refactor: Import fewer private modules (#1069)

* fix import

* introduce compat

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* fix the rest

* add download file func

* precommit

* unused import

* apply type annotations

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

Author: selmanozleyen

src/squidpy/_compat.py

@@ -0,0 +1,37 @@
+from __future__ import annotations
+
+from importlib.metadata import version
+
+from packaging.version import Version
+from scanpy.plotting._tools.scatterplots import _add_categorical_legend as add_categorical_legend
+from scanpy.plotting._tools.scatterplots import _panel_grid as panel_grid
+from scanpy.plotting._utils import add_colors_for_categorical_sample_annotation
+
+__all__ = [
+    # scanpy
+    "set_default_colors_for_categorical_obs",
+    "add_categorical_legend",
+    "panel_grid",
+    "add_colors_for_categorical_sample_annotation",
+    # anndata
+    "ArrayView",
+    "SparseCSCView",
+    "SparseCSRView",
+]
+
+# See https://github.com/scverse/squidpy/issues/1061 for more details
+_SET_DEFAULT_COLORS_FOR_CATEGORICAL_OBS_CHANGED = Version(version("scanpy")) >= Version("0.12.0rc1")
+
+if _SET_DEFAULT_COLORS_FOR_CATEGORICAL_OBS_CHANGED:
+    from scanpy.plotting._utils import _set_default_colors_for_categorical_obs as set_default_colors_for_categorical_obs
+else:
+    from scanpy.plotting._utils import set_default_colors_for_categorical_obs
+
+
+CAN_USE_SPARSE_ARRAY = Version(version("anndata")) >= Version("0.11.0rc1")
+if CAN_USE_SPARSE_ARRAY:
+    from anndata._core.views import ArrayView
+    from anndata._core.views import SparseCSCMatrixView as SparseCSCView
+    from anndata._core.views import SparseCSRMatrixView as SparseCSRView
+else:
+    from anndata._core.views import ArrayView, SparseCSCView, SparseCSRView

src/squidpy/datasets/_10x_datasets.py

@@ -7,10 +7,9 @@
 import spatialdata as sd
 from anndata import AnnData
 from scanpy import settings
-from scanpy._utils import check_presence_download
 
 from squidpy._constants._constants import TenxVersions
-from squidpy.datasets._utils import DEFAULT_CACHE_DIR, PathLike, _get_zipped_dataset
+from squidpy.datasets._utils import DEFAULT_CACHE_DIR, PathLike, _get_zipped_dataset, download_file
 
 __all__ = ["visium"]
 
@@ -110,23 +109,20 @@ def visium(
 
     # download spatial data
     tar_pth = sample_dir / visium_files.spatial_attrs
-    check_presence_download(filename=tar_pth, backup_url=url_prefix + visium_files.spatial_attrs)
+    download_file(filename=tar_pth, backup_url=url_prefix + visium_files.spatial_attrs)
     with tarfile.open(tar_pth) as f:
         for el in f:
             if not (sample_dir / el.name).exists():
                 f.extract(el, sample_dir)
 
     # download counts
-    check_presence_download(
+    download_file(
         filename=sample_dir / "filtered_feature_bc_matrix.h5",
         backup_url=url_prefix + visium_files.feature_matrix,
     )
 
     if include_hires_tiff:  # download image
-        check_presence_download(
-            filename=sample_dir / "image.tif",
-            backup_url=url_prefix + visium_files.tif_image,
-        )
+        download_file(filename=sample_dir / "image.tif", backup_url=url_prefix + visium_files.tif_image)
         return read_visium(
             base_dir / sample_id,
             source_image_path=base_dir / sample_id / "image.tif",

src/squidpy/datasets/_utils.py

@@ -10,11 +10,11 @@
 from typing import Any, TypeAlias
 
 import anndata
+import pooch
 import spatialdata as sd
 from anndata import AnnData
 from scanpy import logging as logg
 from scanpy import read
-from scanpy._utils import check_presence_download
 
 from squidpy.im import ImageContainer
 
@@ -170,7 +170,7 @@ def _create_signature(self) -> Signature:
         )
 
     def _download(self, fpath: PathLike, backup_url: str, **kwargs: Any) -> Any:
-        check_presence_download(Path(fpath), backup_url)
+        download_file(filename=Path(fpath), backup_url=backup_url)
 
         img = ImageContainer()
         img.add_img(fpath, layer="image", library_id=self.library_id, **kwargs)
@@ -204,10 +204,7 @@ def _get_zipped_dataset(folderpath: Path, dataset_name: str, figshare_id: str) -
     if not download_zip.exists():
         logg.info(f"Downloading Visium H&E SpatialData to {download_zip}")
         try:
-            check_presence_download(
-                filename=download_zip,
-                backup_url=f"https://ndownloader.figshare.com/files/{figshare_id}",
-            )
+            download_file(filename=download_zip, backup_url=f"https://ndownloader.figshare.com/files/{figshare_id}")
         except Exception as e:
             raise RuntimeError(f"Failed to download dataset: {e}") from e
 
@@ -223,3 +220,16 @@ def _get_zipped_dataset(folderpath: Path, dataset_name: str, figshare_id: str) -
         raise RuntimeError(f"Expected extracted data at {extracted_path}, but not found")
 
     return sd.read_zarr(extracted_path)
+
+
+def download_file(filename: PathLike, backup_url: str) -> None:
+    """
+    Replacement for scanpy._utils.check_presence_download using Pooch.
+    Saves to the exact local path specified in 'filename'.
+    """
+    pooch.retrieve(
+        url=backup_url,
+        known_hash=None,
+        fname=os.path.basename(filename),
+        path=os.path.dirname(filename) or ".",  # Handles current dir if no folder
+    )

src/squidpy/gr/_utils.py

@@ -4,30 +4,21 @@
 
 from collections.abc import Hashable, Iterable, Sequence
 from contextlib import contextmanager
-from importlib.metadata import version
 from typing import TYPE_CHECKING, Any
 
 import numpy as np
 import pandas as pd
 from anndata import AnnData
 from anndata.utils import make_index_unique
-from packaging.version import Version
 from pandas import CategoricalDtype
 from pandas.api.types import infer_dtype
 from scanpy import logging as logg
 from scipy.sparse import csc_matrix, csr_matrix, issparse, spmatrix
 
+from squidpy._compat import ArrayView, SparseCSCView, SparseCSRView
 from squidpy._docs import d
 from squidpy._utils import NDArrayA, _unique_order_preserving
 
-CAN_USE_SPARSE_ARRAY = Version(version("anndata")) >= Version("0.11.0rc1")
-if CAN_USE_SPARSE_ARRAY:
-    from anndata._core.views import ArrayView
-    from anndata._core.views import SparseCSCMatrixView as SparseCSCView
-    from anndata._core.views import SparseCSRMatrixView as SparseCSRView
-else:
-    from anndata._core.views import ArrayView, SparseCSCView, SparseCSRView
-
 
 def _check_tuple_needles(
     needles: Sequence[tuple[Any, Any]],

src/squidpy/pl/_color_utils.py

@@ -11,8 +11,8 @@
 from cycler import Cycler, cycler
 from matplotlib.colors import ListedColormap, to_hex, to_rgba
 from scanpy import logging as logg
-from scanpy.plotting._utils import add_colors_for_categorical_sample_annotation
 
+from squidpy._compat import add_colors_for_categorical_sample_annotation
 from squidpy._constants._pkg_constants import Key
 
 Palette_t: TypeAlias = str | ListedColormap | None

src/squidpy/pl/_interactive/_utils.py

@@ -7,11 +7,11 @@
 from matplotlib.colors import to_hex, to_rgb
 from numba import njit
 from pandas import CategoricalDtype
-from pandas._libs.lib import infer_dtype
+from pandas.api.types import infer_dtype
 from scanpy import logging as logg
-from scanpy.plotting._utils import add_colors_for_categorical_sample_annotation
 from scipy.spatial import KDTree
 
+from squidpy._compat import add_colors_for_categorical_sample_annotation
 from squidpy._constants._pkg_constants import Key
 from squidpy._utils import NDArrayA
 

src/squidpy/pl/_spatial_utils.py

@@ -30,12 +30,12 @@
 from pandas import CategoricalDtype
 from scanpy import logging as logg
 from scanpy import settings as sc_settings
-from scanpy.plotting._tools.scatterplots import _add_categorical_legend
 from skimage.color import label2rgb
 from skimage.morphology import erosion, square
 from skimage.segmentation import find_boundaries
 from skimage.util import map_array
 
+from squidpy._compat import add_categorical_legend
 from squidpy._constants._constants import ScatterShape
 from squidpy._constants._pkg_constants import Key
 from squidpy._utils import NDArrayA
@@ -665,7 +665,7 @@ def _decorate_axs(
                 palette=palette,
                 alpha=alpha,
             )
-            _add_categorical_legend(
+            add_categorical_legend(
                 ax,
                 color_source_vector,
                 palette=palette,

src/squidpy/pl/_utils.py

@@ -21,7 +21,7 @@
 from mpl_toolkits.axes_grid1 import make_axes_locatable
 from numba import njit, prange
 from pandas import CategoricalDtype
-from pandas._libs.lib import infer_dtype
+from pandas.api.types import infer_dtype
 from pandas.core.dtypes.common import (
     is_bool_dtype,
     is_integer_dtype,

src/squidpy/pl/_var_by_distance.py

@@ -14,10 +14,9 @@
 from cycler import Cycler
 from matplotlib import rcParams
 from matplotlib.axes import Axes
-from scanpy.plotting._tools.scatterplots import _panel_grid
-from scanpy.plotting._utils import _set_default_colors_for_categorical_obs
 from scipy.sparse import issparse
 
+from squidpy._compat import panel_grid, set_default_colors_for_categorical_obs
 from squidpy._docs import d
 from squidpy.pl._utils import save_fig
 
@@ -99,7 +98,7 @@ def var_by_distance(
 
     # if several variables are plotted, make a panel grid
     if isinstance(var, list) and not stack_vars:
-        fig, grid = _panel_grid(
+        fig, grid = panel_grid(
             hspace=0.25,
             wspace=0.75 / rcParams["figure.figsize"][0] + 0.02,
             ncols=4,
@@ -174,7 +173,7 @@ def var_by_distance(
             elif covariate is not None and not stack_vars:
                 # make a categorical color palette if none was specified and there are several regplots to be plotted
                 if isinstance(line_palette, str) or line_palette is None:
-                    _set_default_colors_for_categorical_obs(adata, covariate)
+                    set_default_colors_for_categorical_obs(adata, covariate)
                     line_palette = adata.uns[covariate + "_colors"]
                 covariate_instances = df[covariate].unique()