Maint/fix warnings (#798)

* Remove `applymap`

* Remove `dtype` arg from `AnnData`

* Fix `ChainedAssignmentError`

* Use `plt.colormaps` in tests

* Use `errorbar="sd"`

* Remove `is_categorical_dtype`

* Fix `axis` in `DataFrame.groupby`

* Fix docs in `var_by_distance`

* Fix inplace operation in `var_by_distance`

* Use `.iloc` for `row_order` in `_heatmap`

* Convert library ids to list

* Use `Dataset.sizes` in appropriate places

* Set `observed=True` in `groupby`

* Use `na_action=None` in `Categorical.map`

* Ignore `pkg_resources` warnings

* Remove `get_cmap` usage

* Remove old warning filters

src/squidpy/gr/_ligrec.py

@@ -730,7 +730,7 @@ def extractor(res: Sequence[TempResult]) -> TempResult:
 
         return TempResult(means=means, pvalues=pvalues)
 
-    groups = data.groupby("clusters")
+    groups = data.groupby("clusters", observed=True)
     clustering = np.array(data["clusters"].values, dtype=np.int32)
 
     mean = groups.mean().values.T  # (n_genes, n_clusters)

src/squidpy/gr/_utils.py

@@ -18,7 +18,7 @@
 from anndata._core.views import ArrayView, SparseCSCView, SparseCSRView
 from anndata.utils import make_index_unique
 from pandas import CategoricalDtype
-from pandas.api.types import infer_dtype, is_categorical_dtype
+from pandas.api.types import infer_dtype
 from scanpy import logging as logg
 from scipy.sparse import csc_matrix, csr_matrix, issparse, spmatrix
 

src/squidpy/im/_container.py

@@ -151,7 +151,7 @@ def concat(
         # check that imgs are not already 3d
         imgs = list(imgs)
         for img in imgs:
-            if img.data.dims["z"] > 1:
+            if img.data.sizes["z"] > 1:
                 raise ValueError(
                     f"Currently, can concatenate only images with 1 Z-dimension, found `{img.data.dims['z']}`."
                 )
@@ -605,10 +605,10 @@ def _rescale(arr: xr.DataArray) -> xr.DataArray:
             data.attrs = _update_attrs_scale(attrs, scale)
 
         if mask_circle:
-            if data.dims["y"] != data.dims["x"]:
+            if data.sizes["y"] != data.sizes["x"]:
                 raise ValueError(
                     f"Masking circle is only available for square crops, "
-                    f"found crop of shape `{(data.dims['y'], data.dims['x'])}`."
+                    f"found crop of shape `{(data.sizes['y'], data.sizes['x'])}`."
                 )
             c = data.x.shape[0] // 2
             # manually reassign coordinates
@@ -789,7 +789,7 @@ def generate_spot_crops(
                     raise e
         else:
             try:
-                obs_library_ids = adata.obs[library_id]
+                obs_library_ids = list(adata.obs[library_id])
             except KeyError:
                 logg.debug(
                     f"Unable to find library ids in `adata.obs[{library_id!r}]`. "
@@ -1368,7 +1368,7 @@ def shape(self) -> tuple[int, int]:
         """Image shape ``(y, x)``."""
         if not len(self):
             return 0, 0
-        return self.data.dims["y"], self.data.dims["x"]
+        return self.data.sizes["y"], self.data.sizes["x"]
 
     def copy(self, deep: bool = False) -> ImageContainer:
         """
@@ -1596,8 +1596,7 @@ def _repr_html_(self) -> str:
         for i, layer in enumerate(self.data.keys()):
             s += f"<p style={style!r}><strong>{html.escape(str(layer))}</strong>: "
             s += ", ".join(
-                f"<em>{html.escape(str(dim))}</em> ({shape})"
-                for dim, shape in zip(self.data[layer].dims, self.data[layer].shape)
+                f"<em>{html.escape(str(dim))}</em> ({shape})" for dim, shape in self.data[layer].sizes.items()
             )
             s += "</p>"
             if i == 9 and i < len(self) - 1:  # show only first 10 layers

src/squidpy/im/_feature_mixin.py

@@ -330,9 +330,9 @@ def convert_to_full_image_coordinates(x: NDArrayA, y: NDArrayA) -> NDArrayA:
                 return np.array([[]], dtype=np.float64)
 
             if self.data.attrs.get("mask_circle", False):
-                if self.data.dims["y"] != self.data.dims["x"]:
-                    raise ValueError(f"Crop is not a square: `{self.data.dims}`.")
-                c = self.data.dims["x"] // 2  # center
+                if self.data.sizes["y"] != self.data.sizes["x"]:
+                    raise ValueError(f"Crop is not a square: `{self.data.sizes}`.")
+                c = self.data.sizes["x"] // 2  # center
                 mask = (x - c) ** 2 + (y - c) ** 2 <= c**2
                 y = y[mask]
                 x = x[mask]
@@ -341,7 +341,7 @@ def convert_to_full_image_coordinates(x: NDArrayA, y: NDArrayA) -> NDArrayA:
                 return np.array([[]], dtype=np.float64)  # because of masking, should not happen
 
             coord = self.data.attrs.get(
-                Key.img.coords, CropCoords(x0=0, y0=0, x1=self.data.dims["x"], y1=self.data.dims["y"])
+                Key.img.coords, CropCoords(x0=0, y0=0, x1=self.data.sizes["x"], y1=self.data.sizes["y"])
             )  # fall back to default (i.e no crop) coordinates
             padding = self.data.attrs.get(Key.img.padding, _NULL_PADDING)  # fallback to no padding
             y_slc, x_slc = coord.to_image_coordinates(padding).slice

src/squidpy/pl/_color_utils.py

@@ -53,7 +53,7 @@ def _get_palette(
 
     len_cat = len(categories)
     if isinstance(palette, str):
-        cmap = plt.get_cmap(palette)
+        cmap = plt.colormaps[palette]
         palette = [to_hex(x, keep_alpha=True) for x in cmap(np.linspace(0, 1, len_cat), alpha=alpha)]
     elif isinstance(palette, ListedColormap):
         palette = [to_hex(x, keep_alpha=True) for x in palette(np.linspace(0, 1, len_cat), alpha=alpha)]

src/squidpy/pl/_graph.py

@@ -159,7 +159,7 @@ def interaction_matrix(
     _assert_categorical_obs(adata, key=cluster_key)
     array = _get_data(adata, cluster_key=cluster_key, func_name="interaction_matrix")
 
-    ad = AnnData(X=array, obs={cluster_key: pd.Categorical(adata.obs[cluster_key].cat.categories)}, dtype=array.dtype)
+    ad = AnnData(X=array, obs={cluster_key: pd.Categorical(adata.obs[cluster_key].cat.categories)})
     _maybe_set_colors(source=adata, target=ad, key=cluster_key, palette=palette)
     if title is None:
         title = "Interaction matrix"
@@ -224,7 +224,7 @@ def nhood_enrichment(
     _assert_categorical_obs(adata, key=cluster_key)
     array = _get_data(adata, cluster_key=cluster_key, func_name="nhood_enrichment")[mode]
 
-    ad = AnnData(X=array, obs={cluster_key: pd.Categorical(adata.obs[cluster_key].cat.categories)}, dtype=array.dtype)
+    ad = AnnData(X=array, obs={cluster_key: pd.Categorical(adata.obs[cluster_key].cat.categories)})
     _maybe_set_colors(source=adata, target=ad, key=cluster_key, palette=palette)
     if title is None:
         title = "Neighborhood enrichment"
@@ -315,7 +315,7 @@ def ripley(
         **kwargs,
     )
     if plot_sims:
-        sns.lineplot(y="stats", x="bins", ci="sd", alpha=0.01, color="gray", data=res["sims_stat"], ax=ax)
+        sns.lineplot(y="stats", x="bins", errorbar="sd", alpha=0.01, color="gray", data=res["sims_stat"], ax=ax)
     ax.legend(**legend_kwargs)
     ax.set_ylabel("value")
     ax.set_title(f"Ripley's {mode.s}")

src/squidpy/pl/_interactive/_controller.py

@@ -10,7 +10,6 @@
 from napari import Viewer
 from napari.layers import Points, Shapes
 from pandas import CategoricalDtype
-from pandas.core.dtypes.common import is_categorical_dtype
 from PyQt5.QtWidgets import QGridLayout, QLabel, QWidget
 from scanpy import logging as logg
 

src/squidpy/pl/_interactive/_utils.py

@@ -8,7 +8,6 @@
 from numba import njit
 from pandas import CategoricalDtype
 from pandas._libs.lib import infer_dtype
-from pandas.core.dtypes.common import is_categorical_dtype
 from scanpy import logging as logg
 from scanpy.plotting._utils import add_colors_for_categorical_sample_annotation
 from scipy.spatial import KDTree

src/squidpy/pl/_ligrec.py

@@ -107,7 +107,7 @@ def _plot_size_legend(self, size_legend_ax: Axes) -> None:
             ax.set_position([l + w, b, w, h])
 
     def _plot_colorbar(self, color_legend_ax: Axes, normalize: bool) -> None:
-        cmap = plt.get_cmap(self.cmap)
+        cmap = plt.colormaps[self.cmap]
 
         ColorbarBase(
             color_legend_ax,
@@ -303,7 +303,7 @@ def get_dendrogram(adata: AnnData, linkage: str = "complete") -> Mapping[str, An
         pvals = pvals.T
         means = means.T
 
-    for cls, size in (pvals.groupby(level=0, axis=1)).size().to_dict().items():
+    for cls, size in (pvals.T.groupby(level=0)).size().to_dict().items():
         label_ranges[cls] = (start, start + size - 1)
         start += size
     label_ranges = {k: label_ranges[k] for k in sorted(label_ranges.keys())}
@@ -322,7 +322,7 @@ def get_dendrogram(adata: AnnData, linkage: str = "complete") -> Mapping[str, An
     var = pd.DataFrame(pvals.columns)
     var = var.set_index(var.columns[0])
 
-    adata = AnnData(pvals.values, obs={"groups": pd.Categorical(pvals.index)}, var=var, dtype=pvals.values.dtype)
+    adata = AnnData(pvals.values, obs={"groups": pd.Categorical(pvals.index)}, var=var)
     adata.obs_names = pvals.index
     minn = np.nanmin(adata.X)
     delta = np.nanmax(adata.X) - minn

src/squidpy/pl/_spatial_utils.py

@@ -26,7 +26,6 @@
 from matplotlib import colors, patheffects, rcParams
 from matplotlib import pyplot as plt
 from matplotlib.axes import Axes
-from matplotlib.cm import get_cmap
 from matplotlib.collections import Collection, PatchCollection
 from matplotlib.colors import (
     ColorConverter,
@@ -40,7 +39,6 @@
 from matplotlib.patches import Circle, Polygon, Rectangle
 from matplotlib_scalebar.scalebar import ScaleBar
 from pandas import CategoricalDtype
-from pandas.core.dtypes.common import is_categorical_dtype
 from scanpy import logging as logg
 from scanpy._settings import settings as sc_settings
 from scanpy.plotting._tools.scatterplots import _add_categorical_legend
@@ -385,7 +383,7 @@ def subset_by_key(
     ) -> tuple[AnnData, NDArrayA]:
         if key is None or values is None:
             return adata, coords
-        if key not in adata.obs or not is_categorical_dtype(adata.obs[key]):
+        if key not in adata.obs or not isinstance(adata.obs[key].dtype, CategoricalDtype):
             return adata, coords
         try:
             mask = adata.obs[key].isin(values).values
@@ -468,7 +466,7 @@ def _set_color_source_vec(
     else:
         color_source_vector = adata.obs_vector(value_to_plot, layer=layer)
 
-    if not is_categorical_dtype(color_source_vector):
+    if not isinstance(color_source_vector.dtype, CategoricalDtype):
         return None, color_source_vector, False
 
     color_source_vector = pd.Categorical(color_source_vector)  # convert, e.g., `pd.Series`
@@ -480,7 +478,7 @@ def _set_color_source_vec(
     if color_map is None:
         raise ValueError("Unable to create color palette.")
     # do not rename categories, as colors need not be unique
-    color_vector = color_source_vector.map(color_map)
+    color_vector = color_source_vector.map(color_map, na_action=None)
     if color_vector.isna().any():
         color_vector = color_vector.add_categories([to_hex(na_color)])
         color_vector = color_vector.fillna(to_hex(na_color))
@@ -613,7 +611,7 @@ def _decorate_axs(
     adata: AnnData,
     coords: NDArrayA,
     value_to_plot: str,
-    color_source_vector: pd.Series[CategoricalDtype],
+    color_source_vector: pd.Series[CategoricalDtype] | None,
     img: NDArrayA | None = None,
     img_cmap: str | None = None,
     img_alpha: float | None = None,
@@ -646,7 +644,7 @@ def _decorate_axs(
             path_effect = []
 
         # Adding legends
-        if is_categorical_dtype(color_source_vector):
+        if color_source_vector is not None and isinstance(color_source_vector.dtype, CategoricalDtype):
             clusters = color_source_vector.categories
             palette = _get_palette(adata, cluster_key=value_to_plot, categories=clusters, palette=palette, alpha=alpha)
             _add_categorical_legend(
@@ -691,7 +689,7 @@ def _map_color_seg(
 ) -> NDArrayA:
     cell_id = np.array(cell_id)
 
-    if is_categorical_dtype(color_vector):
+    if isinstance(color_vector.dtype, CategoricalDtype):
         if isinstance(na_color, tuple) and len(na_color) == 4 and np.any(color_source_vector.isna()):
             cell_id[color_source_vector.isna()] = 0
         val_im: NDArrayA = map_array(seg, cell_id, color_vector.codes + 1)  # type: ignore[union-attr]
@@ -822,7 +820,10 @@ def _prepare_params_plot(
             fig, ax = plt.subplots(figsize=figsize, dpi=dpi, constrained_layout=True)
 
     # set cmap and norm
-    cmap = copy(get_cmap(cmap))
+    if cmap is None:
+        cmap = plt.rcParams["image.cmap"]
+    if isinstance(cmap, str):
+        cmap = plt.colormaps[cmap]
     cmap.set_bad("lightgray" if na_color is None else na_color)
 
     if isinstance(norm, Normalize):

src/squidpy/pl/_utils.py

@@ -1,7 +1,6 @@
 from __future__ import annotations
 
 import os
-from copy import copy
 from functools import wraps
 from inspect import signature
 from pathlib import Path
@@ -34,7 +33,6 @@
 from pandas._libs.lib import infer_dtype
 from pandas.core.dtypes.common import (
     is_bool_dtype,
-    is_categorical_dtype,
     is_integer_dtype,
     is_numeric_dtype,
     is_object_dtype,
@@ -482,7 +480,7 @@ def _annotate_heatmap(
 ) -> None:
     # modified from matplotlib's site
     if isinstance(cmap, str):
-        cmap = plt.get_cmap(cmap)
+        cmap = plt.colormaps[cmap]
 
     data = im.get_array()
     kw = {"ha": "center", "va": "center"}
@@ -554,7 +552,7 @@ def _heatmap(
         row_order = col_order = np.arange(len(adata.uns[Key.uns.colors(key)])).tolist()
 
     row_order = row_order[::-1]
-    row_labels = adata.obs[key][row_order]
+    row_labels = adata.obs[key].iloc[row_order]
     data = adata[row_order, col_order].X
 
     row_cmap, col_cmap, row_norm, col_norm, n_cls = _get_cmap_norm(adata, key, order=(row_order, col_order))
@@ -563,7 +561,8 @@ def _heatmap(
     col_sm = mpl.cm.ScalarMappable(cmap=col_cmap, norm=col_norm)
 
     norm = mpl.colors.Normalize(vmin=kwargs.pop("vmin", np.nanmin(data)), vmax=kwargs.pop("vmax", np.nanmax(data)))
-    cont_cmap = copy(plt.get_cmap(cont_cmap))
+    if isinstance(cont_cmap, str):
+        cont_cmap = plt.colormaps[cont_cmap]
     cont_cmap.set_bad(color="grey")
 
     im = ax.imshow(data[::-1], cmap=cont_cmap, norm=norm)

src/squidpy/pl/_var_by_distance.py

@@ -67,24 +67,31 @@ def var_by_distance(
     line_palette
         Categorical color palette used in case a covariate is specified.
     scatter_palette
-        Color palette for the scatter plot underlying the `sns.regplot`
-    %(plotting_save)s
+        Color palette for the scatter plot underlying the :func:`seaborn.regplot`.
+    dpi
+        Dots per inch.
+    figsize
+        Size of the figure in inches.
+    save
+        Whether to save the plot.
     title
         Panel titles.
     axis_label
         Panel axis labels.
+    return_ax
+        Whether to return :class:`matplotlib.axes.Axes` object(s).
     regplot_kwargs
-        Kwargs for `sns.regplot`
+        Kwargs for :func:`seaborn.regplot`.
     scatterplot_kwargs
-        Kwargs for `sns.scatter`
+        Kwargs for :func:`matplotlib.pyplot.scatter`.
 
     Returns
     -------
     %(plotting_returns)s
     """
     dpi = rcParams["figure.dpi"] if dpi is None else dpi
     regplot_kwargs = dict(regplot_kwargs)
-    scatterplot_kwargs = dict(regplot_kwargs)
+    scatterplot_kwargs = dict(scatterplot_kwargs)
 
     df = adata.obsm[design_matrix_key]  # get design matrix
     df[var] = np.array(adata[:, var].X.A) if issparse(adata[:, var].X) else np.array(adata[:, var].X)  # add var column

src/squidpy/read/_read.py

@@ -232,7 +232,6 @@ def nanostring(
 
     adata = AnnData(
         csr_matrix(counts.loc[common_index, :].values),
-        dtype=counts.values.dtype,
         obs=obs.loc[common_index, :],
         uns={Key.uns.spatial: {}},
     )