Python 3.10+ & use uv for docs & fix RTD & support numpy 2 (#830)

* uv docs

Signed-off-by: zethson <lukas.heumos@posteo.net>

* submodule

Signed-off-by: zethson <lukas.heumos@posteo.net>

* fix types

Signed-off-by: zethson <lukas.heumos@posteo.net>

* Install scvelo from github

Signed-off-by: zethson <lukas.heumos@posteo.net>

* cellrank from github

Signed-off-by: zethson <lukas.heumos@posteo.net>

* cellrank fix?

Signed-off-by: zethson <lukas.heumos@posteo.net>

* cellrank out

Signed-off-by: zethson <lukas.heumos@posteo.net>

---------

Signed-off-by: zethson <lukas.heumos@posteo.net>

.github/workflows/run_notebooks.yml

@@ -13,7 +13,7 @@ jobs:
                         "docs/tutorials/notebooks/ehrapy_introduction.ipynb",
                         "docs/tutorials/notebooks/mimic_2_introduction.ipynb",
                         "docs/tutorials/notebooks/mimic_2_survival_analysis.ipynb",
-                        "docs/tutorials/notebooks/mimic_2_fate.ipynb",
+                        # "docs/tutorials/notebooks/mimic_2_fate.ipynb",  # https://github.com/theislab/cellrank/issues/1235
                         "docs/tutorials/notebooks/mimic_2_causal_inference.ipynb",
                         # "docs/tutorials/notebooks/mimic_3_demo.ipynb",
                         # "docs/tutorials/notebooks/medcat.ipynb",
@@ -34,5 +34,8 @@ jobs:
             - name: Install ehrapy and additional dependencies
               run: uv pip install --system . cellrank nbconvert ipykernel
 
+            - name: Install scvelo from Github
+              run: uv pip install --system git+https://github.com/theislab/scvelo.git
+
             - name: Run ${{ matrix.notebook }} Notebook
               run: jupyter nbconvert --to notebook --execute ${{ matrix.notebook }}

.readthedocs.yml

@@ -3,16 +3,13 @@ build:
     os: ubuntu-22.04
     tools:
         python: "3.11"
-    jobs:
-        pre_build:
-            - python -c "import ehrapy"
-            - pip freeze
-        post_create_environment:
-            - pip install uv
-        post_install:
-            # VIRTUAL_ENV needs to be set manually for now.
-            # See https://github.com/readthedocs/readthedocs.org/pull/11152/
-            - VIRTUAL_ENV=$READTHEDOCS_VIRTUALENV_PATH pip install .[docs]
+    commands:
+        - asdf plugin add uv
+        - asdf install uv latest
+        - asdf global uv latest
+        - uv venv
+        - uv pip install .[docs]
+        - .venv/bin/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: false

docs/_ext/edit_on_github.py

@@ -20,7 +20,7 @@ def get_github_repo(app: Sphinx, path: str) -> str:
 
 
 def _html_page_context(
-    app: Sphinx, _pagename: str, templatename: str, context: dict[str, Any], doctree: Optional[Any]
+    app: Sphinx, _pagename: str, templatename: str, context: dict[str, Any], doctree: Any | None
 ) -> None:
     # doctree is None - otherwise viewcode fails
     if templatename != "page.html" or doctree is None:

ehrapy/_utils_doc.py

@@ -1,9 +1,9 @@
 import inspect
+from collections.abc import Callable
 from textwrap import dedent
-from typing import Callable, Optional, Union
 
 
-def getdoc(c_or_f: Union[Callable, type]) -> Optional[str]:  # pragma: no cover
+def getdoc(c_or_f: Callable | type) -> str | None:  # pragma: no cover
     if getattr(c_or_f, "__doc__", None) is None:
         return None
     doc = inspect.getdoc(c_or_f)

ehrapy/anndata/anndata_ext.py

@@ -404,7 +404,7 @@ def _detect_binary_columns(df: pd.DataFrame, numerical_columns: list[str]) -> li
     for column in numerical_columns:
         # checking for float and int as well as NaNs (this is safe since checked columns are numericals only)
         # only columns that contain at least one 0 and one 1 are counted as binary (or 0.0/1.0)
-        if df[column].isin([0.0, 1.0, np.NaN, 0, 1]).all() and df[column].nunique() == 2:
+        if df[column].isin([0.0, 1.0, np.nan, 0, 1]).all() and df[column].nunique() == 2:
             binary_columns.append(column)
 
     return binary_columns
@@ -423,7 +423,7 @@ def _cast_obs_columns(obs: pd.DataFrame) -> pd.DataFrame:
     # type cast each non-numerical column to either bool (if possible) or category else
     obs[object_columns] = obs[object_columns].apply(
         lambda obs_name: obs_name.astype("category")
-        if not set(pd.unique(obs_name)).issubset({False, True, np.NaN})
+        if not set(pd.unique(obs_name)).issubset({False, True, np.nan})
         else obs_name.astype("bool"),
         axis=0,
     )

ehrapy/data/_datasets.py

@@ -743,7 +743,7 @@ def synthea_1k_sample(
 
     df = anndata_to_df(adata)
     df.drop(
-        columns=[col for col in df.columns if any(isinstance(x, (list, dict)) for x in df[col].dropna())], inplace=True
+        columns=[col for col in df.columns if any(isinstance(x, list | dict) for x in df[col].dropna())], inplace=True
     )
     df.drop(columns=df.columns[df.isna().all()], inplace=True)
     adata = df_to_anndata(df, index_column="id")

ehrapy/plot/_scanpy_pl_api.py

@@ -1,10 +1,10 @@
 from __future__ import annotations
 
-from collections.abc import Collection, Iterable, Mapping, Sequence
+from collections.abc import Callable, Collection, Iterable, Mapping, Sequence
 from enum import Enum
 from functools import partial
 from types import MappingProxyType
-from typing import TYPE_CHECKING, Any, Callable, Literal, Union
+from typing import TYPE_CHECKING, Any, Literal
 
 import scanpy as sc
 from scanpy.plotting import DotPlot, MatrixPlot, StackedViolin
@@ -36,12 +36,12 @@
     from scanpy.plotting._utils import _AxesSubplot
 
 _Basis = Literal["pca", "tsne", "umap", "diffmap", "draw_graph_fr"]
-_VarNames = Union[str, Sequence[str]]
-ColorLike = Union[str, tuple[float, ...]]
+_VarNames = str | Sequence[str]
+ColorLike = str | tuple[float, ...]
 _IGraphLayout = Literal["fa", "fr", "rt", "rt_circular", "drl", "eq_tree", ...]  # type: ignore
 _FontWeight = Literal["light", "normal", "medium", "semibold", "bold", "heavy", "black"]
 _FontSize = Literal["xx-small", "x-small", "small", "medium", "large", "x-large", "xx-large"]
-VBound = Union[str, float, Callable[[Sequence[float]], float]]
+VBound = str | float | Callable[[Sequence[float]], float]
 
 
 @_doc_params(scatter_temp=doc_scatter_basic, show_save_ax=doc_show_save_ax)

ehrapy/preprocessing/_encoding.py

@@ -73,7 +73,7 @@ def encode(
 
     if isinstance(encodings, str) and not autodetect:
         raise ValueError("Passing a string for parameter encodings is only possible when using autodetect=True!")
-    elif autodetect and not isinstance(encodings, (str, type(None))):
+    elif autodetect and not isinstance(encodings, str | type(None)):
         raise ValueError(
             f"Setting encode mode with autodetect=True only works by passing a string (encode mode name) or None not {type(encodings)}!"
         )
@@ -630,7 +630,7 @@ def _update_obs(adata: AnnData, categorical_names: list[str]) -> pd.DataFrame:
             updated_obs[var_name] = adata.X[::, idx : idx + 1].flatten()
             # note: this will count binary columns (0 and 1 only) as well
             # needed for writing to .h5ad files
-            if set(pd.unique(updated_obs[var_name])).issubset({False, True, np.NaN}):
+            if set(pd.unique(updated_obs[var_name])).issubset({False, True, np.nan}):
                 updated_obs[var_name] = updated_obs[var_name].astype("bool")
     # get all non bool object columns and cast them to category dtype
     object_columns = list(updated_obs.select_dtypes(include="object").columns)

ehrapy/preprocessing/_imputation.py

@@ -63,7 +63,7 @@ def explicit_impute(
         _warn_imputation_threshold(adata, var_names=replacement.keys(), threshold=warning_threshold)  # type: ignore
 
     # 1: Replace all missing values with the specified value
-    if isinstance(replacement, (int, str)):
+    if isinstance(replacement, int | str):
         _replace_explicit(adata.X, replacement, impute_empty_strings)
 
     # 2: Replace all missing values in a subset of columns with a specified value per column or a default value, when the column is not explicitly named

ehrapy/preprocessing/_normalization.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from typing import TYPE_CHECKING, Callable
+from typing import TYPE_CHECKING
 
 import numpy as np
 import sklearn.preprocessing as sklearn_pp
@@ -20,7 +20,7 @@
 )
 
 if TYPE_CHECKING:
-    from collections.abc import Sequence
+    from collections.abc import Callable, Sequence
 
     import pandas as pd
     from anndata import AnnData

ehrapy/preprocessing/_scanpy_pp_api.py

@@ -1,7 +1,8 @@
 from __future__ import annotations
 
+from collections.abc import Callable
 from types import MappingProxyType
-from typing import TYPE_CHECKING, Any, Callable, Literal, Union
+from typing import TYPE_CHECKING, Any, Literal, TypeAlias
 
 import numpy as np
 import scanpy as sc
@@ -15,7 +16,7 @@
 
     from ehrapy.preprocessing._types import KnownTransformer
 
-AnyRandom = Union[int, np.random.RandomState, None]
+AnyRandom: TypeAlias = int | np.random.RandomState | None
 
 
 def pca(
@@ -193,7 +194,7 @@ def combat(
     "sqeuclidean",
     "yule",
 ]
-_Metric = Union[_MetricSparseCapable, _MetricScipySpatial]
+_Metric = _MetricSparseCapable | _MetricScipySpatial
 
 
 def neighbors(

ehrapy/tools/_method_options.py

@@ -1,11 +1,11 @@
-from typing import Literal, Optional
+from typing import Literal
 
 _InitPos = Literal["paga", "spectral", "random"]
 
 _LAYOUTS = ("fr", "drl", "kk", "grid_fr", "lgl", "rt", "rt_circular", "fa")
 _Layout = Literal[_LAYOUTS]  # type: ignore
 
-_rank_features_groups_method = Optional[Literal["logreg", "t-test", "wilcoxon", "t-test_overestim_var"]]
+_rank_features_groups_method = Literal["logreg", "t-test", "wilcoxon", "t-test_overestim_var"] | None
 _correction_method = Literal["benjamini-hochberg", "bonferroni"]
 _rank_features_groups_cat_method = Literal[
     "chi-square", "g-test", "freeman-tukey", "mod-log-likelihood", "neyman", "cressie-read"