spatialdata-io 0.1.7__tar.gz → 0.2.0__tar.gz

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/.github/release.yml

@@ -5,18 +5,24 @@ changelog:
         authors:
             - pre-commit-ci
     categories:
-        - title: Major
+        - title: Added
           labels:
-              - "release-major"
-        - title: Minor
-          labels:
-              - "release-minor"
+              - "release-added"
         - title: Changed
           labels:
               - "release-changed"
+        - title: Deprecated
+          labels:
+              - "release-deprecated"
+        - title: Removed
+          labels:
+              - "release-removed"
         - title: Fixed
           labels:
               - "release-fixed"
+        - title: Security
+          labels:
+              - "release-security"
         - title: Other Changes
           labels:
               - "*"

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/.github/workflows/prepare_test_data.yaml

@@ -1,6 +1,8 @@
 name: Prepare test data
 
 on:
+    schedule:
+        - cron: "0 0 1 * *" # run once a month to prevent artifact expiration
     workflow_dispatch:
 # uncomment and adjust the branch name if you need to add new datasets to the artifact
 #    push:
@@ -44,7 +46,7 @@ jobs:
                   done
 
             - name: Upload artifacts
-              uses: actions/upload-artifact@v3
+              uses: actions/upload-artifact@v4
               with:
                   name: data
                   path: ./data

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/.github/workflows/test.yaml

@@ -37,7 +37,7 @@ jobs:
               run: |
                   echo "::set-output name=dir::$(pip cache dir)"
             - name: Restore pip cache
-              uses: actions/cache@v2
+              uses: actions/cache@v3
               with:
                   path: ${{ steps.pip-cache-dir.outputs.dir }}
                   key: pip-${{ runner.os }}-${{ env.pythonLocation }}-${{ hashFiles('**/pyproject.toml') }}
@@ -53,18 +53,12 @@ jobs:
 
             - name: Download artifact of test data
               if: matrix.python == '3.12'
-              uses: dawidd6/action-download-artifact@v2
+              uses: dawidd6/action-download-artifact@v9
               with:
                   workflow: prepare_test_data.yaml
                   name: data
                   path: ./data
 
-            - name: List the data directory
-              if: matrix.python == '3.12'
-              run: |
-                  ls -l ./data
-                  pwd
-
             - name: Test
               env:
                   MPLBACKEND: agg

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/.pre-commit-config.yaml

@@ -7,11 +7,11 @@ default_stages:
 minimum_pre_commit_version: 2.16.0
 repos:
     - repo: https://github.com/psf/black
-      rev: 24.10.0
+      rev: 25.1.0
       hooks:
           - id: black
     - repo: https://github.com/rbubley/mirrors-prettier
-      rev: v3.4.2
+      rev: v3.5.1
       hooks:
           - id: prettier
     - repo: https://github.com/asottile/blacken-docs
@@ -19,11 +19,11 @@ repos:
       hooks:
           - id: blacken-docs
     - repo: https://github.com/PyCQA/isort
-      rev: 5.13.2
+      rev: 6.0.0
       hooks:
           - id: isort
     - repo: https://github.com/pre-commit/mirrors-mypy
-      rev: v1.14.1
+      rev: v1.15.0
       hooks:
           - id: mypy
             additional_dependencies: [numpy, types-PyYAML]
@@ -59,7 +59,7 @@ repos:
                 - --remove-unused-variable
                 - --ignore-init-module-imports
     - repo: https://github.com/PyCQA/flake8
-      rev: 7.1.1
+      rev: 7.1.2
       hooks:
           - id: flake8
             additional_dependencies:

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/LICENSE

@@ -1,6 +1,6 @@
 BSD 3-Clause License
 
-Copyright (c) 2022, scverse
+Copyright (c) 2022, scverse®
 All rights reserved.
 
 Redistribution and use in source and binary forms, with or without

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: spatialdata-io
-Version: 0.1.7
+Version: 0.2.0
 Summary: SpatialData IO for common techs
 Project-URL: Documentation, https://spatialdata-io.readthedocs.io/
 Project-URL: Source, https://github.com/scverse/spatialdata-io
@@ -9,7 +9,7 @@ Author: scverse
 Maintainer-email: scverse <scverse@scverse.scverse>
 License: BSD 3-Clause License
         
-        Copyright (c) 2022, scverse
+        Copyright (c) 2022, scverse®
         All rights reserved.
         
         Redistribution and use in source and binary forms, with or without
@@ -181,7 +181,20 @@ Marconato, L., Palla, G., Yamauchi, K.A. et al. SpatialData: an open and univers
 
 [scverse-discourse]: https://discourse.scverse.org/
 [issue-tracker]: https://github.com/scverse/spatialdata-io/issues
-[changelog]: https://spatialdata.scverse.org/projects/io/en/latest/changelog.html
-[link-docs]: https://spatialdata.scverse.org/projects/io/en/latest/
-[link-api]: https://spatialdata.scverse.org/projects/io/en/latest/api.html
-[link-cli]: https://spatialdata.scverse.org/projects/io/en/latest/cli.html
+[changelog]: https://spatialdata.scverse.org/projects/io/en/stable/changelog.html
+[link-docs]: https://spatialdata.scverse.org/projects/io/en/stable/
+[link-api]: https://spatialdata.scverse.org/projects/io/en/stable/api.html
+[link-cli]: https://spatialdata.scverse.org/projects/io/en/stable/cli.html
+[//]: # "numfocus-fiscal-sponsor-attribution"
+
+spatialdata-io is part of the scverse® project ([website](https://scverse.org), [governance](https://scverse.org/about/roles)) and is fiscally sponsored by [NumFOCUS](https://numfocus.org/).
+If you like scverse® and want to support our mission, please consider making a tax-deductible [donation](https://numfocus.org/donate-to-scverse) to help the project pay for developer time, professional services, travel, workshops, and a variety of other needs.
+
+<div align="center">
+<a href="https://numfocus.org/project/scverse">
+  <img
+    src="https://raw.githubusercontent.com/numfocus/templates/master/images/numfocus-logo.png"
+    width="200"
+  >
+</a>
+</div>

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/README.md

@@ -107,7 +107,20 @@ Marconato, L., Palla, G., Yamauchi, K.A. et al. SpatialData: an open and univers
 
 [scverse-discourse]: https://discourse.scverse.org/
 [issue-tracker]: https://github.com/scverse/spatialdata-io/issues
-[changelog]: https://spatialdata.scverse.org/projects/io/en/latest/changelog.html
-[link-docs]: https://spatialdata.scverse.org/projects/io/en/latest/
-[link-api]: https://spatialdata.scverse.org/projects/io/en/latest/api.html
-[link-cli]: https://spatialdata.scverse.org/projects/io/en/latest/cli.html
+[changelog]: https://spatialdata.scverse.org/projects/io/en/stable/changelog.html
+[link-docs]: https://spatialdata.scverse.org/projects/io/en/stable/
+[link-api]: https://spatialdata.scverse.org/projects/io/en/stable/api.html
+[link-cli]: https://spatialdata.scverse.org/projects/io/en/stable/cli.html
+[//]: # "numfocus-fiscal-sponsor-attribution"
+
+spatialdata-io is part of the scverse® project ([website](https://scverse.org), [governance](https://scverse.org/about/roles)) and is fiscally sponsored by [NumFOCUS](https://numfocus.org/).
+If you like scverse® and want to support our mission, please consider making a tax-deductible [donation](https://numfocus.org/donate-to-scverse) to help the project pay for developer time, professional services, travel, workshops, and a variety of other needs.
+
+<div align="center">
+<a href="https://numfocus.org/project/scverse">
+  <img
+    src="https://raw.githubusercontent.com/numfocus/templates/master/images/numfocus-logo.png"
+    width="200"
+  >
+</a>
+</div>

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/_version.py

@@ -1,8 +1,13 @@
-# file generated by setuptools_scm
+# file generated by setuptools-scm
 # don't change, don't track in version control
+
+__all__ = ["__version__", "__version_tuple__", "version", "version_tuple"]
+
 TYPE_CHECKING = False
 if TYPE_CHECKING:
-    from typing import Tuple, Union
+    from typing import Tuple
+    from typing import Union
+
     VERSION_TUPLE = Tuple[Union[int, str], ...]
 else:
     VERSION_TUPLE = object
@@ -12,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.1.7'
-__version_tuple__ = version_tuple = (0, 1, 7)
+__version__ = version = '0.2.0'
+__version_tuple__ = version_tuple = (0, 2, 0)

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/__main__.py

@@ -621,8 +621,8 @@ def macsima_wrapper(
 @click.option("--name", "-n", type=str, help="name of the element to be stored")
 @click.option(
     "--data-axes",
-    type=click.Choice(["cyx", "czyx"], case_sensitive=False),
-    help="Axes of the data for image files. Valid values are 'cyx' and 'czyx'.",
+    type=str,
+    help="Axes of the data for image files. Valid values are permutations of 'cyx' and 'czyx'.",
 )
 @click.option(
     "--coordinate-system",
@@ -638,6 +638,8 @@ def read_generic_wrapper(
     coordinate_system: str | None = None,
 ) -> None:
     """Read generic data to SpatialData"""
+    if data_axes is not None and "".join(sorted(data_axes)) not in ["cxy", "cxyz"]:
+        raise ValueError("data_axes must be a permutation of 'cyx' or 'czyx'.")
     generic_to_zarr(input=input, output=output, name=name, data_axes=data_axes, coordinate_system=coordinate_system)
 
 

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/_constants/_constants.py

@@ -148,6 +148,9 @@ class XeniumKeys(ModeEnum):
 
     # specs keys
     ANALYSIS_SW_VERSION = "analysis_sw_version"
+    # spec which contains xeniumranger version whenever xeniumranger is used to resegment the data; the new, resegmented data
+    # needs to be parsed by considering the xeniumranger version
+    XENIUM_RANGER = "xenium_ranger"
 
     # zarr file with labels file and cell summary keys
     CELLS_ZARR = "cells.zarr.zip"

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/converters/generic_to_zarr.py

@@ -69,14 +69,14 @@ def generic_to_zarr(
     )
 
     if output.exists():
-        sdata = SpatialData.read_zarr(output)
+        sdata = SpatialData.read(output)
         if name in sdata:
             raise ValueError(
                 f"Name {name} already exists in {output}; please provide a different name or delete the "
                 f"existing element."
             )
         sdata[name] = element
-        sdata.write_element(name, element_name=name)
+        sdata.write_element(element_name=name)
         print(f"Element {name} written to {output}")
     else:
         sdata = SpatialData.init_from_elements(elements={name: element})

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/readers/cosmx.py

@@ -15,14 +15,10 @@ from anndata import AnnData
 from dask.dataframe import DataFrame as DaskDataFrame
 from dask_image.imread import imread
 from scipy.sparse import csr_matrix
-
-# from spatialdata._core.core_utils import xy_cs
 from skimage.transform import estimate_transform
 from spatialdata import SpatialData
 from spatialdata._logging import logger
 from spatialdata.models import Image2DModel, Labels2DModel, PointsModel, TableModel
-
-# from spatialdata._core.ngff.ngff_coordinate_system import NgffAxis  # , CoordinateSystem
 from spatialdata.transformations.transformations import Affine, Identity
 
 from spatialdata_io._constants._constants import CosmxKeys
@@ -30,10 +26,6 @@ from spatialdata_io._docs import inject_docs
 
 __all__ = ["cosmx"]
 
-# x_axis = NgffAxis(name="x", type="space", unit="discrete")
-# y_axis = NgffAxis(name="y", type="space", unit="discrete")
-# c_axis = NgffAxis(name="c", type="channel", unit="index")
-
 
 @inject_docs(cx=CosmxKeys)
 def cosmx(
@@ -177,7 +169,7 @@ def cosmx(
     fovs_images_and_labels = set(fovs_images).intersection(set(fovs_labels))
     fovs_diff = fovs_images_and_labels.difference(set(fovs_counts))
     if len(fovs_diff):
-        raise logger.warning(
+        logger.warning(
             f"Found images and labels for {len(fovs_images)} FOVs, but only {len(fovs_counts)} FOVs in the counts file.\n"
             + f"The following FOVs are missing: {fovs_diff} \n"
             + "... will use only fovs in Table."

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/readers/generic.py

@@ -49,6 +49,8 @@ def generic(
     -------
     Parsed spatial element.
     """
+    if isinstance(input, str):
+        input = Path(input)
     if coordinate_system is None:
         coordinate_system = DEFAULT_COORDINATE_SYSTEM
     if input.suffix in VALID_SHAPE_TYPES:

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/readers/visium.py

@@ -33,6 +33,7 @@ def visium(
     fullres_image_file: str | Path | None = None,
     tissue_positions_file: str | Path | None = None,
     scalefactors_file: str | Path | None = None,
+    var_names_make_unique: bool = True,
     imread_kwargs: Mapping[str, Any] = MappingProxyType({}),
     image_models_kwargs: Mapping[str, Any] = MappingProxyType({}),
     **kwargs: Any,
@@ -72,6 +73,8 @@ def visium(
         Path to the tissue positions file.
     scalefactors_file
         Path to the scalefactors file.
+    var_names_make_unique
+        If `True`, call `.var_names_make_unique()` on each `AnnData` table.
     imread_kwargs
         Keyword arguments passed to :func:`dask_image.imread.imread`.
     image_models_kwargs
@@ -114,7 +117,10 @@ def visium(
     assert counts_file is not None
 
     if library_id is None and dataset_id is None:
-        raise ValueError("Cannot determine the `library_id`. Please provide `dataset_id`.")
+        raise ValueError(
+            "Cannot determine the `library_id`. Please provide `dataset_id`; the `dataset_id` value will be used to "
+            "name the elements in the `SpatialData` object."
+        )
 
     if dataset_id is not None:
         if dataset_id != library_id and library_id is not None:
@@ -202,14 +208,16 @@ def visium(
         radius=scalefactors["spot_diameter_fullres"] / 2.0,
         index=adata.obs["spot_id"].copy(),
         transformations={
-            "global": transform_original,
-            "downscaled_hires": transform_hires,
-            "downscaled_lowres": transform_lowres,
+            dataset_id: transform_original,
+            f"{dataset_id}_downscaled_hires": transform_hires,
+            f"{dataset_id}_downscaled_lowres": transform_lowres,
         },
     )
     shapes[dataset_id] = circles
     adata.obs["region"] = dataset_id
     table = TableModel.parse(adata, region=dataset_id, region_key="region", instance_key="spot_id")
+    if var_names_make_unique:
+        table.var_names_make_unique()
 
     images = {}
     if fullres_image_file is not None:
@@ -220,7 +228,7 @@ def visium(
             images[dataset_id + "_full_image"] = Image2DModel.parse(
                 full_image,
                 scale_factors=[2, 2, 2, 2],
-                transformations={"global": transform_original},
+                transformations={dataset_id: transform_original},
                 rgb=None,
                 **image_models_kwargs,
             )
@@ -231,14 +239,16 @@ def visium(
         image_hires = imread(path / VisiumKeys.IMAGE_HIRES_FILE, **imread_kwargs).squeeze().transpose(2, 0, 1)
         image_hires = DataArray(image_hires, dims=("c", "y", "x"))
         images[dataset_id + "_hires_image"] = Image2DModel.parse(
-            image_hires, transformations={"downscaled_hires": Identity(), "global": transform_hires.inverse()}, rgb=None
+            image_hires,
+            transformations={f"{dataset_id}_downscaled_hires": Identity(), dataset_id: transform_hires.inverse()},
+            rgb=None,
         )
     if (path / VisiumKeys.IMAGE_LOWRES_FILE).exists():
         image_lowres = imread(path / VisiumKeys.IMAGE_LOWRES_FILE, **imread_kwargs).squeeze().transpose(2, 0, 1)
         image_lowres = DataArray(image_lowres, dims=("c", "y", "x"))
         images[dataset_id + "_lowres_image"] = Image2DModel.parse(
             image_lowres,
-            transformations={"downscaled_lowres": Identity(), "global": transform_lowres.inverse()},
+            transformations={f"{dataset_id}_downscaled_lowres": Identity(), dataset_id: transform_lowres.inverse()},
             rgb=None,
         )
 

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/readers/visium_hd.py

@@ -17,16 +17,17 @@ from geopandas import GeoDataFrame
 from imageio import imread as imread2
 from multiscale_spatial_image import MultiscaleSpatialImage
 from numpy.random import default_rng
+from skimage.transform import ProjectiveTransform, warp
 from spatial_image import SpatialImage
-from spatialdata import SpatialData, rasterize_bins, rasterize_bins_link_table_to_labels
-from spatialdata.models import Image2DModel, ShapesModel, TableModel
-from spatialdata.transformations import (
-    Affine,
-    Identity,
-    Scale,
-    Sequence,
-    set_transformation,
+from spatialdata import (
+    SpatialData,
+    get_extent,
+    rasterize_bins,
+    rasterize_bins_link_table_to_labels,
 )
+from spatialdata._types import ArrayLike
+from spatialdata.models import Image2DModel, ShapesModel, TableModel
+from spatialdata.transformations import Affine, Identity, Scale, set_transformation
 from xarray import DataArray
 
 from spatialdata_io._constants._constants import VisiumHDKeys
@@ -45,6 +46,7 @@ def visium_hd(
     annotate_table_by_labels: bool = False,
     fullres_image_file: str | Path | None = None,
     load_all_images: bool = False,
+    var_names_make_unique: bool = True,
     imread_kwargs: Mapping[str, Any] = MappingProxyType({}),
     image_models_kwargs: Mapping[str, Any] = MappingProxyType({}),
     anndata_kwargs: Mapping[str, Any] = MappingProxyType({}),
@@ -61,7 +63,8 @@ def visium_hd(
     path
         Path to directory containing the *10x Genomics* Visium HD output.
     dataset_id
-        Unique identifier of the dataset. If `None`, it tries to infer it from the file name of the feature slice file.
+        Unique identifier of the dataset, used to name the elements of the `SpatialData` object. If `None`, it tries to
+         infer it from the file name of the feature slice file.
     filtered_counts_file
         It sets the value of `counts_file` to ``{vx.FILTERED_COUNTS_FILE!r}`` (when `True`) or to
         ``{vx.RAW_COUNTS_FILE!r}`` (when `False`).
@@ -80,6 +83,8 @@ def visium_hd(
     load_all_images
         If `False`, load only the full resolution, high resolution and low resolution images. If `True`, also the
         following images: ``{vx.IMAGE_CYTASSIST!r}``.
+    var_names_make_unique
+        If `True`, call `.var_names_make_unique()` on each `AnnData` table.
     imread_kwargs
         Keyword arguments for :func:`imageio.imread`.
     image_models_kwargs
@@ -100,7 +105,8 @@ def visium_hd(
 
     if dataset_id is None:
         dataset_id = _infer_dataset_id(path)
-    filename_prefix = f"{dataset_id}_"
+
+    filename_prefix = _get_filename_prefix(path, dataset_id)
 
     def load_image(path: Path, suffix: str, scale_factors: list[int] | None = None) -> None:
         _load_image(
@@ -236,9 +242,9 @@ def visium_hd(
         shapes_name = dataset_id + "_" + bin_size_str
         radius = scalefactors[VisiumHDKeys.SCALEFACTORS_SPOT_DIAMETER_FULLRES] / 2.0
         transformations = {
-            "global": transform_original,
-            "downscaled_hires": transform_hires,
-            "downscaled_lowres": transform_lowres,
+            dataset_id: transform_original,
+            f"{dataset_id}_downscaled_hires": transform_hires,
+            f"{dataset_id}_downscaled_lowres": transform_lowres,
         }
         circles = ShapesModel.parse(
             adata.obsm["spatial"],
@@ -265,6 +271,8 @@ def visium_hd(
             region_key=str(VisiumHDKeys.REGION_KEY),
             instance_key=str(VisiumHDKeys.INSTANCE_KEY),
         )
+        if var_names_make_unique:
+            tables[bin_size_str].var_names_make_unique()
 
     # read full resolution image
     if fullres_image_file is not None:
@@ -314,7 +322,10 @@ def visium_hd(
     )
     set_transformation(
         images[dataset_id + "_hires_image"],
-        {"downscaled_hires": Identity(), "global": transform_hires.inverse()},
+        {
+            f"{dataset_id}_downscaled_hires": Identity(),
+            dataset_id: transform_hires.inverse(),
+        },
         set_all=True,
     )
 
@@ -332,7 +343,10 @@ def visium_hd(
     )
     set_transformation(
         images[dataset_id + "_lowres_image"],
-        {"downscaled_lowres": Identity(), "global": transform_lowres.inverse()},
+        {
+            f"{dataset_id}_downscaled_lowres": Identity(),
+            dataset_id: transform_lowres.inverse(),
+        },
         set_all=True,
     )
 
@@ -351,15 +365,60 @@ def visium_hd(
         )
         image = images[dataset_id + "_cytassist_image"]
         transform_matrices = _get_transform_matrices(metadata, hd_layout)
-        affine0 = transform_matrices["cytassist_colrow_to_spot_colrow"]
-        affine1 = transform_matrices["spot_colrow_to_microscope_colrow"]
-        set_transformation(image, Sequence([affine0, affine1]), "global")
+        projective0 = transform_matrices["cytassist_colrow_to_spot_colrow"]
+        projective1 = transform_matrices["spot_colrow_to_microscope_colrow"]
+        projective = projective1 @ projective0
+        projective /= projective[2, 2]
+        if _projective_matrix_is_affine(projective):
+            affine = Affine(projective, input_axes=("x", "y"), output_axes=("x", "y"))
+            set_transformation(image, affine, dataset_id)
+        else:
+            # the projective matrix is not affine, we will separate the affine part and the projective shift, and apply
+            # the projective shift to the image
+            affine_matrix, projective_shift = _decompose_projective_matrix(projective)
+            affine = Affine(affine_matrix, input_axes=("x", "y"), output_axes=("x", "y"))
+
+            # determine the size of the transformed image
+            bounding_box = get_extent(image, coordinate_system=dataset_id)
+            x0, x1 = bounding_box["x"]
+            y0, y1 = bounding_box["y"]
+            x1 -= 1
+            y1 -= 1
+            corners = [(x0, y0), (x1, y0), (x1, y1), (x0, y1)]
+
+            transformed_corners = []
+            for x, y in corners:
+                px, py = _projective_matrix_transform_point(projective_shift, x, y)
+                transformed_corners.append((px, py))
+            transformed_corners_array = np.array(transformed_corners)
+            transformed_bounds = (
+                np.min(transformed_corners_array[:, 0]),
+                np.min(transformed_corners_array[:, 1]),
+                np.max(transformed_corners_array[:, 0]),
+                np.max(transformed_corners_array[:, 1]),
+            )
+            # the first two components are <= 0, we just discard them since the cytassist image has a lot of padding
+            # and therefore we can safely discard pixels with negative coordinates
+            transformed_shape = (np.ceil(transformed_bounds[2]), np.ceil(transformed_bounds[3]))
+
+            # flip xy
+            transformed_shape = (transformed_shape[1], transformed_shape[0])
+
+            # the cytassist image is a small, single-scale image, so we can compute it in memory
+            numpy_data = image.transpose("y", "x", "c").data.compute()
+            warped = warp(
+                numpy_data, ProjectiveTransform(projective_shift).inverse, output_shape=transformed_shape, order=1
+            )
+            warped = np.round(warped * 255).astype(np.uint8)
+            warped = Image2DModel.parse(warped, dims=("y", "x", "c"), transformations={dataset_id: affine}, rgb=True)
+
+            # we replace the cytassist image with the warped image
+            images[dataset_id + "_cytassist_image"] = warped
 
     sdata = SpatialData(tables=tables, images=images, shapes=shapes, labels=labels)
 
     if annotate_table_by_labels:
         for bin_size_str in bin_sizes:
-
             shapes_name = dataset_id + "_" + bin_size_str
 
             # add labels layer (rasterized bins).
@@ -388,7 +447,8 @@ def _infer_dataset_id(path: Path) -> str:
     files = [file.name for file in path.iterdir() if file.is_file() and file.name.endswith(suffix)]
     if len(files) == 0 or len(files) > 1:
         raise ValueError(
-            f"Cannot infer `dataset_id` from the feature slice file in {path}, please pass `dataset_id` as an argument."
+            f"Cannot infer `dataset_id` from the feature slice file in {path}, please pass `dataset_id` as an "
+            f"argument. The `dataset_id` value will be used to name the elements in the `SpatialData` object."
         )
     return files[0].replace(suffix, "")
 
@@ -424,20 +484,63 @@ def _load_image(
             ), "When the image is not in RGB, the first dimension should be the number of channels."
 
         image = DataArray(data, dims=("c", "y", "x"))
-        parsed = Image2DModel.parse(image, scale_factors=scale_factors, rgb=None, **image_models_kwargs)
+        parsed = Image2DModel.parse(
+            image,
+            scale_factors=scale_factors,
+            rgb=None,
+            transformations={dataset_id: Identity()},
+            **image_models_kwargs,
+        )
         images[dataset_id + suffix] = parsed
     else:
         warnings.warn(f"File {path} does not exist, skipping it.", UserWarning, stacklevel=2)
     return None
 
 
-def _get_affine(coefficients: list[int]) -> Affine:
-    matrix = np.array(coefficients).reshape(3, 3)
-    # the last row doesn't match with machine precision, let's check the matrix it's still close to a homogeneous
-    # matrix, and fix this
-    assert np.allclose(matrix[2], [0, 0, 1], atol=1e-2), matrix
-    matrix[2] = [0, 0, 1]
-    return Affine(matrix, input_axes=("x", "y"), output_axes=("x", "y"))
+def _projective_matrix_transform_point(projective_shift: ArrayLike, x: float, y: float) -> tuple[float, float]:
+    v = np.array([x, y, 1])
+    v = projective_shift @ v
+    v /= v[2]
+    return v[0], v[1]
+
+
+def _projective_matrix_is_affine(projective_matrix: ArrayLike) -> bool:
+    assert np.allclose(projective_matrix[2, 2], 1), "A projective matrix should have a 1 in the bottom right corner."
+    return np.allclose(projective_matrix[2, :2], [0, 0])
+
+
+def _decompose_projective_matrix(projective_matrix: ArrayLike) -> tuple[ArrayLike, ArrayLike]:
+    """
+    Decompose a projective transformation matrix into an affine transformation and a projective shift.
+
+    Parameters
+    ----------
+    projective_matrix
+        Projective transformation matrix.
+
+    Returns
+    -------
+    A tuple where the first element is the affine matrix and the second element is the projective shift.
+
+    Let P be the initial projective matrix and A the affine matrix. The projective shift S is defined as: S = A^-1 @ P.
+    """
+    assert np.allclose(projective_matrix[2, 2], 1), "A projective matrix should have a 1 in the bottom right corner."
+    affine_matrix = projective_matrix.copy()
+    affine_matrix[2] = [0, 0, 1]
+    # equivalent to np.linalg.inv(affine_matrix) @ projective_matrix, but more numerically stable
+    projective_shift = np.linalg.solve(affine_matrix, projective_matrix)
+    projective_shift /= projective_shift[2, 2]
+    return affine_matrix, projective_shift
+
+
+def _get_filename_prefix(path: Path, dataset_id: str) -> str:
+    if (path / f"{dataset_id}_{VisiumHDKeys.FEATURE_SLICE_FILE.value}").exists():
+        return f"{dataset_id}_"
+    assert (path / VisiumHDKeys.FEATURE_SLICE_FILE.value).exists(), (
+        f"Cannot locate the feature slice file, please ensure the file is present in the {path} directory and/or adjust"
+        "the `dataset_id` parameter"
+    )
+    return ""
 
 
 def _parse_metadata(path: Path, filename_prefix: str) -> tuple[dict[str, Any], dict[str, Any]]:
@@ -447,11 +550,29 @@ def _parse_metadata(path: Path, filename_prefix: str) -> tuple[dict[str, Any], d
     return metadata, hd_layout
 
 
-def _get_transform_matrices(metadata: dict[str, Any], hd_layout: dict[str, Any]) -> dict[str, Affine]:
+def _get_transform_matrices(metadata: dict[str, Any], hd_layout: dict[str, Any]) -> dict[str, ArrayLike]:
+    """
+    Gets 4 projective transformation matrices, describing how to align the CytAssist, spots and microscope coordinates.
+
+    Parameters
+    ----------
+    metadata
+        Metadata of the Visium HD dataset parsed using `_parse_metadata()` from the feature slice file.
+    hd_layout
+        Layout of the Visium HD dataset parsed using `_parse_metadata()` from the feature slice file.
+
+    Returns
+    -------
+    A dictionary containing four projective transformation matrices:
+    - CytAssist col/row to Spot col/row
+    - Spot col/row to CytAssist col/row
+    - Microscope col/row to Spot col/row
+    - Spot col/row to Microscope col/row
+    """
     transform_matrices = {}
 
-    # not used
-    transform_matrices["hd_layout_transform"] = _get_affine(hd_layout[VisiumHDKeys.TRANSFORM])
+    # this transformation is parsed but not used in the current implementation
+    transform_matrices["hd_layout_transform"] = np.array(hd_layout[VisiumHDKeys.TRANSFORM]).reshape(3, 3)
 
     for key in [
         VisiumHDKeys.CYTASSIST_COLROW_TO_SPOT_COLROW,
@@ -459,7 +580,7 @@ def _get_transform_matrices(metadata: dict[str, Any], hd_layout: dict[str, Any])
         VisiumHDKeys.MICROSCOPE_COLROW_TO_SPOT_COLROW,
         VisiumHDKeys.SPOT_COLROW_TO_MICROSCOPE_COLROW,
     ]:
-        data = metadata[VisiumHDKeys.TRANSFORM_MATRICES][key]
-        transform_matrices[key.value] = _get_affine(data)
+        coefficients = metadata[VisiumHDKeys.TRANSFORM_MATRICES][key]
+        transform_matrices[key] = np.array(coefficients).reshape(3, 3)
 
     return transform_matrices

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/src/spatialdata_io/readers/xenium.py

@@ -298,7 +298,6 @@ def xenium(
                     f"{XeniumKeys.MORPHOLOGY_FOCUS_CHANNEL_IMAGE.value.format(0)} to "
                     f"{XeniumKeys.MORPHOLOGY_FOCUS_CHANNEL_IMAGE.value.format(len(files) - 1)}, found {files}"
                 )
-            # the 'dummy' channel is a temporary workaround, see _get_images() for more details
             if len(files) == 1:
                 channel_names = {
                     0: XeniumKeys.MORPHOLOGY_FOCUS_CHANNEL_0.value,
@@ -309,7 +308,6 @@ def xenium(
                     1: XeniumKeys.MORPHOLOGY_FOCUS_CHANNEL_1.value,
                     2: XeniumKeys.MORPHOLOGY_FOCUS_CHANNEL_2.value,
                     3: XeniumKeys.MORPHOLOGY_FOCUS_CHANNEL_3.value,
-                    4: "dummy",
                 }
             # this reads the scale 0 for all the 1 or 4 channels (the other files are parsed automatically)
             # dask.image.imread will call tifffile.imread which will give a warning saying that reading multi-file
@@ -601,6 +599,8 @@ def xenium_aligned_image(
     imread_kwargs: Mapping[str, Any] = MappingProxyType({}),
     image_models_kwargs: Mapping[str, Any] = MappingProxyType({}),
     dims: tuple[str, ...] | None = None,
+    rgba: bool = False,
+    c_coords: list[str] | None = None,
 ) -> DataTree:
     """
     Read an image aligned to a Xenium dataset, with an optional alignment file.
@@ -619,6 +619,13 @@ def xenium_aligned_image(
         Example: for an image with shape (1, y, 1, x, 3), use dims=("anystring", "y", "dummy", "x", "c"). Values that
         are not "c", "x" or "y" are considered dummy dimensions and will be squeezed (the data must have len 1 for
         those axes).
+    rgba
+        Interprets the `c` channel as RGBA, by setting the channel names to `r`, `g`, `b` (`a`). When `c_coords` is not
+        `None`, this argument is ignored.
+    c_coords
+        Channel names for the image. By default, the function will try to infer the channel names from the image
+        shape and name (by detecting if the name suggests that the image is a H&E image). Example: for an RGB image with
+        shape (3, y, x), use c_coords=["r", "g", "b"].
 
     Returns
     -------
@@ -645,10 +652,6 @@ def xenium_aligned_image(
         else:
             assert len(image.shape) == 3
             assert image.shape[0] in [3, 4]
-            if image.shape[0] == 4:
-                # as explained before in _get_images(), we need to add a dummy channel until we support 4-channel images as
-                # non-RGBA images in napari
-                image = da.concatenate([image, da.zeros_like(image[0:1])], axis=0)
             dims = ("c", "y", "x")
     else:
         logging.info(f"Image has shape {image.shape}, parsing with dims={dims}.")
@@ -667,10 +670,19 @@ def xenium_aligned_image(
         alignment = pd.read_csv(alignment_file, header=None).values
         transformation = Affine(alignment, input_axes=("x", "y"), output_axes=("x", "y"))
 
+    if c_coords is None and (rgba or image_path.name.endswith(XeniumKeys.ALIGNED_HE_IMAGE_SUFFIX)):
+        c_index = dims.index("c")
+        n_channels = image.shape[c_index]
+        if n_channels == 3:
+            c_coords = ["r", "g", "b"]
+        elif n_channels == 4:
+            c_coords = ["r", "g", "b", "a"]
+
     return Image2DModel.parse(
         image,
         dims=dims,
         transformations={"global": transformation},
+        c_coords=c_coords,
         **image_models_kwargs,
     )
 
@@ -719,7 +731,17 @@ def _parse_version_of_xenium_analyzer(
     specs: dict[str, Any],
     hide_warning: bool = True,
 ) -> packaging.version.Version | None:
-    string = specs[XeniumKeys.ANALYSIS_SW_VERSION]
+
+    # After using xeniumranger (e.g. 3.0.1.1) to resegment data from previous versions (e.g. xenium-1.6.0.7), a new dict is added to
+    # `specs`, named 'xenium_ranger', which contains the key 'version' and whose value specifies the version of xeniumranger used to
+    # resegment the data (e.g. 'xenium-3.0.1.1').
+    # When parsing the outs/ folder from the resegmented data, this version (rather than the original 'analysis_sw_version') is used
+    # whenever a code branch is dependent on the data version
+    if specs.get(XeniumKeys.XENIUM_RANGER):
+        string = specs[XeniumKeys.XENIUM_RANGER]["version"]
+    else:
+        string = specs[XeniumKeys.ANALYSIS_SW_VERSION]
+
     pattern = r"^(?:x|X)enium-(\d+\.\d+\.\d+(\.\d+-\d+)?)"
 
     result = re.search(pattern, string)
@@ -780,6 +802,3 @@ def prefix_suffix_uint32_from_cell_id_str(cell_id_str: ArrayLike) -> tuple[Array
     cell_id_prefix = [int(x, 16) for x in cell_id_prefix_hex]
 
     return np.array(cell_id_prefix, dtype=np.uint32), np.array(dataset_suffix_int)
-
-
-##

{spatialdata_io-0.1.7 → spatialdata_io-0.2.0}/tests/test_generic.py

@@ -87,7 +87,7 @@ def test_cli_read_generic_image_invalid_data_axes(runner: CliRunner) -> None:
             ],
         )
         assert result.exit_code != 0, result.output
-        assert "'invalid_axes' is not one of 'cyx', 'czyx'." in result.output
+        assert "data_axes must be a permutation of 'cyx' or 'czyx'." in result.exc_info[1].args[0]
 
 
 @pytest.mark.parametrize("cli", [True, False])