Add old tiling back.

Author: newmana

stlearn/image_preprocessing/image_tiling.py

@@ -1,12 +1,104 @@
-import os
-from pathlib import Path
-
-import numpy as np
+from typing import Optional, Union
 from anndata import AnnData
 from PIL import Image
-
-# Test progress bar
+from pathlib import Path
 from tqdm import tqdm
+import numpy as np
+import os
+
+
+def old_tiling(
+    adata: AnnData,
+    out_path: Union[Path, str] = "./tiling",
+    library_id: Union[str, None] = None,
+    crop_size: int = 40,
+    target_size: int = 299,
+    img_fmt: str = "JPEG",
+    verbose: bool = False,
+    copy: bool = False,
+) -> Optional[AnnData]:
+    """\
+    Tiling H&E images to small tiles based on spot spatial location
+
+    Parameters
+    ----------
+    adata
+        Annotated data matrix.
+    out_path
+        Path to save spot image tiles
+    library_id
+        Library id stored in AnnData.
+    crop_size
+        Size of tiles
+    verbose
+        Verbose output
+    copy
+        Return a copy instead of writing to adata.
+    target_size
+        Input size for convolutional neuron network
+    Returns
+    -------
+    Depending on `copy`, returns or updates `adata` with the following fields.
+    **tile_path** : `adata.obs` field
+        Saved path for each spot image tiles
+    """
+
+    if library_id is None:
+        library_id = list(adata.uns["spatial"].keys())[0]
+
+    # Check the exist of out_path
+    if not os.path.isdir(out_path):
+        os.mkdir(out_path)
+
+    image = adata.uns["spatial"][library_id]["images"][
+        adata.uns["spatial"][library_id]["use_quality"]
+    ]
+    if image.dtype == np.float32 or image.dtype == np.float64:
+        image = (image * 255).astype(np.uint8)
+    img_pillow = Image.fromarray(image)
+
+    if img_pillow.mode == "RGBA":
+        img_pillow = img_pillow.convert("RGB")
+
+    tile_names = []
+
+    with tqdm(
+        total=len(adata),
+        desc="Tiling image",
+        bar_format="{l_bar}{bar} [ time left: {remaining} ]",
+    ) as pbar:
+        for imagerow, imagecol in zip(adata.obs["imagerow"], adata.obs["imagecol"]):
+            imagerow_down = imagerow - crop_size / 2
+            imagerow_up = imagerow + crop_size / 2
+            imagecol_left = imagecol - crop_size / 2
+            imagecol_right = imagecol + crop_size / 2
+            tile = img_pillow.crop(
+                (imagecol_left, imagerow_down, imagecol_right, imagerow_up)
+            )
+            tile.thumbnail((target_size, target_size), Image.Resampling.LANCZOS)
+            tile = tile.resize((target_size, target_size))
+            tile_name = str(imagecol) + "-" + str(imagerow) + "-" + str(crop_size)
+
+            if img_fmt == "JPEG":
+                out_tile = Path(out_path) / (tile_name + ".jpeg")
+                tile_names.append(str(out_tile))
+                tile.save(out_tile, "JPEG")
+            else:
+                out_tile = Path(out_path) / (tile_name + ".png")
+                tile_names.append(str(out_tile))
+                tile.save(out_tile, "PNG")
+
+            if verbose:
+                print(
+                    "generate tile at location ({}, {})".format(
+                        str(imagecol), str(imagerow)
+                    )
+                )
+
+            pbar.update(1)
+
+    adata.obs["tile_path"] = tile_names
+    return adata if copy else None
 
 
 def tiling(

stlearn/pp.py

@@ -1,5 +1,5 @@
 from .image_preprocessing.feature_extractor import extract_feature, old_extract_feature
-from .image_preprocessing.image_tiling import tiling
+from .image_preprocessing.image_tiling import tiling, old_tiling
 from .preprocessing.filter_genes import filter_genes
 from .preprocessing.graph import neighbors
 from .preprocessing.log_scale import log1p, scale
@@ -12,6 +12,7 @@
     "scale",
     "neighbors",
     "tiling",
+    "old_tiling",
     "extract_feature",
     "old_extract_feature",
 ]