_path=True),
+    type=click.Path(file_okay=False, path_type=Path),
     required=True,
     help="Directory to store patch results. If directory exists, will skip"
     " whole slides for which outputs exist.",
@@ -32,7 +32,7 @@
     help="Physical spacing of the patch in micrometers per pixel.",
 )
 @click.option(
-    "--thumbsize",
+    "--seg-thumbsize",
     default=(2048, 2048),
     type=(int, int),
     help="The size of the slide thumbnail (in pixels) used for tissue segmentation."
@@ -40,25 +40,25 @@
     " max(thumbsize).",
 )
 @click.option(
-    "--median-filter-size",
+    "--seg-median-filter-size",
     default=7,
     type=click.IntRange(min=3),
     help="The kernel size for median filtering. Must be greater than 1 and odd.",
 )
 @click.option(
-    "--binary-threshold",
+    "--seg-binary-threshold",
     default=7,
     type=click.IntRange(min=1),
     help="The threshold for image binarization.",
 )
 @click.option(
-    "--closing-kernel-size",
+    "--seg-closing-kernel-size",
     default=6,
     type=click.IntRange(min=1),
     help="The kernel size for binary closing (morphological operation).",
 )
 @click.option(
-    "--min-object-size-um2",
+    "--seg-min-object-size-um2",
     default=200**2,
     type=click.FloatRange(min=0),
     help="The minimum size of an object to keep during tissue detection. If a"
@@ -66,7 +66,7 @@
     " The default is 200um x 200um. The units of this argument are microns squared.",
 )
 @click.option(
-    "--min-hole-size-um2",
+    "--seg-min-hole-size-um2",
     default=190**2,
     type=click.FloatRange(min=0),
     help="The minimum size of a hole to keep as a hole. If a hole is smaller than this"
@@ -78,12 +78,12 @@ def patch(
     results_dir: str,
     patch_size_px: int,
     patch_spacing_um_px: float,
-    thumbsize: tuple[int, int],
-    median_filter_size: int,
-    binary_threshold: int,
-    closing_kernel_size: int,
-    min_object_size_um2: float,
-    min_hole_size_um2: float,
+    seg_thumbsize: tuple[int, int],
+    seg_median_filter_size: int,
+    seg_binary_threshold: int,
+    seg_closing_kernel_size: int,
+    seg_min_object_size_um2: float,
+    seg_min_hole_size_um2: float,
 ) -> None:
     """Patch a directory of whole slide iamges."""
     segment_and_patch_directory_of_slides(
@@ -91,10 +91,10 @@ def patch(
         save_dir=results_dir,
         patch_size_px=patch_size_px,
         patch_spacing_um_px=patch_spacing_um_px,
-        thumbsize=thumbsize,
-        median_filter_size=median_filter_size,
-        binary_threshold=binary_threshold,
-        closing_kernel_size=closing_kernel_size,
-        min_object_size_um2=min_object_size_um2,
-        min_hole_size_um2=min_hole_size_um2,
+        thumbsize=seg_thumbsize,
+        median_filter_size=seg_median_filter_size,
+        binary_threshold=seg_binary_threshold,
+        closing_kernel_size=seg_closing_kernel_size,
+        min_object_size_um2=seg_min_object_size_um2,
+        min_hole_size_um2=seg_min_hole_size_um2,
     )
diff --git a/wsinfer/patchlib/__init__.py b/wsinfer/patchlib/__init__.py
index 38776fd..10b7989 100644
--- a/wsinfer/patchlib/__init__.py
+++ b/wsinfer/patchlib/__init__.py
@@ -85,8 +85,8 @@ def segment_and_patch_one_slide(
     None
     """
 
-    save_dir = Path(save_dir).resolve()
-    slide_path = Path(slide_path).resolve()
+    save_dir = Path(save_dir)
+    slide_path = Path(slide_path)
     slide_prefix = slide_path.stem
 
     logger.info(f"Segmenting and patching slide {slide_path}")

From a371d2732d8621da7d46055d3d158c609aa2c4cd Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 13:50:57 -0500
Subject: [PATCH 12/15] add strided patches

fixes #202

This adds a command line option '--patch-overlap-ratio' that controls
the level of overlap between adjacent patches. Negative values create
space between patches, and values closer to 1 makes patches overlap
more.
---
 wsinfer/cli/infer.py         | 12 ++++++++++++
 wsinfer/patchlib/__init__.py |  8 ++++++--
 wsinfer/patchlib/patch.py    | 20 +++++++++++++++++---
 3 files changed, 35 insertions(+), 5 deletions(-)

diff --git a/wsinfer/cli/infer.py b/wsinfer/cli/infer.py
index 5c03c2e..705ce0c 100644
--- a/wsinfer/cli/infer.py
+++ b/wsinfer/cli/infer.py
@@ -303,6 +303,16 @@ def get_stdout(args: list[str]) -> str:
     " area, it is filled with foreground. The default is 190um x 190um. The units of"
     " this argument are microns squared.",
 )
+@click.option(
+    "--patch-overlap-ratio",
+    default=0.0,
+    type=click.FloatRange(min=None, max=1, max_open=True),
+    help="The ratio of overlap among patches. The default value of 0 produces"
+    " non-overlapping patches. A value in (0, 1) will produce overlapping patches."
+    " Negative values will add space between patches. A value of -1 would skip"
+    " every other patch. A value of 0.5 will provide 50%% of overlap between patches."
+    " Values must be in (-inf, 1).",
+)
 def run(
     ctx: click.Context,
     *,
@@ -321,6 +331,7 @@ def run(
     seg_closing_kernel_size: int,
     seg_min_object_size_um2: float,
     seg_min_hole_size_um2: float,
+    patch_overlap_ratio: float = 0.0,
 ) -> None:
     """Run model inference on a directory of whole slide images.
 
@@ -398,6 +409,7 @@ def run(
         closing_kernel_size=seg_closing_kernel_size,
         min_object_size_um2=seg_min_object_size_um2,
         min_hole_size_um2=seg_min_hole_size_um2,
+        overlap=patch_overlap_ratio,
     )
 
     if not results_dir.joinpath("patches").exists():
diff --git a/wsinfer/patchlib/__init__.py b/wsinfer/patchlib/__init__.py
index 10b7989..ee15598 100644
--- a/wsinfer/patchlib/__init__.py
+++ b/wsinfer/patchlib/__init__.py
@@ -14,7 +14,7 @@
 from ..wsi import _validate_wsi_directory
 from ..wsi import get_avg_mpp
 from .patch import get_multipolygon_from_binary_arr
-from .patch import get_nonoverlapping_patch_coordinates_within_polygon
+from .patch import get_patch_coordinates_within_polygon
 from .segment import segment_tissue
 
 logger = logging.getLogger(__name__)
@@ -34,6 +34,7 @@ def segment_and_patch_one_slide(
     closing_kernel_size: int = 6,
     min_object_size_um2: float = 200**2,
     min_hole_size_um2: float = 190**2,
+    overlap: float = 0.0,
 ) -> None:
     """Get non-overlapping patch coordinates in tissue regions of a whole slide image.
 
@@ -171,12 +172,13 @@ def segment_and_patch_one_slide(
     half_patch_size = round(patch_size / 2)
 
     # Nx4 --> N x (minx, miny, width, height)
-    coords = get_nonoverlapping_patch_coordinates_within_polygon(
+    coords = get_patch_coordinates_within_polygon(
         slide_width=slide_width,
         slide_height=slide_height,
         patch_size=patch_size,
         half_patch_size=half_patch_size,
         polygon=polygon,
+        overlap=overlap,
     )
     logger.info(f"Found {len(coords)} patches within tissue")
 
@@ -299,6 +301,7 @@ def segment_and_patch_directory_of_slides(
     closing_kernel_size: int = 6,
     min_object_size_um2: float = 200**2,
     min_hole_size_um2: float = 190**2,
+    overlap: float = 0.0,
 ) -> None:
     """Get non-overlapping patch coordinates in tissue regions for a directory of whole
     slide images.
@@ -373,6 +376,7 @@ def segment_and_patch_directory_of_slides(
                 closing_kernel_size=closing_kernel_size,
                 min_object_size_um2=min_object_size_um2,
                 min_hole_size_um2=min_hole_size_um2,
+                overlap=overlap,
             )
         except Exception as e:
             logger.error(f"Failed to segment and patch slide\n{slide_path}", exc_info=e)
diff --git a/wsinfer/patchlib/patch.py b/wsinfer/patchlib/patch.py
index 5d2687d..ee8d030 100644
--- a/wsinfer/patchlib/patch.py
+++ b/wsinfer/patchlib/patch.py
@@ -128,12 +128,13 @@ def merge_polygons(polygon: MultiPolygon, idx: int, add: bool) -> MultiPolygon:
     return polygon, contours_unscaled, hierarchy[np.newaxis]
 
 
-def get_nonoverlapping_patch_coordinates_within_polygon(
+def get_patch_coordinates_within_polygon(
     slide_width: int,
     slide_height: int,
     patch_size: int,
     half_patch_size: int,
     polygon: Polygon,
+    overlap: float = 0.0,
 ) -> npt.NDArray[np.int_]:
     """Get coordinates of patches within a polygon.
 
@@ -149,6 +150,12 @@ def get_nonoverlapping_patch_coordinates_within_polygon(
         Half of the length of a patch in pixels.
     polygon : Polygon
         A shapely Polygon representing the presence of tissue.
+    overlap : float
+        The proportion of the patch_size to overlap. A value of 0.5
+        would have an overlap of 50%. A value of 0.2 would have an
+        overlap of 20%. Negative values will add space between patches.
+        A value of -1 would skip every other patch. Value must be in (-inf, 1).
+        The default value of 0.0 produces non-overlapping patches.
 
     Returns
     -------
@@ -157,12 +164,19 @@ def get_nonoverlapping_patch_coordinates_within_polygon(
         contains the coordinates of the top-left of a tile: (minx, miny).
     """
 
+    if overlap >= 1:
+        raise ValueError(f"overlap must be in (-inf, 1) but got {overlap}")
+
+    # Handle potentially overlapping slides.
+    step_size = round((1 - overlap) * patch_size)
+    logger.info(f"Patches are {patch_size} px, with step size of {step_size} px.")
+
     # Make an array of Nx2, where each row is (x, y) centroid of the patch.
     tile_centroids_arr: npt.NDArray[np.int_] = np.array(
         list(
             itertools.product(
-                range(0 + half_patch_size, slide_width, patch_size),
-                range(0 + half_patch_size, slide_height, patch_size),
+                range(0 + half_patch_size, slide_width, step_size),
+                range(0 + half_patch_size, slide_height, step_size),
             )
         )
     )

From c974d1b563683e836bdb9dd15729622886cf0c33 Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 13:57:00 -0500
Subject: [PATCH 13/15] save jsons with geojson ext

---
 wsinfer/write_geojson.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/wsinfer/write_geojson.py b/wsinfer/write_geojson.py
index ecd52bc..1ab072a 100644
--- a/wsinfer/write_geojson.py
+++ b/wsinfer/write_geojson.py
@@ -71,7 +71,7 @@ def make_geojson(csv: Path, results_dir: Path) -> None:
     if not prob_cols:
         raise KeyError("Did not find any columns with prob_ prefix.")
     geojson = _dataframe_to_geojson(df, prob_cols)
-    with open(results_dir / "model-outputs-geojson" / f"{filename}.json", "w") as f:
+    with open(results_dir / "model-outputs-geojson" / f"{filename}.geojson", "w") as f:
         json.dump(geojson, f)
 
 
@@ -95,7 +95,7 @@ def write_geojsons(csvs: list[Path], results_dir: Path, num_workers: int) -> Non
             "that contains model-outputs, masks, and patches."
         )
     if output.exists():
-        geojsons = list((results_dir / "model-outputs-geojson").glob("*.json"))
+        geojsons = list((results_dir / "model-outputs-geojson").glob("*.geojson"))
 
         # Makes a list of filenames for both geojsons and csvs
         geojson_filenames = [filename.stem for filename in geojsons]

From da971951d2ef804ba10907ab7430db6c2909c489 Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 13:57:11 -0500
Subject: [PATCH 14/15] print some logs

---
 wsinfer/cli/infer.py | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/wsinfer/cli/infer.py b/wsinfer/cli/infer.py
index 705ce0c..9b937cd 100644
--- a/wsinfer/cli/infer.py
+++ b/wsinfer/cli/infer.py
@@ -447,9 +447,12 @@ def run(
     with open(run_metadata_outpath, "w") as f:
         json.dump(run_metadata, f, indent=2)
 
-    click.secho("Finished.", fg="green")
-
+    click.echo("Writing inference results to GeoJSON files")
     csvs = list((results_dir / "model-outputs-csv").glob("*.csv"))
     write_geojsons(csvs, results_dir, num_workers)
+
     if qupath:
+        click.echo("Creating QuPath project with results")
         make_qupath_project(wsi_dir, results_dir)
+
+    click.secho("Finished.", fg="green")

From 39d8e9d79768010898b9caf4348d098d7d4d16ff Mon Sep 17 00:00:00 2001
From: kaczmarj <jakub.kaczmarzyk@gmail.com>
Date: Thu, 22 Feb 2024 14:17:08 -0500
Subject: [PATCH 15/15] use geojson ext for purple.json

---
 tests/test_all.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/test_all.py b/tests/test_all.py
index a6c47d0..b4d2b70 100644
--- a/tests/test_all.py
+++ b/tests/test_all.py
@@ -145,7 +145,7 @@ def test_cli_run_with_registered_models(
     geojson_dir = results_dir / "model-outputs-geojson"
     # result = runner.invoke(cli, ["togeojson", str(results_dir), str(geojson_dir)])
     assert result.exit_code == 0
-    with open(geojson_dir / "purple.json") as f:
+    with open(geojson_dir / "purple.geojson") as f:
         d: geojsonlib.GeoJSON = geojsonlib.load(f)
     assert d.is_valid, "geojson not valid!"
     assert len(d["features"]) == len(df_ref)