Catch all warnings (scverse#3724)

Author: flying-sheep

.github/workflows/ci.yml

@@ -12,6 +12,9 @@ env:
   FORCE_COLOR: "1"
   MPLBACKEND: agg
   UV_CONSTRAINT: ci/constraints.txt
+  # It’s impossible to ignore SyntaxWarnings for a single module,
+  # so because leidenalg 0.10.0 has them, we pre-compile things: https://github.com/vtraag/leidenalg/issues/173
+  UV_COMPILE_BYTECODE: "1"
 
 defaults:
   run:

docs/installation.md

@@ -24,7 +24,7 @@ After installing installing e.g. [Miniconda][], run:
 $ conda install -c conda-forge scanpy python-igraph leidenalg
 ```
 
-Pull Scanpy [from PyPI][] (consider using `pip3` to access Python 3):
+Pull Scanpy [from PyPI][]:
 
 ```console
 $ pip install scanpy

docs/release-notes/3724.bugfix.md

@@ -0,0 +1 @@
+Raise fewer redundant warnings, mainly in {mod}`scanpy.pl` and {mod}`scanpy.datasets` functions {smaller}`P Angerer`

pyproject.toml

@@ -170,26 +170,20 @@ markers = [
     "anndata_dask_support: tests that require dask support in anndata",
 ]
 filterwarnings = [
-    # legacy-api-wrap: internal use of positional API
-    "error:The specified parameters:FutureWarning",
-    # When calling `.show()` in tests, this is raised
-    "ignore:FigureCanvasAgg is non-interactive:UserWarning",
-
-    # We explicitly handle the below errors in tests
-    "error:`anndata.read` is deprecated:FutureWarning",
-    "error:Observation names are not unique:UserWarning",
-    "error:The dtype argument is deprecated and will be removed:FutureWarning",
-    "error:The behavior of DataFrame\\.sum with axis=None is deprecated:FutureWarning",
-    "error:The default of observed=False is deprecated:FutureWarning",
-    "error:Series\\.__getitem__ treating keys as positions is deprecated:FutureWarning",
-    "error:The default value of 'ignore' for the `na_action` parameter in pandas\\.Categorical\\.map:FutureWarning",
-    "error:The provided callable.* is currently using:FutureWarning",
-    "error:The behavior of DataFrame concatenation with empty or all-NA entries is deprecated:FutureWarning",
-    "error:A value is trying to be set on a copy of a slice from a DataFrame",
-    "error:No data for colormapping provided via 'c'",
-    "error:\\n*The `scale` parameter has been renamed and will be removed",
-    "error:\\n*Passing `palette` without assigning `hue` is deprecated",
-    "error:\\n*Setting a gradient palette using color= is deprecated",
+    'error',
+    # Umap warns when tensorflow isn’t installed.
+    'ignore::ImportWarning:umap',
+    # seaborn≤0.13.2 causes some matplotlib warnings
+    'ignore::PendingDeprecationWarning:seaborn',
+    # matplotlib<3.10.4 causes a Pillow warning
+    'ignore:.*Pillow:DeprecationWarning',
+    # networkx 2.x warns about scipy.sparse changes
+    'ignore:\n*.*scipy\.sparse array:DeprecationWarning:networkx',
+    # h5py <3.9 uses deprecated np.product
+    'ignore:.*product.*deprecated.*NumPy:DeprecationWarning',
+    # we want to see and eventually fix these
+    'default::numba.core.errors.NumbaPerformanceWarning',
+    'default:.*TSNE.*random.*to.*pca:FutureWarning',      # we should set init=obsm["X_pca"] or so
 ]
 
 [tool.coverage.run]

src/scanpy/_utils/__init__.py

@@ -84,7 +84,7 @@ def ensure_igraph() -> None:
     msg = (
         "Please install the igraph package: "
         "`conda install -c conda-forge python-igraph` or "
-        "`pip3 install igraph`."
+        "`pip install igraph`."
     )
     raise ImportError(msg)
 
@@ -127,6 +127,7 @@ def renamed_arg(old_name, new_name, *, pos_0: bool = False):
     def decorator(func):
         @wraps(func)
         def wrapper(*args, **kwargs):
+            __tracebackhide__ = True
             if old_name in kwargs:
                 f_name = func.__name__
                 pos_str = (
@@ -832,8 +833,8 @@ def warn_with_traceback(  # noqa: PLR0917
     settings.write(warnings.formatwarning(message, category, filename, lineno, line))
 
 
-def warn_once(msg: str, category: type[Warning], stacklevel: int = 1):
-    warnings.warn(msg, category, stacklevel=stacklevel)
+def warn_once(msg: str, category: type[Warning], stacklevel: int = 0) -> None:
+    warnings.warn(msg, category, stacklevel=stacklevel + 1)
     # You'd think `'once'` works, but it doesn't at the repl and in notebooks
     warnings.filterwarnings("ignore", category=category, message=re.escape(msg))
 

src/scanpy/datasets/_datasets.py

@@ -6,14 +6,14 @@
 
 import numpy as np
 import pandas as pd
-from anndata import AnnData
+from anndata import AnnData, OldFormatWarning
 
 from .. import _utils
 from .._compat import deprecated, old_positionals
 from .._settings import settings
 from .._utils._doctests import doctest_internet, doctest_needs
 from ..readwrite import read, read_visium
-from ._utils import check_datasetdir_exists, filter_oldformatwarning
+from ._utils import check_datasetdir_exists
 
 if TYPE_CHECKING:
     from typing import Literal
@@ -123,6 +123,8 @@ def burczynski06() -> AnnData:
     --------
     >>> import scanpy as sc
     >>> sc.datasets.burczynski06()
+    UserWarning: Variable names are not unique. To make them unique, call `.var_names_make_unique`.
+        utils.warn_names_duplicates("var")
     AnnData object with n_obs × n_vars = 127 × 22283
         obs: 'groups'
 
@@ -196,6 +198,8 @@ def moignard15() -> AnnData:
     --------
     >>> import scanpy as sc
     >>> sc.datasets.moignard15()
+    UserWarning: Unknown extension is not supported and will be removed
+        warn(msg)
     AnnData object with n_obs × n_vars = 3934 × 42
         obs: 'exp_groups'
         uns: 'iroot', 'exp_groups_colors'
@@ -318,7 +322,6 @@ def toggleswitch() -> AnnData:
     return adata
 
 
-@filter_oldformatwarning
 def pbmc68k_reduced() -> AnnData:
     r"""Subsampled and processed 68k PBMCs.
 
@@ -354,12 +357,12 @@ def pbmc68k_reduced() -> AnnData:
     """
     filename = HERE / "10x_pbmc68k_reduced.h5ad"
     with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=OldFormatWarning, module="anndata")
         warnings.filterwarnings("ignore", category=FutureWarning, module="anndata")
         return read(filename)
 
 
 @doctest_internet
-@filter_oldformatwarning
 @check_datasetdir_exists
 def pbmc3k() -> AnnData:
     r"""3k PBMCs from 10x Genomics.
@@ -406,12 +409,13 @@ def pbmc3k() -> AnnData:
 
     """
     url = "https://falexwolf.de/data/pbmc3k_raw.h5ad"
-    adata = read(settings.datasetdir / "pbmc3k_raw.h5ad", backup_url=url)
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=OldFormatWarning, module="anndata")
+        adata = read(settings.datasetdir / "pbmc3k_raw.h5ad", backup_url=url)
     return adata
 
 
 @doctest_internet
-@filter_oldformatwarning
 @check_datasetdir_exists
 def pbmc3k_processed() -> AnnData:
     """Processed 3k PBMCs from 10x Genomics.
@@ -447,6 +451,7 @@ def pbmc3k_processed() -> AnnData:
     url = "https://raw.githubusercontent.com/chanzuckerberg/cellxgene/main/example-dataset/pbmc3k.h5ad"
 
     with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=OldFormatWarning, module="anndata")
         warnings.filterwarnings("ignore", category=FutureWarning, module="anndata")
         return read(settings.datasetdir / "pbmc3k_processed.h5ad", backup_url=url)
 
@@ -540,6 +545,10 @@ def visium_sge(
     --------
     >>> import scanpy as sc
     >>> sc.datasets.visium_sge(sample_id="V1_Breast_Cancer_Block_A_Section_1")
+    FutureWarning: Use `squidpy.datasets.visium` instead.
+        sc.datasets.visium_sge(sample_id="V1_Breast_Cancer_Block_A_Section_1")
+    UserWarning: Variable names are not unique. To make them unique, call `.var_names_make_unique`.
+        utils.warn_names_duplicates("var")
     AnnData object with n_obs × n_vars = 3798 × 36601
         obs: 'in_tissue', 'array_row', 'array_col'
         var: 'gene_ids', 'feature_types', 'genome'
@@ -552,4 +561,6 @@ def visium_sge(
         sample_id, spaceranger_version, download_image=include_hires_tiff
     )
     source_image_path = sample_dir / "image.tif" if include_hires_tiff else None
-    return read_visium(sample_dir, source_image_path=source_image_path)
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", r".*squidpy\.read", FutureWarning)
+        return read_visium(sample_dir, source_image_path=source_image_path)

src/scanpy/datasets/_utils.py

@@ -1,12 +1,8 @@
 from __future__ import annotations
 
-import warnings
 from functools import wraps
 from typing import TYPE_CHECKING
 
-import anndata as ad
-from packaging.version import Version
-
 from .._settings import settings
 
 if TYPE_CHECKING:
@@ -24,18 +20,3 @@ def wrapper(*args: P.args, **kwargs: P.kwargs) -> R:
         return f(*args, **kwargs)
 
     return wrapper
-
-
-def filter_oldformatwarning(f: Callable[P, R]) -> Callable[P, R]:
-    """Filter anndata.OldFormatWarning from being thrown by the wrapped function."""
-
-    @wraps(f)
-    def wrapper(*args: P.args, **kwargs: P.kwargs) -> R:
-        with warnings.catch_warnings():
-            if Version(ad.__version__).release >= (0, 8):
-                warnings.filterwarnings(
-                    "ignore", category=ad.OldFormatWarning, module="anndata"
-                )
-            return f(*args, **kwargs)
-
-    return wrapper

src/scanpy/logging.py

@@ -78,11 +78,12 @@ def _set_log_file(settings: SettingsMeta) -> None:
     file = settings.logfile
     name = settings.logpath
     root = settings._root_logger
+    for handler in list(root.handlers):
+        root.removeHandler(handler)
+        handler.close()
     h = logging.StreamHandler(file) if name is None else logging.FileHandler(name)
     h.setFormatter(_LogFormatter())
     h.setLevel(root.level)
-    for handler in list(root.handlers):
-        root.removeHandler(handler)
     root.addHandler(h)
 
 

src/scanpy/plotting/_tools/__init__.py

@@ -1325,7 +1325,7 @@ def rank_genes_groups_violin(  # noqa: PLR0913
         _ax.set_title(f"{group_name} vs. {reference}")
         _ax.legend_.remove()
         _ax.set_ylabel("expression")
-        _ax.set_xticklabels(new_gene_names, rotation="vertical")
+        _ax.set_xticks(range(len(new_gene_names)), new_gene_names, rotation="vertical")
         writekey = (
             f"rank_genes_groups_{adata.uns[key]['params']['groupby']}_{group_name}"
         )

src/scanpy/plotting/_tools/paga.py

@@ -1289,7 +1289,7 @@ def moving_average(a):
             ),
         )
         if show_yticks:
-            groups_axis.set_yticklabels(["", xlabel, ""], fontsize=ytick_fontsize)
+            groups_axis.set_yticks(range(3), [xlabel, "", ""], fontsize=ytick_fontsize)
         else:
             groups_axis.set_yticks([])
         groups_axis.set_frame_on(False)
@@ -1331,7 +1331,7 @@ def moving_average(a):
                 cmap=color_map_anno,
             )
             if show_yticks:
-                anno_axis.set_yticklabels(["", anno, ""], fontsize=ytick_fontsize)
+                anno_axis.set_yticks(range(3), [anno, "", ""], fontsize=ytick_fontsize)
                 anno_axis.tick_params(axis="both", which="both", length=0)
             else:
                 anno_axis.set_yticks([])