spatialdata 0.2.0rc1__tar.gz → 0.2.2__tar.gz

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/.pre-commit-config.yaml

@@ -9,7 +9,7 @@ ci:
     skip: []
 repos:
     - repo: https://github.com/psf/black
-      rev: 24.4.2
+      rev: 24.8.0
       hooks:
           - id: black
     - repo: https://github.com/pre-commit/mirrors-prettier
@@ -17,17 +17,17 @@ repos:
       hooks:
           - id: prettier
     - repo: https://github.com/asottile/blacken-docs
-      rev: 1.16.0
+      rev: 1.18.0
       hooks:
           - id: blacken-docs
     - repo: https://github.com/pre-commit/mirrors-mypy
-      rev: v1.10.0
+      rev: v1.11.1
       hooks:
           - id: mypy
             additional_dependencies: [numpy, types-requests]
             exclude: tests/|docs/
     - repo: https://github.com/astral-sh/ruff-pre-commit
-      rev: v0.4.10
+      rev: v0.5.6
       hooks:
           - id: ruff
             args: [--fix, --exit-non-zero-on-fix]

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/.readthedocs.yaml

@@ -13,6 +13,7 @@ python:
           path: .
           extra_requirements:
               - docs
+              - torch
 submodules:
     include:
         - "docs/tutorials/notebooks"

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/CHANGELOG.md

@@ -8,12 +8,37 @@ and this project adheres to [Semantic Versioning][].
 [keep a changelog]: https://keepachangelog.com/en/1.0.0/
 [semantic versioning]: https://semver.org/spec/v2.0.0.html
 
-## [0.2.0] - 2024-xx-xx
+## [0.x.x] - 2024-xx-xx
+
+## [0.2.2] - 2024-08-07
+
+# Major
+
+-   New disk format for shapes using `GeoParquet` (the change is backward compatible) #542
+
+# Minor
+
+-   Add `return_background` as argument to `get_centroids` and `get_element_instances` #621
+-   Ability to save data using older disk formats #542
+
+# Fixed
+
+-   Circles validation now checks for inf or nan radii #653
+-   Bug with table name in torch dataset #654 @LLehner
+
+## [0.2.1] - 2024-07-04
+
+### Minor
+
+-   Relaxing `spatial-image` package requirement #616
+
+## [0.2.0] - 2024-07-03
 
 ### Changed
 
 -   Using `DataArray` directly instead of the subclass `SpatialImage` (removed install constraint for the `spatial_image` package) #587
 -   Using `DataTree` directly instead of the subclass `MultiscaleSpatialImage` (removed install constraint for the `multiscale_spatial_image` package) #587
+-   Changed `element`parameter (deprecation in v0.3.0) of `transform_element_to_coordinate_system` to a string `element_name` #611
 
 ### Major
 
@@ -34,6 +59,7 @@ and this project adheres to [Semantic Versioning][].
 ### Fixed
 
 -   Preserve channel names of multi-scale images in `transform` (#379)
+-   Fix `filter_by_coordinate_system` with SpatialData object having a table not annotating an element (#619)
 
 ## [0.1.2] - 2024-03-30
 

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: spatialdata
-Version: 0.2.0rc1
+Version: 0.2.2
 Summary: Spatial data format.
 Project-URL: Documentation, https://spatialdata.scverse.org/en/latest
 Project-URL: Source, https://github.com/scverse/spatialdata.git
@@ -41,7 +41,7 @@ Requires-Python: >=3.9
 Requires-Dist: anndata>=0.9.1
 Requires-Dist: click
 Requires-Dist: dask-image
-Requires-Dist: dask>=2024.2.1
+Requires-Dist: dask>=2024.4.1
 Requires-Dist: fsspec<=2023.6
 Requires-Dist: geopandas>=0.14
 Requires-Dist: multiscale-spatial-image>=1.0.0
@@ -56,7 +56,7 @@ Requires-Dist: scikit-image
 Requires-Dist: scipy
 Requires-Dist: setuptools
 Requires-Dist: shapely>=2.0.1
-Requires-Dist: spatial-image>=1.1.0
+Requires-Dist: spatial-image>=1.0.0
 Requires-Dist: typing-extensions>=4.8.0
 Requires-Dist: xarray
 Requires-Dist: xarray-datatree
@@ -76,7 +76,7 @@ Requires-Dist: sphinx-pytest; extra == 'docs'
 Requires-Dist: sphinx>=4.5; extra == 'docs'
 Requires-Dist: sphinxcontrib-bibtex>=1.0.0; extra == 'docs'
 Provides-Extra: extra
-Requires-Dist: napari-spatialdata>=0.2.4; extra == 'extra'
+Requires-Dist: napari-spatialdata[all]>=0.2.4; extra == 'extra'
 Requires-Dist: spatialdata-io; extra == 'extra'
 Requires-Dist: spatialdata-plot; extra == 'extra'
 Provides-Extra: test
@@ -135,13 +135,17 @@ Another useful resource to get started is the source code of the [`spatialdata-i
 
 ## Installation
 
-Check out the docs for more complete [installation instructions](https://spatialdata.scverse.org/en/latest/installation.html). To get started with the "batteries included" installation, you can install via pip:
+Check out the docs for more complete [installation instructions](https://spatialdata.scverse.org/en/stable/installation.html). To get started with the "batteries included" installation, you can install via pip:
 
 ```bash
 pip install "spatialdata[extra]"
 ```
 
-Note: if you are using a Mac with an M1/M2 chip, please follow the installation instructions.
+or via conda:
+
+```bash
+mamba install -c conda-forge spatialdata napari-spatialdata spatialdata-io spatialdata-plot
+```
 
 ## Limitations
 
@@ -151,7 +155,7 @@ Note: if you are using a Mac with an M1/M2 chip, please follow the installation
 
 To get involved in the discussion, or if you need help to get started, you are welcome to use the following options.
 
--   <ins>Chat</ins> via [`scverse` Zulip](https://imagesc.zulipchat.com/#narrow/stream/329057-scverse/topic/segmentation) (public or 1 to 1).
+-   <ins>Chat</ins> via [`scverse` Zulip](https://scverse.zulipchat.com/#narrow/stream/315824-spatial) (public or 1 to 1).
 -   <ins>Forum post</ins> in the [scverse discourse forum](https://discourse.scverse.org/).
 -   <ins>Bug report/feature request</ins> via the [GitHub issue tracker][issue-tracker].
 -   <ins>Zoom call</ins> as part of the SpatialData Community Meetings, held every 2 weeks on Thursday, [schedule here](https://hackmd.io/enWU826vRai-JYaL7TZaSw).

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/README.md

@@ -47,13 +47,17 @@ Another useful resource to get started is the source code of the [`spatialdata-i
 
 ## Installation
 
-Check out the docs for more complete [installation instructions](https://spatialdata.scverse.org/en/latest/installation.html). To get started with the "batteries included" installation, you can install via pip:
+Check out the docs for more complete [installation instructions](https://spatialdata.scverse.org/en/stable/installation.html). To get started with the "batteries included" installation, you can install via pip:
 
 ```bash
 pip install "spatialdata[extra]"
 ```
 
-Note: if you are using a Mac with an M1/M2 chip, please follow the installation instructions.
+or via conda:
+
+```bash
+mamba install -c conda-forge spatialdata napari-spatialdata spatialdata-io spatialdata-plot
+```
 
 ## Limitations
 
@@ -63,7 +67,7 @@ Note: if you are using a Mac with an M1/M2 chip, please follow the installation
 
 To get involved in the discussion, or if you need help to get started, you are welcome to use the following options.
 
--   <ins>Chat</ins> via [`scverse` Zulip](https://imagesc.zulipchat.com/#narrow/stream/329057-scverse/topic/segmentation) (public or 1 to 1).
+-   <ins>Chat</ins> via [`scverse` Zulip](https://scverse.zulipchat.com/#narrow/stream/315824-spatial) (public or 1 to 1).
 -   <ins>Forum post</ins> in the [scverse discourse forum](https://discourse.scverse.org/).
 -   <ins>Bug report/feature request</ins> via the [GitHub issue tracker][issue-tracker].
 -   <ins>Zoom call</ins> as part of the SpatialData Community Meetings, held every 2 weeks on Thursday, [schedule here](https://hackmd.io/enWU826vRai-JYaL7TZaSw).

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.2.0rc1'
-__version_tuple__ = version_tuple = (0, 2, 0)
+__version__ = version = '0.2.2'
+__version_tuple__ = version_tuple = (0, 2, 2)

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/pyproject.toml

@@ -25,7 +25,7 @@ dependencies = [
     "anndata>=0.9.1",
     "click",
     "dask-image",
-    "dask>=2024.2.1",
+    "dask>=2024.4.1",
     "fsspec<=2023.6",
     "geopandas>=0.14",
     "multiscale_spatial_image>=1.0.0",
@@ -38,7 +38,7 @@ dependencies = [
     "rich",
     "setuptools",
     "shapely>=2.0.1",
-    "spatial_image>=1.1.0",
+    "spatial_image>=1.0.0",
     "scikit-image",
     "scipy",
     "typing_extensions>=4.8.0",
@@ -73,7 +73,7 @@ torch = [
     "torch"
 ]
 extra  = [
-    "napari-spatialdata>=0.2.4",
+    "napari-spatialdata[all]>=0.2.4",
     "spatialdata-plot",
     "spatialdata-io",
 ]

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/src/spatialdata/_core/centroids.py

@@ -35,6 +35,7 @@ def _validate_coordinate_system(e: SpatialElement, coordinate_system: str) -> No
 def get_centroids(
     e: SpatialElement,
     coordinate_system: str = "global",
+    return_background: bool = False,
 ) -> DaskDataFrame:
     """
     Get the centroids of the geometries contained in a SpatialElement, as a new Points element.
@@ -45,6 +46,8 @@ def get_centroids(
         The SpatialElement. Only points, shapes (circles, polygons and multipolygons) and labels are supported.
     coordinate_system
         The coordinate system in which the centroids are computed.
+    return_background
+        If True, the centroid of the background label (0) is included in the output.
 
     Notes
     -----
@@ -69,7 +72,7 @@ def _get_centroids_for_axis(xdata: xr.DataArray, axis: str) -> pd.DataFrame:
     -------
     pd.DataFrame
         A DataFrame containing one column, named after "axis", with the centroids of the labels along that axis.
-        The index of the DataFrame is the collection of label values, sorted ascendingly.
+        The index of the DataFrame is the collection of label values, sorted in ascending order.
     """
     centroids: dict[int, float] = defaultdict(float)
     for i in xdata[axis]:
@@ -95,6 +98,7 @@ def _get_centroids_for_axis(xdata: xr.DataArray, axis: str) -> pd.DataFrame:
 def _(
     e: DataArray | DataTree,
     coordinate_system: str = "global",
+    return_background: bool = False,
 ) -> DaskDataFrame:
     """Get the centroids of a Labels element (2D or 3D)."""
     model = get_model(e)
@@ -110,6 +114,8 @@ def _(
     for axis in get_axes_names(e):
         dfs.append(_get_centroids_for_axis(e, axis))
     df = pd.concat(dfs, axis=1)
+    if not return_background and 0 in df.index:
+        df = df.drop(index=0)  # drop the background label
     t = get_transformation(e, coordinate_system)
     centroids = PointsModel.parse(df, transformations={coordinate_system: t})
     return transform(centroids, to_coordinate_system=coordinate_system)

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/src/spatialdata/_core/operations/rasterize_bins.py

@@ -9,6 +9,7 @@ from dask.dataframe import DataFrame as DaskDataFrame
 from geopandas import GeoDataFrame
 from numpy.random import default_rng
 from scipy.sparse import csc_matrix
+from shapely import MultiPolygon, Point, Polygon
 from skimage.transform import estimate_transform
 from xarray import DataArray
 
@@ -65,6 +66,9 @@ def rasterize_bins(
     The returned image will have one pixel for each bin, and a coordinate transformation to map the image to the
     original data orientation. In particular, the bins of Visium HD data are in a grid that is slightly rotated;
     the coordinate transformation will adjust for this, so that the returned data is aligned to the original geometries.
+
+    If `spatialdata-plot` is used to visualized the returned image, the parameter `scale='full'` needs to be passed to
+    `.render_shapes()`, to disable an automatic rasterization that would confict with the rasterization performed here.
     """
     element = sdata[bins]
     table = sdata.tables[table_name]
@@ -152,8 +156,13 @@ def rasterize_bins(
 
     src = np.stack([sub_table.obs[col_key] - min_col, sub_table.obs[row_key] - min_row], axis=1)
     if isinstance(sub_df, GeoDataFrame):
-        sub_x = sub_df.geometry.x.values
-        sub_y = sub_df.geometry.y.values
+        if isinstance(sub_df.iloc[0].geometry, Point):
+            sub_x = sub_df.geometry.x.values
+            sub_y = sub_df.geometry.y.values
+        else:
+            assert isinstance(sub_df.iloc[0].geometry, (Polygon, MultiPolygon))
+            sub_x = sub_df.centroid.x
+            sub_y = sub_df.centroid.y
     else:
         assert isinstance(sub_df, DaskDataFrame)
         sub_x = sub_df.x.compute().values

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/src/spatialdata/_core/query/relational_query.py

@@ -86,6 +86,7 @@ def _filter_table_by_element_names(table: AnnData | None, element_names: str | l
 @singledispatch
 def get_element_instances(
     element: SpatialElement,
+    return_background: bool = False,
 ) -> pd.Index:
     """
     Get the instances (index values) of the SpatialElement.
@@ -94,6 +95,8 @@ def get_element_instances(
     ----------
     element
         The SpatialElement.
+    return_background
+        If True, the background label (0) is included in the output.
 
     Returns
     -------
@@ -106,6 +109,7 @@ def get_element_instances(
 @get_element_instances.register(DataTree)
 def _(
     element: DataArray | DataTree,
+    return_background: bool = False,
 ) -> pd.Index:
     model = get_model(element)
     assert model in [Labels2DModel, Labels3DModel], "Expected a `Labels` element. Found an `Image` instead."
@@ -119,7 +123,10 @@ def _(
         xdata = next(iter(v))
         # can be slow
         instances = da.unique(xdata.data).compute()
-    return pd.Index(np.sort(instances))
+    index = pd.Index(np.sort(instances))
+    if not return_background and 0 in index:
+        return index.drop(0)  # drop the background label
+    return index
 
 
 @get_element_instances.register(GeoDataFrame)
@@ -568,7 +575,8 @@ def join_spatialelement_table(
     both the SpatialElement and table.
 
     For Points and Shapes elements every valid join for argument how is supported. For Labels elements only
-     the ``'left'`` and ``'right_exclusive'`` joins are supported.
+    the ``'left'`` and ``'right_exclusive'`` joins are supported.
+    For Labels, the background label (0) is not included in the output and it will not be returned.
 
     Parameters
     ----------
@@ -896,7 +904,7 @@ def get_values(
             x = matched_table[:, value_key_values].X
             import scipy
 
-            if isinstance(x, scipy.sparse.csr_matrix):
+            if isinstance(x, (scipy.sparse.csr_matrix, scipy.sparse.csc_matrix, scipy.sparse.coo_matrix)):
                 x = x.todense()
             df = pd.DataFrame(x, columns=value_key_values)
         if origin == "obsm":

{spatialdata-0.2.0rc1 → spatialdata-0.2.2}/src/spatialdata/_core/query/spatial_query.py

@@ -2,9 +2,10 @@ from __future__ import annotations
 
 import warnings
 from abc import abstractmethod
+from collections.abc import Mapping
 from dataclasses import dataclass
 from functools import singledispatch
-from typing import Any, Callable
+from typing import TYPE_CHECKING, Any, Callable
 
 import dask.array as da
 import dask.dataframe as dd
@@ -49,7 +50,7 @@ def _get_bounding_box_corners_in_intrinsic_coordinates(
     min_coordinate: list[Number] | ArrayLike,
     max_coordinate: list[Number] | ArrayLike,
     target_coordinate_system: str,
-) -> tuple[ArrayLike, tuple[str, ...]]:
+) -> tuple[DataArray, tuple[str, ...]]:
     """Get all corners of a bounding box in the intrinsic coordinates of an element.
 
     Parameters
@@ -74,38 +75,58 @@ def _get_bounding_box_corners_in_intrinsic_coordinates(
 
     The axes of the intrinsic coordinate system.
     """
-    from spatialdata.transformations import get_transformation
-
     min_coordinate = _parse_list_into_array(min_coordinate)
     max_coordinate = _parse_list_into_array(max_coordinate)
-    # get the transformation from the element's intrinsic coordinate system
-    # to the query coordinate space
-    transform_to_query_space = get_transformation(element, to_coordinate_system=target_coordinate_system)
+
+    # compute the output axes of the transformation, remove c from input and output axes, return the matrix without c
+    # and then build an affine transformation from that
     m_without_c, input_axes_without_c, output_axes_without_c = _get_axes_of_tranformation(
         element, target_coordinate_system
     )
-    axes, min_coordinate, max_coordinate = _adjust_bounding_box_to_real_axes(
+    spatial_transform = Affine(m_without_c, input_axes=input_axes_without_c, output_axes=output_axes_without_c)
+
+    # we identified 5 cases (see the responsible function for details), cases 1 and 5 correspond to invertible
+    # transformations; we focus on them
+    m_without_c_linear = m_without_c[:-1, :-1]
+    _ = _get_case_of_bounding_box_query(m_without_c_linear, input_axes_without_c, output_axes_without_c)
+
+    # adjust the bounding box to the real axes, dropping or adding eventually mismatching axes; the order of the axes is
+    # not adjusted
+    axes_adjusted, min_coordinate, max_coordinate = _adjust_bounding_box_to_real_axes(
         axes, min_coordinate, max_coordinate, output_axes_without_c
     )
+    if set(axes_adjusted) != set(output_axes_without_c):
+        raise ValueError("The axes of the bounding box must match the axes of the transformation.")
+
+    # let's get the bounding box corners and inverse transform then to the intrinsic coordinate system; since we are
+    # in case 1 or 5, the transformation is invertible
+    spatial_transform_bb_axes = Affine(
+        spatial_transform.to_affine_matrix(input_axes=input_axes_without_c, output_axes=axes_adjusted),
+        input_axes=input_axes_without_c,
+        output_axes=axes_adjusted,
+    )
 
-    # get the coordinates of the bounding box corners
     bounding_box_corners = get_bounding_box_corners(
         min_coordinate=min_coordinate,
         max_coordinate=max_coordinate,
-        axes=axes,
-    ).data
-
-    # transform the coordinates to the intrinsic coordinate system
-    intrinsic_axes = get_axes_names(element)
-    transform_to_intrinsic = transform_to_query_space.inverse().to_affine_matrix(  # type: ignore[union-attr]
-        input_axes=axes, output_axes=intrinsic_axes
+        axes=axes_adjusted,
     )
-    rotation_matrix = transform_to_intrinsic[0:-1, 0:-1]
-    translation = transform_to_intrinsic[0:-1, -1]
 
-    intrinsic_bounding_box_corners = bounding_box_corners @ rotation_matrix.T + translation
+    inverse = spatial_transform_bb_axes.inverse()
+    if not isinstance(inverse, Affine):
+        raise RuntimeError("This should not happen")
+    rotation_matrix = inverse.matrix[0:-1, 0:-1]
+    translation = inverse.matrix[0:-1, -1]
+
+    intrinsic_bounding_box_corners = bounding_box_corners.data @ rotation_matrix.T + translation
 
-    return intrinsic_bounding_box_corners, intrinsic_axes
+    return (
+        DataArray(
+            intrinsic_bounding_box_corners,
+            coords={"corner": range(len(bounding_box_corners)), "axis": list(inverse.output_axes)},
+        ),
+        input_axes_without_c,
+    )
 
 
 def _get_polygon_in_intrinsic_coordinates(
@@ -227,6 +248,11 @@ def _adjust_bounding_box_to_real_axes(
             M = np.finfo(np.float32).max - 1
             min_coordinate = np.append(min_coordinate, -M)
             max_coordinate = np.append(max_coordinate, M)
+    else:
+        indices = [axes_bb.index(ax) for ax in axes_out_without_c]
+        min_coordinate = min_coordinate[np.array(indices)]
+        max_coordinate = max_coordinate[np.array(indices)]
+        axes_bb = axes_out_without_c
     return axes_bb, min_coordinate, max_coordinate
 
 
@@ -407,6 +433,7 @@ def bounding_box_query(
     min_coordinate: list[Number] | ArrayLike,
     max_coordinate: list[Number] | ArrayLike,
     target_coordinate_system: str,
+    return_request_only: bool = False,
     filter_table: bool = True,
     **kwargs: Any,
 ) -> SpatialElement | SpatialData | None:
@@ -426,6 +453,9 @@ def bounding_box_query(
     filter_table
         If `True`, the table is filtered to only contain rows that are annotating regions
         contained within the bounding box.
+    return_request_only
+        If `True`, the function returns the bounding box coordinates in the target coordinate system.
+        Only valid with `DataArray` and `DataTree` elements.
 
     Returns
     -------
@@ -472,7 +502,8 @@ def _(
     min_coordinate: list[Number] | ArrayLike,
     max_coordinate: list[Number] | ArrayLike,
     target_coordinate_system: str,
-) -> DataArray | DataTree | None:
+    return_request_only: bool = False,
+) -> DataArray | DataTree | Mapping[str, slice] | None:
     """Implement bounding box query for Spatialdata supported DataArray.
 
     Notes
@@ -493,48 +524,11 @@ def _(
         max_coordinate=max_coordinate,
     )
 
-    # compute the output axes of the transformation, remove c from input and output axes, return the matrix without c
-    # and then build an affine transformation from that
-    m_without_c, input_axes_without_c, output_axes_without_c = _get_axes_of_tranformation(
-        image, target_coordinate_system
-    )
-    spatial_transform = Affine(m_without_c, input_axes=input_axes_without_c, output_axes=output_axes_without_c)
-
-    # we identified 5 cases (see the responsible function for details), cases 1 and 5 correspond to invertible
-    # transformations; we focus on them
-    m_without_c_linear = m_without_c[:-1, :-1]
-    case = _get_case_of_bounding_box_query(m_without_c_linear, input_axes_without_c, output_axes_without_c)
-    assert case in [1, 5]
-
-    # adjust the bounding box to the real axes, dropping or adding eventually mismatching axes; the order of the axes is
-    # not adjusted
-    axes, min_coordinate, max_coordinate = _adjust_bounding_box_to_real_axes(
-        axes, min_coordinate, max_coordinate, output_axes_without_c
-    )
-    assert set(axes) == set(output_axes_without_c)
-
-    # since the order of the axes is arbitrary, let's adjust the affine transformation without c to match those axes
-    spatial_transform_bb_axes = Affine(
-        spatial_transform.to_affine_matrix(input_axes=input_axes_without_c, output_axes=axes),
-        input_axes=input_axes_without_c,
-        output_axes=axes,
-    )
-
-    # let's get the bounding box corners and inverse transform then to the intrinsic coordinate system; since we are
-    # in case 1 or 5, the transformation is invertible
-    bounding_box_corners = get_bounding_box_corners(
-        min_coordinate=min_coordinate,
-        max_coordinate=max_coordinate,
-        axes=axes,
-    )
-    inverse = spatial_transform_bb_axes.inverse()
-    assert isinstance(inverse, Affine)
-    rotation_matrix = inverse.matrix[0:-1, 0:-1]
-    translation = inverse.matrix[0:-1, -1]
-    intrinsic_bounding_box_corners = DataArray(
-        bounding_box_corners.data @ rotation_matrix.T + translation,
-        coords={"corner": range(len(bounding_box_corners)), "axis": list(inverse.output_axes)},
+    intrinsic_bounding_box_corners, axes = _get_bounding_box_corners_in_intrinsic_coordinates(
+        image, axes, min_coordinate, max_coordinate, target_coordinate_system
     )
+    if TYPE_CHECKING:
+        assert isinstance(intrinsic_bounding_box_corners, DataArray)
 
     # build the request: now that we have the bounding box corners in the intrinsic coordinate system, we can use them
     # to build the request to query the raster data using the xarray APIs
@@ -555,6 +549,9 @@ def _(
         else:
             translation_vector.append(0)
 
+    if return_request_only:
+        return selection
+
     # query the data
     query_result = image.sel(selection)
     if isinstance(image, DataArray):
@@ -652,6 +649,7 @@ def _(
         max_coordinate=max_coordinate,
         target_coordinate_system=target_coordinate_system,
     )
+    intrinsic_bounding_box_corners = intrinsic_bounding_box_corners.data
     min_coordinate_intrinsic = intrinsic_bounding_box_corners.min(axis=0)
     max_coordinate_intrinsic = intrinsic_bounding_box_corners.max(axis=0)
 
@@ -722,14 +720,14 @@ def _(
     )
 
     # get the four corners of the bounding box
-    (intrinsic_bounding_box_corners, intrinsic_axes) = _get_bounding_box_corners_in_intrinsic_coordinates(
+    (intrinsic_bounding_box_corners, _) = _get_bounding_box_corners_in_intrinsic_coordinates(
         element=polygons,
         axes=axes,
         min_coordinate=min_coordinate,
         max_coordinate=max_coordinate,
         target_coordinate_system=target_coordinate_system,
     )
-
+    intrinsic_bounding_box_corners = intrinsic_bounding_box_corners.data
     bounding_box_non_axes_aligned = Polygon(intrinsic_bounding_box_corners)
     indices = polygons.geometry.intersects(bounding_box_non_axes_aligned)
     queried = polygons[indices]
@@ -841,6 +839,7 @@ def _(
     image: DataArray | DataTree,
     polygon: Polygon | MultiPolygon,
     target_coordinate_system: str,
+    return_request_only: bool = False,
     **kwargs: Any,
 ) -> DataArray | DataTree | None:
     _check_deprecated_kwargs(kwargs)
@@ -852,6 +851,7 @@ def _(
         max_coordinate=[max_x, max_y],
         axes=("x", "y"),
         target_coordinate_system=target_coordinate_system,
+        return_request_only=return_request_only,
     )