save tiles to .tar

.github/workflows/pr-test.yaml

@@ -26,8 +26,8 @@ jobs:
           test -d tests
           test -f tests/fixtures/input/test-wsi.tif
           test -f tests/fixtures/input/test-mask.tif
-          test -f tests/fixtures/gt/test-wsi.tiles.npz
-          test -f tests/fixtures/gt/test-wsi.tiles.meta.json
+          test -f tests/fixtures/gt/test-wsi.coordinates.npz
+          test -f tests/fixtures/gt/test-wsi.coordinates.meta.json
 
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v3

README.md

@@ -81,8 +81,8 @@ result = tile_slide(
 
 artifacts = save_tiling_result(result, output_dir=Path("output"))
 
-print(artifacts.tiles_npz_path)   # output/coordinates/slide-1.tiles.npz ; more info in docs/artifacts.md
-print(artifacts.tiles_meta_path)  # output/coordinates/slide-1.tiles.meta.json ; more info in docs/artifacts.md
+print(artifacts.coordinates_npz_path)   # output/tiles/slide-1.coordinates.npz ; more info in docs/artifacts.md
+print(artifacts.coordinates_meta_path)  # output/tiles/slide-1.coordinates.meta.json ; more info in docs/artifacts.md
 
 tiling_preview_path = write_tiling_preview(
     result=result,
@@ -155,8 +155,8 @@ More CLI details: [docs/cli.md](docs/cli.md)
 
 `hs2p` writes explicit named artifacts rather than anonymous coordinate dumps.
 
-- Tiling writes `coordinates/{sample_id}.tiles.npz` and `coordinates/{sample_id}.tiles.meta.json`
-- Sampling writes the same pair under `coordinates/<annotation>/`
+- Tiling writes `tiles/{sample_id}.coordinates.npz` and `tiles/{sample_id}.coordinates.meta.json`
+- Sampling writes the same pair under `tiles/<annotation>/`
 - Batch runs also write `process_list.csv`
 - Saved coordinate arrays use a deterministic column-major order: numeric `x` first, then numeric `y` within each shared `x`
 

docs/README.md

@@ -5,6 +5,6 @@
 - [CLI guide](cli.md)
   - Input CSV schema, batch commands, and current config layout
 - [Artifact reference](artifacts.md)
-  - `.tiles.npz`, `.tiles.meta.json`, and `process_list.csv`
+  - `.coordinates.npz`, `.coordinates.meta.json`, and `process_list.csv`
 - [Tissue mask generation](tissue-mask-generation.md)
   - Standalone script for producing pyramidal tissue masks

docs/api.md

@@ -28,7 +28,7 @@ The config dataclasses keep the main knobs explicit and fill secondary options f
 - `TilingResult`
   - In-memory tile coordinates plus read-level metadata for one slide
 - `TilingArtifacts`
-  - Paths to the saved `.tiles.npz` and `.tiles.meta.json` outputs
+  - Paths to the saved `.coordinates.npz` and `.coordinates.meta.json` outputs
   - Can also carry preview-image paths when batch QC is enabled in `tile_slides()`
 
 For field-by-field details, see the dataclass docstrings in [hs2p/api.py](../hs2p/api.py).
@@ -185,8 +185,8 @@ You can round-trip a saved result later:
 from hs2p import load_tiling_result
 
 loaded = load_tiling_result(
-    tiles_npz_path=artifacts.tiles_npz_path,
-    tiles_meta_path=artifacts.tiles_meta_path,
+    coordinates_npz_path=artifacts.coordinates_npz_path,
+    coordinates_meta_path=artifacts.coordinates_meta_path,
 )
 ```
 

docs/artifacts.md

@@ -4,15 +4,15 @@ HS2P writes explicit named artifacts for both the Python API and the CLI.
 
 ## Coordinate artifact locations
 
-- Tiling writes one pair per slide under `coordinates/`
-- Sampling writes one pair per annotation under `coordinates/<annotation>/`
+- Tiling writes one pair per slide under `tiles/`
+- Sampling writes one pair per annotation under `tiles/<annotation>/`
 
 Each successful output produces:
 
-- `{sample_id}.tiles.npz`
-- `{sample_id}.tiles.meta.json`
+- `{sample_id}.coordinates.npz`
+- `{sample_id}.coordinates.meta.json`
 
-## `.tiles.npz`
+## `.coordinates.npz`
 
 - `tile_index`
   - Contiguous tile ids from `0` to `num_tiles - 1`
@@ -25,7 +25,7 @@ Each successful output produces:
 - `tissue_fraction`
   - Optional per-tile tissue coverage measured during extraction
 
-## `.tiles.meta.json`
+## `.coordinates.meta.json`
 
 - `sample_id`
   - Sample identifier that produced the artifact and names the files
@@ -69,8 +69,8 @@ Tiling columns:
 - `mask_path`
 - `tiling_status`
 - `num_tiles`
-- `tiles_npz_path`
-- `tiles_meta_path`
+- `coordinates_npz_path`
+- `coordinates_meta_path`
 - `error`
 - `traceback`
 

docs/cli.md

@@ -71,8 +71,8 @@ Detailed logs still go to `output_dir/logs/log.txt`, which is the best place to
 
 ## Current config areas
 
-- `tiling.read_tiles_from`
-  - Optional directory containing precomputed `{sample_id}.tiles.npz` and `{sample_id}.tiles.meta.json`
+- `tiling.read_coordinates_from`
+  - Optional directory containing precomputed `{sample_id}.coordinates.npz` and `{sample_id}.coordinates.meta.json`
 - `tiling.params`
   - Core tiling resolution, tile size, overlap, and tissue-threshold settings
 - `tiling.seg_params`
@@ -124,6 +124,6 @@ When enabled, every candidate tile that passes the tissue mask check is read fro
 
 - `resume: true` expects the current `process_list.csv` schema and current-format artifacts
 - reused tiling artifacts are validated against `sample_id`, `config_hash`, `image_path`, and `mask_path`
-- `tiling.read_tiles_from` is the supported way to reuse precomputed tiling outputs
+- `tiling.read_coordinates_from` is the supported way to reuse precomputed tiling outputs
 
 For the exact output files and field meanings, see [artifacts.md](artifacts.md).