Merge branch 'main' into move-to-nanobind

Author: Intron7

.readthedocs.yml

@@ -5,20 +5,17 @@ submodules:
 build:
   os: ubuntu-24.04
   tools:
-    python: "3.12"
+    python: '3.13'
+  jobs:
+    post_checkout:
+      # unshallow so version can be derived from tag
+      - git fetch --unshallow || true
+    create_environment:
+      - asdf plugin add uv
+      - asdf install uv latest
+      - asdf global uv latest
+    build:
+      html:
+        - CMAKE_ARGS="-DRSC_BUILD_EXTENSIONS=OFF" uvx hatch run docs:build
+        - mv docs/_build $READTHEDOCS_OUTPUT
 
-  commands:
-    # Install and set up uv
-    - asdf plugin add uv
-    - asdf install uv latest
-    - asdf global uv latest
-
-    # Use uv to synchronize dependencies
-    - CMAKE_ARGS="-DRSC_BUILD_EXTENSIONS=OFF" uv pip install --system ".[doc]"
-
-    # Build documentation using sphinx
-    - python -m sphinx -T -b html -d docs/_build/doctrees -D language=en docs $READTHEDOCS_OUTPUT/html
-
-sphinx:
-  configuration: docs/conf.py
-  fail_on_warning: true

docs/Installation.md

@@ -1,29 +1,52 @@
 # Installation
 ## Conda
-The easiest way to install *rapids-singlecell* is to use one of the *yaml* file provided in the [conda](https://github.com/scverse/rapids_singlecell/tree/main/conda) folder. These *yaml* files install everything needed to run the example notebooks and get you started.
+The easiest way to install *rapids-singlecell* is to use one of the *yaml* files provided in the [conda](https://github.com/scverse/rapids_singlecell/tree/main/conda) folder. These *yaml* files install everything needed to run the example notebooks and get you started.
+
+`````{tab-set}
+````{tab-item} CUDA 13
+```bash
+conda env create -f conda/rsc_rapids_25.12.yml
+# or
+mamba env create -f conda/rsc_rapids_25.12.yml
 ```
-conda env create -f conda/rsc_rapids_25.12.yml #default CUDA-13.0 PYTHON-3.13
+*Python 3.13, CUDA 13.0*
+````
+````{tab-item} CUDA 12
+```bash
+conda env create -f conda/rsc_rapids_25.10.yml
 # or
-mamba env create -f conda/rsc_rapids_25.10.yml #default CUDA-12.9 PYTHON-3.13
+mamba env create -f conda/rsc_rapids_25.10.yml
 ```
+*Python 3.13, CUDA 12.9*
+````
+`````
+
 ```{note}
 RAPIDS currently doesn't support `channel_priority: strict`; use `channel_priority: flexible` instead
 ```
 
 ## PyPI
 *rapids-singlecell* is also on PyPI.
-```
+```bash
 pip install rapids-singlecell
 ```
 The default installer doesn't cover RAPIDS nor CuPy. Information on how to install RAPIDS & CuPy can be found [here](https://rapids.ai/start.html).
 
-If you want to use RAPIDS new PyPI packages, the whole library with all dependencies can be install with:
-```
-uv pip install 'rapids-singlecell[rapids12]' --extra-index-url=https://pypi.nvidia.com #CUDA12
-uv pip install 'rapids-singlecell[rapids13]' --extra-index-url=https://pypi.nvidia.com #CUDA13
-
+If you want to use RAPIDS new PyPI packages, the whole library with all dependencies can be installed with:
 
+`````{tab-set}
+````{tab-item} CUDA 13
+```bash
+uv pip install 'rapids-singlecell[rapids13]' --extra-index-url=https://pypi.nvidia.com
 ```
+````
+````{tab-item} CUDA 12
+```bash
+uv pip install 'rapids-singlecell[rapids12]' --extra-index-url=https://pypi.nvidia.com
+```
+````
+`````
+
 It is important to ensure that the CUDA environment is set up correctly so that RAPIDS and CuPy can locate the necessary libraries.
 
 ```{note}

docs/conf.py

@@ -58,13 +58,13 @@
     "sphinx.ext.githubpages",
     "sphinx_autodoc_typehints",
     "sphinx.ext.extlinks",
-    "readthedocs_ext.readthedocs",
     "sphinx.ext.imgconverter",
     "sphinx_copybutton",
     "nbsphinx",
     "scanpydoc",
     "sphinx.ext.linkcode",
-    "sphinx_copybutton",
+    "sphinx_tabs.tabs",
+    "sphinxext.opengraph",
 ]
 
 autosummary_generate = True
@@ -130,7 +130,14 @@
 # List of patterns, relative to source directory, that match files and
 # directories to ignore when looking for source files.
 # This pattern also affects html_static_path and html_extra_path.
-exclude_patterns = ["_build", "Thumbs.db", ".DS_Store", "._*", "*.ipynb_checkpoints"]
+exclude_patterns = [
+    "_build",
+    "Thumbs.db",
+    ".DS_Store",
+    "._*",
+    "*.ipynb_checkpoints",
+    "release-notes/blank.md",
+]
 
 
 # -- Options for HTML output -------------------------------------------------
@@ -152,6 +159,11 @@
 html_css_files = ["_static/css/override.css"]
 html_title = "rapids-singlecell"
 
+# OpenGraph metadata for social media previews
+ogp_site_url = "https://rapids-singlecell.readthedocs.io/"
+ogp_site_name = "rapids-singlecell"
+ogp_image = "_static/logo_RTD.svg"
+
 qualname_overrides = {
     "numpy.bool_": "numpy.bool",  # Since numpy 2, numpy.bool is the canonical dtype
 }
@@ -163,6 +175,7 @@
     ("py:class", "anndata._core.views.ArrayView"),
     ("py:class", "anndata._core.raw.Raw"),
     ("py:class", "scanpy._utils.Empty"),
+    ("py:data", "typing.Union"),
     *[
         ("py:class", f"anndata._core.aligned_mapping.{cls}{kind}")
         for cls in "Layers AxisArrays PairwiseArrays".split()

docs/release-notes/0.13.0.md

@@ -5,7 +5,7 @@
 * Add support for aggregate operations on CSC matrices, Fortran-ordered arrays, and Dask with sparse CSR and dense matrices {pr}`395` {smaller}`S Dicks`
 * Adds dask support for `tl.score_genes` & `tl.score_genes_cell_cycle` {pr}`408` {smaller}`S Dicks`
 * Adds dask support for `tl.rank_genes_groups_logreg` {pr}`413` {smaller}`S Dicks`
-* Updates {mod}`dcg` to work `Decoupler 2.0.0` {pr}`418` {smaller}`S Dicks`
+* Updates {mod}`rapids_singlecell.dcg` to work with `Decoupler 2.0.0` {pr}`418` {smaller}`S Dicks`
 
 ```{rubric} Performance
 ```

pyproject.toml

@@ -42,9 +42,9 @@ doc = [
     "nbsphinx>=0.8.12",
     "myst-nb",
     "sphinx-design",
+    "sphinx-tabs",
+    "sphinxext-opengraph",
     "scanpydoc[typehints,theme]>=0.9.4",
-    "readthedocs-sphinx-ext",
-    "sphinx_copybutton",
     "dask",
     "pytest",
 ]

src/rapids_singlecell/decoupler_gpu/_method_waggr.py

@@ -175,7 +175,7 @@ def _func_waggr(
     verbose: bool = False,
 ) -> tuple[np.ndarray, np.ndarray]:
     r"""
-    Weighted Aggregate (WAGGR) :cite:`decoupler`.
+    Weighted Aggregate (WAGGR).
 
     This approach aggregates the molecular features :math:`x_i` from one observation :math:`i` with
     the feature weights :math:`w` of a given feature set :math:`j` into an enrichment score :math:`ES`.

src/rapids_singlecell/decoupler_gpu/_method_zscore.py

@@ -18,7 +18,7 @@ def _func_zscore(
     verbose: bool = False,
 ) -> tuple[cp.ndarray, cp.ndarray]:
     r"""
-    Z-score (ZSCORE) :cite:`zscore`.
+    Z-score (ZSCORE).
 
     This approach computes the mean value of the molecular features for known targets,
     optionally subtracts the overall mean of all measured features,

src/rapids_singlecell/preprocessing/_neighbors/__init__.py

@@ -187,12 +187,12 @@ def neighbors(
         * 'n_lists': Number of inverted lists for IVF indexing. Default is 2 * next_power_of_2(sqrt(n_samples)).
 
         * 'n_probes': Number of lists to probe during search. Default is 20. Higher values
-        increase accuracy but reduce speed.
+          increase accuracy but reduce speed.
 
         For `nn_descent` algorithm, the following parameters can be specified:
 
         * 'intermediate_graph_degree': The degree of the intermediate graph. Default is None.
-        It is recommended to set it to `>= 1.5 * n_neighbors`.
+          It is recommended to set it to `>= 1.5 * n_neighbors`.
 
         For `all_neighbors` algorithm, the following parameters can be specified:
 
@@ -213,7 +213,7 @@ def neighbors(
         * 'n_lists': Number of inverted lists for IVF indexing. Default is 2 * next_power_of_2(sqrt(n_samples)).
 
         * 'n_probes': Number of lists to probe during search. Default is 20. Higher values
-        increase accuracy but reduce speed.
+          increase accuracy but reduce speed.
 
     key_added
         If not specified, the neighbors data is stored in .uns['neighbors'],
@@ -400,7 +400,7 @@ def bbknn(
         * 'n_lists': Number of inverted lists for IVF indexing. Default is 2 * next_power_of_2(sqrt(n_samples)).
 
         * 'n_probes': Number of lists to probe during search. Default is 20. Higher values
-        increase accuracy but reduce speed.
+          increase accuracy but reduce speed.
 
     trim
         Trim the neighbours of each cell to these many top connectivities.

src/rapids_singlecell/squidpy_gpu/_co_oc.py

@@ -53,9 +53,9 @@ def co_occurrence(
     Otherwise, modifies the ``adata`` with the following keys:
 
         - :attr:`anndata.AnnData.uns` ``['{cluster_key}_co_occurrence']['occ']`` - the co-occurrence probabilities
-        across interval thresholds.
+          across interval thresholds.
         - :attr:`anndata.AnnData.uns` ``['{cluster_key}_co_occurrence']['interval']`` - the distance thresholds
-        computed at ``interval``.
+          computed at ``interval``.
     """
 
     _assert_categorical_obs(adata, key=cluster_key)

src/rapids_singlecell/tools/_clustering.py

@@ -466,9 +466,9 @@ def kmeans(
             computed with default parameters or `n_pcs` if present.
         n_init
             Number of initializations to run the KMeans algorithm
-        random_state: float (default: 42)
+        random_state
             if you want results to be the same when you restart Python, select a
-            state.
+            state. Default is 42.
         key_added
             `adata.obs` key under which to add the cluster labels.
         copy