Backport PR #3864 on branch 1.11.x (fix: various DPT plot fixes) (#3869)

Author: flying-sheep

docs/release-notes/3864.fix.md

@@ -0,0 +1 @@
+Fix {func}`~scanpy.pl.dpt_groups_pseudotime` {smaller}`P Angerer`

src/scanpy/plotting/_tools/__init__.py

@@ -279,7 +279,8 @@ def dpt_groups_pseudotime(
     show: bool | None = None,
     save: bool | str | None = None,
     marker: str | Sequence[str] = ".",
-):
+    return_fig: bool = False,
+) -> Figure | None:
     """Plot groups and pseudotime.
 
     Parameters
@@ -292,9 +293,9 @@ def dpt_groups_pseudotime(
         Marker style. See :mod:`~matplotlib.markers` for details.
 
     """
-    _, (ax_grp, ax_ord) = plt.subplots(2, 1)
+    fig, (ax_grp, ax_ord) = plt.subplots(2, 1)
     timeseries_subplot(
-        adata.obs["dpt_groups"].cat.codes,
+        adata.obs["dpt_groups"].cat.codes.to_numpy(),
         time=adata.obs["dpt_order"].values,
         color=np.asarray(adata.obs["dpt_groups"]),
         highlights_x=adata.uns["dpt_changepoints"],
@@ -321,6 +322,8 @@ def dpt_groups_pseudotime(
         marker=marker,
     )
     savefig_or_show("dpt_groups_pseudotime", save=save, show=show)
+    if return_fig:
+        return fig
 
 
 @old_positionals(

src/scanpy/plotting/_utils.py

@@ -2,6 +2,7 @@
 
 import warnings
 from collections.abc import Callable, Mapping, Sequence
+from itertools import cycle, islice
 from typing import TYPE_CHECKING, Literal, TypedDict, overload
 
 import numpy as np
@@ -153,41 +154,42 @@ def timeseries_subplot(  # noqa: PLR0912, PLR0913
         X with n columns, color is of length n.
 
     """
-    if color is not None:
-        use_color_map = isinstance(color[0], float | np.floating)
+    use_color_map = color is not None and isinstance(color[0], float | np.floating)
     palette = default_palette(palette)
     x_range = np.arange(X.shape[0]) if time is None else time
     if X.ndim == 1:
         X = X[:, None]  # noqa: N806
     if X.shape[1] > 1:
-        colors = palette[: X.shape[1]].by_key()["color"]
+        colors = islice(cycle(palette.by_key()["color"]), X.shape[1])
         subsets = [(x_range, X[:, i]) for i in range(X.shape[1])]
     elif use_color_map:
         colors = [color]
         subsets = [(x_range, X[:, 0])]
     else:
         levels, _ = np.unique(color, return_inverse=True)
-        colors = np.array(palette[: len(levels)].by_key()["color"])
+        colors = islice(cycle(palette.by_key()["color"]), len(levels))
         subsets = [(x_range[color == level], X[color == level, :]) for level in levels]
 
     if isinstance(marker, str):
         marker = [marker]
     if len(marker) != len(subsets) and len(marker) == 1:
-        marker = [marker[0] for _ in range(len(subsets))]
+        marker = [marker[0]] * len(subsets)
+    if not (has_var_names := (len(var_names) > 0)):
+        var_names = [""] * len(subsets)
 
     if ax is None:
         ax = plt.subplot()
-    for i, (x, y) in enumerate(subsets):
+    for (x, y), m, c, var_name in zip(subsets, marker, colors, var_names, strict=True):
         ax.scatter(
             x,
             y,
-            marker=marker[i],
+            marker=m,
             edgecolor="face",
             s=rcParams["lines.markersize"],
-            c=colors[i],
-            label=var_names[i] if len(var_names) > 0 else "",
-            cmap=color_map,
+            c=c,
+            label=var_name,
             rasterized=settings._vector_friendly,
+            **(dict(cmap=color_map) if use_color_map else {}),
         )
     ylim = ax.get_ylim()
     for h in highlights_x:
@@ -199,7 +201,7 @@ def timeseries_subplot(  # noqa: PLR0912, PLR0913
     ax.set_ylabel(ylabel)
     if yticks is not None:
         ax.set_yticks(yticks)
-    if len(var_names) > 0 and legend:
+    if has_var_names and legend:
         ax.legend(frameon=False)
 
 

src/scanpy/tools/_dpt.py

@@ -450,9 +450,8 @@ def select_segment(self, segs, segs_tips, segs_undecided) -> tuple[int, int]:  #
             # if we did not normalize, there would be a danger of simply
             # assigning the highest score to the longest segment
             score = dseg[tips3[2]] / d_seg[tips3[0], tips3[1]]
-            score = (
-                len(seg) if self.choose_largest_segment else score
-            )  # simply the number of points
+            # simply the number of points
+            score = len(seg) if self.choose_largest_segment else score
             logg.debug(
                 f"    group {iseg} score {score} n_points {len(seg)}"
                 f"{' (too small)' if len(seg) < self.min_group_size else ''}"

tests/_images/dpt_groups_pseudotime/expected.png

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import warnings
 from functools import partial
 from itertools import chain, combinations, repeat
 from pathlib import Path
@@ -1320,12 +1321,39 @@ def test_scatter_embedding_add_outline_vmin_vmax_norm_ref(tmp_path, check_same_i
         )
 
 
-def test_timeseries():
+@pytest.fixture(scope="session")
+def pbmc_68k_dpt_session() -> AnnData:
     adata = pbmc68k_reduced()
     sc.pp.neighbors(adata, n_neighbors=5, method="gauss", knn=False)
-    sc.tl.diffmap(adata)
-    sc.tl.dpt(adata, n_branchings=1, n_dcs=10)
-    sc.pl.dpt_timeseries(adata, as_heatmap=True)
+    sc.tl.leiden(adata, resolution=0.5, key_added="leiden_0_5", flavor="leidenalg")
+    adata.uns["iroot"] = np.flatnonzero(adata.obs["leiden_0_5"] == "0")[0]
+    sc.tl.diffmap(adata, n_comps=10)
+    with warnings.catch_warnings():
+        warnings.filterwarnings(
+            "ignore", ".*invalid value encountered in scalar divide"
+        )
+        sc.tl.dpt(adata, n_branchings=3)
+    return adata
+
+
+@needs.leidenalg
+@needs.igraph
+@pytest.mark.parametrize(
+    "func",
+    [sc.pl.dpt_groups_pseudotime, sc.pl.dpt_timeseries],
+)
+def test_dpt_plots(
+    image_comparer, pbmc_68k_dpt_session: AnnData, func: Callable
+) -> None:
+    save_and_compare_images = partial(image_comparer, ROOT, tol=15)
+
+    adata = pbmc_68k_dpt_session.copy()
+    func(
+        adata,
+        show=False,
+        **(dict(as_heatmap=True) if func is sc.pl.dpt_timeseries else {}),
+    )
+    save_and_compare_images(func.__name__)
 
 
 def test_scatter_raw(tmp_path):