Backport PR #3987: ci: circumvent hatch incompatibility

.github/workflows/benchmark.yml

@@ -39,7 +39,7 @@ jobs:
           cache: 'pip'
 
       - name: Cache datasets
-        uses: actions/cache@v4
+        uses: actions/cache@v5
         with:
           path: |
             ~/.cache

.github/workflows/ci.yml

@@ -27,7 +27,7 @@ jobs:
         with: { enable-cache: false }
       - id: get-envs
         run: |
-          ENVS_JSON=$(NO_COLOR=1 uvx hatch env show --json | jq -c 'to_entries
+          ENVS_JSON=$(NO_COLOR=1 uvx '--with=virtualenv<21' hatch env show --json | jq -c 'to_entries
             | map(
               select(.key | startswith("hatch-test"))
               | {
@@ -55,15 +55,15 @@ jobs:
           python-version: ${{ matrix.env.python }}
 
       - name: Cache downloaded data
-        uses: actions/cache@v4
+        uses: actions/cache@v5
         with:
           path: .pytest_cache/d/scanpy-data
           key: pytest
 
       - name: Install dependencies
         run: |
           echo "::group::Install hatch"
-          uv tool install hatch
+          uv tool install hatch '--with=virtualenv<21'
           echo "::endgroup::"
           echo "::group::Create environment"
           hatch -v env create ${{ matrix.env.name }}

.readthedocs.yml

@@ -16,8 +16,8 @@ build:
       - asdf global uv latest
     pre_build:
       # run towncrier to preview the next version’s release notes
-      - ( find docs/release-notes -regex '[^.]+[.][^.]+.md' | grep -q . ) && uvx hatch run towncrier build --keep || true
+      - ( find docs/release-notes -regex '[^.]+[.][^.]+.md' | grep -q . ) && uvx "--with=virtualenv<21" hatch run towncrier build --keep || true
     build:
       html:
-        - uvx hatch run docs:build
+        - uvx "--with=virtualenv<21" hatch run docs:build
         - mv docs/_build $READTHEDOCS_OUTPUT

docs/conf.py

@@ -8,6 +8,7 @@
 from datetime import datetime
 from functools import partial
 from importlib.metadata import version as get_version
+from itertools import product
 from pathlib import Path, PurePosixPath
 from typing import TYPE_CHECKING
 
@@ -262,8 +263,10 @@ def setup(app: Sphinx) -> None:
     "pandas.core.series.Series": "pandas.Series",
     # https://github.com/pandas-dev/pandas/issues/63810
     "pandas.api.typing.aliases.AnyArrayLike": ("doc", "pandas:reference/aliases"),
-    "numpy.bool_": "numpy.bool",  # Since numpy 2, numpy.bool is the canonical dtype
-    "numpy.typing.ArrayLike": ("py:data", "numpy.typing.ArrayLike"),
+    **{
+        f"numpy.{prefix}typing.{name}": ("py:data", f"numpy.typing.{name}")
+        for name, prefix in product(["ArrayLike", "DTypeLike"], ["", "_"])
+    },
 }
 
 nitpick_ignore = [

pyproject.toml

@@ -216,7 +216,7 @@ lint.pylint.max-positional-args = 5
 "legacy_api_wrap.legacy_api".msg = "Use scanpy._compat.old_positionals instead"
 "numba.jit".msg = "Use `scanpy._compat.njit` instead"
 "numba.njit".msg = "Use `scanpy._compat.njit` instead"
-"numpy.bool".msg = "Use `np.bool_` instead for numpy>=1.24<2 compatibility"
+"numpy.bool_".msg = "Use `np.bool` instead"
 "pandas.api.types.is_categorical_dtype".msg = "Use isinstance(s.dtype, CategoricalDtype) instead"
 "pandas.value_counts".msg = "Use pd.Series(a).value_counts() instead"
 "pytest.importorskip".msg = "Use the “@needs” decorator/mark instead"

src/scanpy/_utils/__init__.py

@@ -801,7 +801,7 @@ def select_groups(
     adata: AnnData,
     groups_order_subset: Iterable[str] | Literal["all"] = "all",
     key: str = "groups",
-) -> tuple[list[str], NDArray[np.bool_]]:
+) -> tuple[list[str], NDArray[np.bool]]:
     """Get subset of groups in adata.obs[key]."""
     groups_order = adata.obs[key].cat.categories
     if f"{key}_masks" in adata.uns:

src/scanpy/get/_aggregated.py

@@ -55,7 +55,7 @@ def __init__(
         groupby: pd.Categorical,
         data: Array,
         *,
-        mask: NDArray[np.bool_] | None = None,
+        mask: NDArray[np.bool] | None = None,
     ) -> None:
         self.groupby = groupby
         if (missing := groupby.isna()).any():
@@ -188,7 +188,7 @@ def aggregate(  # noqa: PLR0912
     func: AggType | Iterable[AggType],
     *,
     axis: Literal["obs", 0, "var", 1] | None = None,
-    mask: NDArray[np.bool_] | str | None = None,
+    mask: NDArray[np.bool] | str | None = None,
     dof: int = 1,
     layer: str | None = None,
     obsm: str | None = None,
@@ -337,7 +337,7 @@ def _aggregate(
     by: pd.Categorical,
     func: AggType | Iterable[AggType],
     *,
-    mask: NDArray[np.bool_] | None = None,
+    mask: NDArray[np.bool] | None = None,
     dof: int = 1,
 ) -> dict[AggType, np.ndarray | DaskArray]:
     msg = f"Data type {type(data)} not supported for aggregation"
@@ -353,7 +353,7 @@ def aggregate_dask_mean_var(
     data: DaskArray,
     by: pd.Categorical,
     *,
-    mask: NDArray[np.bool_] | None = None,
+    mask: NDArray[np.bool] | None = None,
     dof: int = 1,
 ) -> MeanVarDict:
     mean = aggregate_dask(data, by, "mean", mask=mask, dof=dof)["mean"]
@@ -374,7 +374,7 @@ def aggregate_dask(
     by: pd.Categorical,
     func: AggType | Iterable[AggType],
     *,
-    mask: NDArray[np.bool_] | None = None,
+    mask: NDArray[np.bool] | None = None,
     dof: int = 1,
 ) -> dict[AggType, DaskArray]:
     if not isinstance(data._meta, CSBase | np.ndarray):
@@ -465,7 +465,7 @@ def aggregate_array(
     by: pd.Categorical,
     func: AggType | Iterable[AggType],
     *,
-    mask: NDArray[np.bool_] | None = None,
+    mask: NDArray[np.bool] | None = None,
     dof: int = 1,
 ) -> dict[AggType, np.ndarray]:
     groupby = Aggregate(groupby=by, data=data, mask=mask)
@@ -549,7 +549,7 @@ def _combine_categories(
 def sparse_indicator(
     categorical: pd.Categorical,
     *,
-    mask: NDArray[np.bool_] | None = None,
+    mask: NDArray[np.bool] | None = None,
     weight: NDArray[np.floating] | None = None,
 ) -> sparse.coo_matrix:
     if mask is not None and weight is None:

src/scanpy/get/get.py

@@ -475,7 +475,7 @@ def _set_obs_rep(
         raise AssertionError(msg)
 
 
-def _check_mask[M: NDArray[np.bool_] | NDArray[np.floating] | pd.Series | None](
+def _check_mask[M: NDArray[np.bool] | NDArray[np.floating] | pd.Series | None](
     data: AnnData | np.ndarray | CSBase | DaskArray,
     mask: str | M,
     dim: Literal["obs", "var"],

src/scanpy/metrics/_common.py

@@ -124,7 +124,7 @@ def _(val: pd.DataFrame | pd.Series) -> NDArray:
 
 def _vals_heterogeneous[V: NDArray | CSRBase](
     vals: V,
-) -> tuple[V, NDArray[np.bool_] | slice, NDArray[np.float64]]:
+) -> tuple[V, NDArray[np.bool] | slice, NDArray[np.float64]]:
     """Check that values wont cause issues in computation.
 
     Returns new set of vals, and indexer to put values back into result.

src/scanpy/plotting/_anndata.py

@@ -229,16 +229,16 @@ def _check_if_annotations(
         other_ax_obj, "var" if axis_name == "obs" else "obs"
     ).index
 
-    def is_annotation(needle: pd.Index) -> NDArray[np.bool_]:
+    def is_annotation(needle: pd.Index) -> NDArray[np.bool]:
         return needle.isin({None}) | needle.isin(annotations) | needle.isin(names)
 
     if not is_annotation(pd.Index([x, y])).all():
         return False
 
     color_idx = pd.Index(colors if colors is not None else [])
     # Colors are valid
-    color_valid: NDArray[np.bool_] = np.fromiter(
-        map(is_color_like, color_idx), dtype=np.bool_, count=len(color_idx)
+    color_valid: NDArray[np.bool] = np.fromiter(
+        map(is_color_like, color_idx), dtype=np.bool, count=len(color_idx)
     )
     # Annotation names are valid too
     color_valid[~color_valid] = is_annotation(color_idx[~color_valid])

src/scanpy/plotting/_tools/scatterplots.py

@@ -61,7 +61,7 @@ def embedding(  # noqa: PLR0912, PLR0913, PLR0915
     basis: str,
     *,
     color: str | Sequence[str] | None = None,
-    mask_obs: NDArray[np.bool_] | str | None = None,
+    mask_obs: NDArray[np.bool] | str | None = None,
     gene_symbols: str | None = None,
     use_raw: bool | None = None,
     sort_order: bool = True,
@@ -1204,7 +1204,7 @@ def _get_color_source_vector(
     adata: AnnData,
     value_to_plot: str,
     *,
-    mask_obs: NDArray[np.bool_] | None = None,
+    mask_obs: NDArray[np.bool] | None = None,
     use_raw: bool = False,
     gene_symbols: str | None = None,
     layer: str | None = None,

src/scanpy/preprocessing/_highly_variable_genes.py

@@ -313,7 +313,7 @@ def in_bounds(
         self,
         mean: NDArray[np.floating] | DaskArray,
         dispersion_norm: NDArray[np.floating] | DaskArray,
-    ) -> NDArray[np.bool_] | DaskArray:
+    ) -> NDArray[np.bool] | DaskArray:
         return (
             (mean > self.min_mean)
             & (mean < self.max_mean)
@@ -482,7 +482,7 @@ def _subset_genes(
     mean: NDArray[np.float64] | DaskArray,
     dispersion_norm: NDArray[np.float64] | DaskArray,
     cutoff: _Cutoffs | int,
-) -> NDArray[np.bool_] | DaskArray:
+) -> NDArray[np.bool] | DaskArray:
     """Get boolean mask of genes with normalized dispersion in bounds."""
     if isinstance(cutoff, _Cutoffs):
         dispersion_norm = np.nan_to_num(dispersion_norm)  # similar to Seurat

src/scanpy/preprocessing/_pca/__init__.py

@@ -66,7 +66,7 @@ def pca(  # noqa: PLR0912, PLR0913, PLR0915
     chunk_size: int | None = None,
     random_state: _LegacyRandom = 0,
     return_info: bool = False,
-    mask_var: NDArray[np.bool_] | str | None | Empty = _empty,
+    mask_var: NDArray[np.bool] | str | None | Empty = _empty,
     use_highly_variable: bool | None = None,
     dtype: DTypeLike = "float32",
     key_added: str | None = None,
@@ -403,7 +403,7 @@ def pca(  # noqa: PLR0912, PLR0913, PLR0915
 
 def _handle_mask_var(
     adata: AnnData,
-    mask_var: NDArray[np.bool_] | str | Empty | None,
+    mask_var: NDArray[np.bool] | str | Empty | None,
     *,
     obsm: str | None = None,
     use_highly_variable: bool | None,

src/scanpy/preprocessing/_scale.py

@@ -78,7 +78,7 @@ def scale[A: _Array](
     copy: bool = False,
     layer: str | None = None,
     obsm: str | None = None,
-    mask_obs: NDArray[np.bool_] | str | None = None,
+    mask_obs: NDArray[np.bool] | str | None = None,
 ) -> AnnData | A | None:
     """Scale data to unit variance and zero mean.
 
@@ -148,7 +148,7 @@ def scale_array[A: _Array](
     max_value: float | None = None,
     copy: bool = False,
     return_mean_std: bool = False,
-    mask_obs: NDArray[np.bool_] | None = None,
+    mask_obs: NDArray[np.bool] | None = None,
 ) -> (
     A
     | tuple[
@@ -175,7 +175,7 @@ def scale_array[A: _Array](
     mask_obs = (
         # For CSR matrices, default to a set mask to take the `scale_array_masked` path.
         # This is faster than the maskless `axis_mul_or_truediv` path.
-        np.ones(x.shape[0], dtype=np.bool_)
+        np.ones(x.shape[0], dtype=np.bool)
         if isinstance(x, CSRBase) and mask_obs is None and not zero_center
         else _check_mask(x, mask_obs, "obs")
     )
@@ -219,7 +219,7 @@ def scale_array[A: _Array](
 
 def scale_array_masked[A: _Array](
     x: A,
-    mask_obs: NDArray[np.bool_],
+    mask_obs: NDArray[np.bool],
     *,
     zero_center: bool = True,
     max_value: float | None = None,
@@ -268,7 +268,7 @@ def scale_and_clip_csr(
     data: NDArray[np.floating],
     *,
     std: NDArray[np.floating],
-    mask_obs: NDArray[np.bool_],
+    mask_obs: NDArray[np.bool],
     max_value: float | None,
 ) -> None:
     for i in numba.prange(len(indptr) - 1):
@@ -289,7 +289,7 @@ def scale_anndata(
     copy: bool = False,
     layer: str | None = None,
     obsm: str | None = None,
-    mask_obs: NDArray[np.bool_] | str | None = None,
+    mask_obs: NDArray[np.bool] | str | None = None,
 ) -> AnnData | None:
     adata = adata.copy() if copy else adata
     str_mean_std = ("mean", "std")

src/scanpy/preprocessing/_scrublet/core.py

@@ -89,7 +89,7 @@ class Scrublet:
 
     # Fields set by methods
 
-    predicted_doublets_: NDArray[np.bool_] | None = field(init=False)
+    predicted_doublets_: NDArray[np.bool] | None = field(init=False)
     """(shape: n_cells)
     Boolean mask of predicted doublets in the observed transcriptomes.
     """
@@ -354,7 +354,7 @@ def _nearest_neighbor_classifier(
         if use_approx_neighbors:
             neighbors = neighbors[:, 1:]
         # Calculate doublet score based on ratio of simulated cell neighbors vs. observed cell neighbors
-        doub_neigh_mask: NDArray[np.bool_] = (
+        doub_neigh_mask: NDArray[np.bool] = (
             manifold.obs["doub_labels"].to_numpy()[neighbors] == "sim"
         )
         n_sim_neigh: NDArray[np.int64] = doub_neigh_mask.sum(axis=1)
@@ -403,7 +403,7 @@ def _nearest_neighbor_classifier(
 
     def call_doublets(
         self, *, threshold: float | None = None, verbose: bool = True
-    ) -> NDArray[np.bool_] | None:
+    ) -> NDArray[np.bool] | None:
         """Call trancriptomes as doublets or singlets.
 
         Parameters

src/scanpy/preprocessing/_simple.py

@@ -853,11 +853,11 @@ def sample(
     fraction: float | None = None,
     *,
     n: int | None = None,
-    rng: RNGLike | SeedLike | None = 0,
+    rng: RNGLike | SeedLike | None = None,
     copy: Literal[False] = False,
     replace: bool = False,
     axis: Literal["obs", 0, "var", 1] = "obs",
-    p: str | NDArray[np.bool_] | NDArray[np.floating] | None = None,
+    p: str | NDArray[np.bool] | NDArray[np.floating] | None = None,
 ) -> None: ...
 @overload
 def sample(
@@ -869,7 +869,7 @@ def sample(
     copy: Literal[True],
     replace: bool = False,
     axis: Literal["obs", 0, "var", 1] = "obs",
-    p: str | NDArray[np.bool_] | NDArray[np.floating] | None = None,
+    p: str | NDArray[np.bool] | NDArray[np.floating] | None = None,
 ) -> AnnData: ...
 @overload
 def sample[A: np.ndarray | CSBase | DaskArray](
@@ -881,7 +881,7 @@ def sample[A: np.ndarray | CSBase | DaskArray](
     copy: bool = False,
     replace: bool = False,
     axis: Literal["obs", 0, "var", 1] = "obs",
-    p: str | NDArray[np.bool_] | NDArray[np.floating] | None = None,
+    p: str | NDArray[np.bool] | NDArray[np.floating] | None = None,
 ) -> tuple[A, NDArray[np.int64]]: ...
 def sample(  # noqa: PLR0912
     data: AnnData | np.ndarray | CSBase | DaskArray,
@@ -892,7 +892,7 @@ def sample(  # noqa: PLR0912
     copy: bool = False,
     replace: bool = False,
     axis: Literal["obs", 0, "var", 1] = "obs",
-    p: str | NDArray[np.bool_] | NDArray[np.floating] | None = None,
+    p: str | NDArray[np.bool] | NDArray[np.floating] | None = None,
 ) -> AnnData | None | tuple[np.ndarray | CSBase | DaskArray, NDArray[np.int64]]:
     r"""Sample observations or variables with or without replacement.
 

src/scanpy/tools/_rank_genes_groups.py

@@ -84,8 +84,8 @@ def _tiecorrect(rankvals: NDArray[np.number]) -> NDArray[np.float64]:
 def _ranks(
     x: NDArray[np.number] | CSBase,
     /,
-    mask_obs: NDArray[np.bool_] | None = None,
-    mask_obs_rest: NDArray[np.bool_] | None = None,
+    mask_obs: NDArray[np.bool] | None = None,
+    mask_obs_rest: NDArray[np.bool] | None = None,
 ) -> Generator[tuple[NDArray[np.float64], int, int], None, None]:
     n_genes = x.shape[1]
 
@@ -125,7 +125,7 @@ def __init__(
         groups: Iterable[str] | Literal["all"],
         groupby: str,
         *,
-        mask_var: NDArray[np.bool_] | None = None,
+        mask_var: NDArray[np.bool] | None = None,
         reference: Literal["rest"] | str = "rest",
         use_raw: bool = True,
         layer: str | None = None,
@@ -507,7 +507,7 @@ def rank_genes_groups(  # noqa: PLR0912, PLR0913, PLR0915
     adata: AnnData,
     groupby: str,
     *,
-    mask_var: NDArray[np.bool_] | str | None = None,
+    mask_var: NDArray[np.bool] | str | None = None,
     use_raw: bool | None = None,
     groups: Literal["all"] | Iterable[str] = "all",
     reference: str = "rest",