Merge branch 'main' into no-format

.gitignore

@@ -28,6 +28,7 @@
 # Python build files
 __pycache__/
 /src/scanpy/_version.py
+/ci/scanpy-min-deps.txt
 /dist/
 /*-env/
 /env-*/

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
 -   repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.7.0
+    rev: v0.8.2
     hooks:
     -   id: ruff
         types_or: [python, pyi, jupyter]

benchmarks/benchmarks/_utils.py

@@ -14,15 +14,16 @@
 import scanpy as sc
 
 if TYPE_CHECKING:
-    from collections.abc import Callable, Sequence, Set
+    from collections.abc import Callable, Sequence
+    from collections.abc import Set as AbstractSet
     from typing import Literal, Protocol, TypeVar
 
     from anndata import AnnData
 
     C = TypeVar("C", bound=Callable)
 
     class ParamSkipper(Protocol):
-        def __call__(self, **skipped: Set) -> Callable[[C], C]: ...
+        def __call__(self, **skipped: AbstractSet) -> Callable[[C], C]: ...
 
     Dataset = Literal["pbmc68k_reduced", "pbmc3k", "bmmc", "lung93k"]
     KeyX = Literal[None, "off-axis"]
@@ -195,7 +196,7 @@ def param_skipper(
     b 5
     """
 
-    def skip(**skipped: Set) -> Callable[[C], C]:
+    def skip(**skipped: AbstractSet) -> Callable[[C], C]:
         skipped_combs = [
             tuple(record.values())
             for record in (

ci/scripts/min-deps.py

@@ -1,9 +1,17 @@
 #!/usr/bin/env python3
+# /// script
+# dependencies = [
+#   "tomli; python_version < '3.11'",
+#   "packaging",
+# ]
+# ///
+
 from __future__ import annotations
 
 import argparse
 import sys
 from collections import deque
+from contextlib import ExitStack
 from pathlib import Path
 from typing import TYPE_CHECKING
 
@@ -16,7 +24,7 @@
 from packaging.version import Version
 
 if TYPE_CHECKING:
-    from collections.abc import Generator, Iterable
+    from collections.abc import Generator, Iterable, Sequence
 
 
 def min_dep(req: Requirement) -> Requirement:
@@ -27,18 +35,21 @@ def min_dep(req: Requirement) -> Requirement:
     -------
 
     >>> min_dep(Requirement("numpy>=1.0"))
-    "numpy==1.0"
+    <Requirement('numpy==1.0.*')>
     """
     req_name = req.name
     if req.extras:
         req_name = f"{req_name}[{','.join(req.extras)}]"
 
-    if not req.specifier:
+    filter_specs = [
+        spec for spec in req.specifier if spec.operator in {"==", "~=", ">=", ">"}
+    ]
+    if not filter_specs:
         return Requirement(req_name)
 
     min_version = Version("0.0.0.a1")
-    for spec in req.specifier:
-        if spec.operator in [">", ">=", "~="]:
+    for spec in filter_specs:
+        if spec.operator in {">", ">=", "~="}:
             min_version = max(min_version, Version(spec.version))
         elif spec.operator == "==":
             min_version = Version(spec.version)
@@ -65,12 +76,19 @@ def extract_min_deps(
             yield min_dep(req)
 
 
-def main():
+class Args(argparse.Namespace):
+    path: Path
+    output: Path | None
+    extras: list[str]
+
+
+def main(argv: Sequence[str] | None = None) -> None:
     parser = argparse.ArgumentParser(
         prog="min-deps",
-        description="""Parse a pyproject.toml file and output a list of minimum dependencies.
-
-        Output is directly passable to `pip install`.""",
+        description=(
+            "Parse a pyproject.toml file and output a list of minimum dependencies. "
+            "Output is optimized for `[uv] pip install` (see `-o`/`--output` for details)."
+        ),
         usage="pip install `python min-deps.py pyproject.toml`",
     )
     parser.add_argument(
@@ -79,8 +97,18 @@ def main():
     parser.add_argument(
         "--extras", type=str, nargs="*", default=(), help="extras to install"
     )
+    parser.add_argument(
+        *("--output", "-o"),
+        type=Path,
+        default=None,
+        help=(
+            "output file (default: stdout). "
+            "Without this option, output is space-separated for direct passing to `pip install`. "
+            "With this option, output written to a file newline-separated file usable as `requirements.txt` or `constraints.txt`."
+        ),
+    )
 
-    args = parser.parse_args()
+    args = parser.parse_args(argv, Args())
 
     pyproject = tomllib.loads(args.path.read_text())
 
@@ -92,7 +120,10 @@ def main():
 
     min_deps = extract_min_deps(deps, pyproject=pyproject)
 
-    print(" ".join(map(str, min_deps)))
+    sep = "\n" if args.output else " "
+    with ExitStack() as stack:
+        f = stack.enter_context(args.output.open("w")) if args.output else sys.stdout
+        print(sep.join(map(str, min_deps)), file=f)
 
 
 if __name__ == "__main__":

ci/scripts/towncrier_automation.py

@@ -1,4 +1,8 @@
 #!/usr/bin/env python3
+# /// script
+# dependencies = [ "towncrier", "packaging" ]
+# ///
+
 from __future__ import annotations
 
 import argparse
@@ -62,9 +66,7 @@ def main(argv: Sequence[str] | None = None) -> None:
         text=True,
         check=True,
     ).stdout.strip()
-    pr_description = (
-        "" if base_branch == "main" else "@meeseeksmachine backport to main"
-    )
+    pr_description = "" if base_branch == "main" else "@meeseeksdev backport to main"
     branch_name = f"release_notes_{args.version}"
 
     # Create a new branch + commit

docs/release-notes/1.10.4.md

@@ -0,0 +1,17 @@
+(v1.10.4)=
+### 1.10.4 {small}`2024-11-12`
+
+### Breaking changes
+
+- Remove Python 3.9 support {smaller}`P Angerer` ({pr}`3283`)
+
+### Bug fixes
+
+- Fix {meth}`scanpy.pl.DotPlot.style`, {meth}`scanpy.pl.MatrixPlot.style`, and {meth}`scanpy.pl.StackedViolin.style` resetting all non-specified parameters {smaller}`P Angerer` ({pr}`3206`)
+- Accept `'group'` instead of `'obs'` for `standard_scale` parameter in {func}`~scanpy.pl.stacked_violin` {smaller}`P Angerer` ({pr}`3243`)
+- Use `density_norm` instead of of `scale` (cont. from {pr}`2844`) in {func}`~scanpy.pl.violin` and {func}`~scanpy.pl.stacked_violin` {smaller}`P Angerer` ({pr}`3244`)
+- Switched all compatibility adapters for positional parameters to {exc}`FutureWarning` {smaller}`P Angerer` ({pr}`3264`)
+- Catch `PerfectSeparationWarning` during {func}`~scanpy.pp.regress_out` {smaller}`J Wagner` ({pr}`3275`)
+- Fix {func}`scanpy.pp.highly_variable_genes` for batches of size 1 {smaller}`P Angerer` ({pr}`3286`)
+- Fix {func}`scanpy.pl.scatter`’s `color` parameter to take collections as advertised {smaller}`P Angerer` ({pr}`3299`)
+- Fix {func}`scanpy.pl.highest_expr_genes` when used with a categorical gene symbol column {smaller}`P Angerer` ({pr}`3302`)

docs/release-notes/3284.performance.md

@@ -0,0 +1 @@
+* Speed up {func}`~scanpy.pp.regress_out` {smaller}`P Ashish, P Angerer & S Dicks`

docs/release-notes/3307.feature.md

@@ -0,0 +1 @@
+Add support {class}`dask.array.Array` to {func}`scanpy.pp.calculate_qc_metrics` {smaller}`I Gold`

docs/release-notes/3324.feature.md

@@ -0,0 +1 @@
+Support `layer` parameter in {func}`scanpy.pl.highest_expr_genes` {smaller}`P Angerer`

docs/release-notes/3335.feature.md

@@ -0,0 +1 @@
+Run numba functions single-threaded when called from inside of a ThreadPool {smaller}`P Angerer`

docs/release-notes/3362.doc.md

@@ -0,0 +1 @@
+Improve {func}`~scanpy.external.pp.harmony_integrate` docs {smaller}`D Kühl`

docs/release-notes/3380.bugfix.md

@@ -0,0 +1 @@
+Raise {exc}`FutureWarning` when calling deprecated {mod}`scanpy.pp` functions {smaller}`P Angerer`

docs/release-notes/3393.bugfix.md

@@ -0,0 +1 @@
+Upper-bound {mod}`sklearn` `<1.6.0` due to {issue}`dask/dask-ml#1002` {smaller}`Ilan Gold`

hatch.toml

@@ -19,14 +19,17 @@ features = ["test", "dask-ml"]
 extra-dependencies = ["ipykernel"]
 overrides.matrix.deps.env-vars = [
     { if = ["pre"], key = "UV_PRERELEASE", value = "allow" },
-    { if = ["min"], key = "UV_RESOLUTION", value = "lowest-direct" },
+    { if = ["min"], key = "UV_CONSTRAINT", value = "ci/scanpy-min-deps.txt" },
+]
+overrides.matrix.deps.pre-install-commands = [
+    { if = ["min"], value = "uv run ci/scripts/min-deps.py pyproject.toml -o ci/scanpy-min-deps.txt" },
 ]
 overrides.matrix.deps.python = [
-    { if = ["min"] , value = "3.10" },
+    { if = ["min"], value = "3.10" },
     { if = ["stable", "full", "pre"], value = "3.12" },
 ]
 overrides.matrix.deps.features = [
-    { if = ["full"] , value = "test-full" },
+    { if = ["full"], value = "test-full" },
 ]
 
 [[envs.hatch-test.matrix]]

pyproject.toml

@@ -52,11 +52,11 @@ dependencies = [
     "pandas >=1.5",
     "scipy>=1.8",
     "seaborn>=0.13",
-    "h5py>=3.6",
+    "h5py>=3.7",
     "tqdm",
-    "scikit-learn>=1.1",
+    "scikit-learn>=1.1,<1.6.0",
     "statsmodels>=0.13",
-    "patsy",
+    "patsy!=1.0.0", # https://github.com/pydata/patsy/issues/215
     "networkx>=2.7",
     "natsort",
     "joblib",
@@ -66,7 +66,7 @@ dependencies = [
     "packaging>=21.3",
     "session-info",
     "legacy-api-wrap>=1.4",  # for positional API deprecations
-    "get-annotations; python_version < '3.10'",
+    "typing-extensions; python_version < '3.13'",
 ]
 dynamic = ["version"]
 
@@ -124,7 +124,7 @@ doc = [
     "ipython>=7.20",  # for nbsphinx code highlighting
     "matplotlib!=3.6.1",
     "sphinxcontrib-bibtex",
-    "setuptools",
+    "setuptools", # undeclared dependency of sphinxcontrib-bibtex→pybtex
     # TODO: remove necessity for being able to import doc-linked classes
     "scanpy[paga,dask-ml]",
     "sam-algorithm",
@@ -167,7 +167,7 @@ addopts = [
     "-ptesting.scanpy._pytest",
     "--pyargs",
 ]
-testpaths = ["./tests", "scanpy"]
+testpaths = ["./tests", "./ci", "scanpy"]
 norecursedirs = ["tests/_images"]
 xfail_strict = true
 nunit_attach_on = "fail"
@@ -212,7 +212,7 @@ exclude_also = [
     "if __name__ == .__main__.:",
     "if TYPE_CHECKING:",
     # https://github.com/numba/numba/issues/4268
-    "@numba.njit.*",
+    '@(numba\.|nb\.)njit.*',
 ]
 
 [tool.ruff]
@@ -232,6 +232,7 @@ select = [
     "TID251",  # Banned imports
     "ICN",     # Follow import conventions
     "PTH",     # Pathlib instead of os.path
+    "PYI",     # Typing
     "PLR0917", # Ban APIs with too many positional parameters
     "FBT",     # No positional boolean parameters
     "PT",      # Pytest style
@@ -246,6 +247,8 @@ ignore = [
     "E262",
     # allow I, O, l as variable names -> I is the identity matrix, i, j, k, l is reasonable indexing notation
     "E741",
+    # `Literal["..."] | str` is useful for autocompletion
+    "PYI051",
 ]
 [tool.ruff.lint.per-file-ignores]
 # Do not assign a lambda expression, use a def
@@ -259,6 +262,8 @@ required-imports = ["from __future__ import annotations"]
 "pandas.value_counts".msg = "Use pd.Series(a).value_counts() instead"
 "legacy_api_wrap.legacy_api".msg = "Use scanpy._compat.old_positionals instead"
 "numpy.bool".msg = "Use `np.bool_` instead for numpy>=1.24<2 compatibility"
+"numba.jit".msg = "Use `scanpy._compat.njit` instead"
+"numba.njit".msg = "Use `scanpy._compat.njit` instead"
 [tool.ruff.lint.flake8-type-checking]
 exempt-modules = []
 strict = true