to_polygons(labels) turns a label containing a hole into an invalid MultiPolygon with the hole counted as area #1235

Summary

A 5×5 label with a 1-pixel hole (true area 24) becomes MultiPolygon, is_valid=False, area 25.0. Inner contours should become interiors of the enclosing polygon.

Severity (agent's assessment): medium — invalid geometries break downstream sjoin/overlay/clip; 24 of 3,556 polygons are invalid on the CosMx sandbox dataset

Where: src/spatialdata/_core/operations/vectorize.py::_region_props_to_polygon (every contour from skimage.measure.find_contours becomes an independent Polygon wrapped in a MultiPolygon)

Expected behaviour

A valid Polygon with one interior ring and area 24.

Reproduction

Save as repro.py and run uv run repro.py (the PEP 723 header pins spatialdata to the commit the bug was found on; replace the URL fragment with @main to test the current main branch).

# /// script
# requires-python = ">=3.12"
# dependencies = [
#     "spatialdata @ git+https://github.com/scverse/spatialdata.git@ccf1ea048d054b6624214bf618008a9f9ae223e0",
# ]
# ///
"""to_polygons(labels) turns a label with a hole into an invalid MultiPolygon whose area includes the hole."""
import warnings
import numpy as np
from spatialdata import to_polygons
from spatialdata.models import Labels2DModel

warnings.simplefilter("ignore")
arr = np.zeros((7, 7), dtype=np.uint16)
arr[1:6, 1:6] = 1
arr[3, 3] = 0  # 5x5 square with a one-pixel hole -> true area 24
polys = to_polygons(Labels2DModel.parse(arr))
g = polys.geometry.iloc[0]
print(f"geometry type={g.geom_type}, is_valid={g.is_valid}, area={g.area} (expected: valid Polygon with one interior, area 24.0)")
bug = (not g.is_valid) or abs(g.area - 24.0) > 1e-9
print("VERDICT:", "BUG REPRODUCED" if bug else "NOT REPRODUCED")

Observed output

geometry type=MultiPolygon, is_valid=False, area=25.0 (expected: valid Polygon with one interior, area 24.0)
VERDICT: BUG REPRODUCED

Possible fix direction (unverified)

Build polygons with shapely.polygonize/build_area from the contour rings, or classify contours by containment (Polygon(outer, holes=[...])); add make_valid() as a safety net and a test with a labelled ring.

Environment

uv run repro.py with the PEP 723 metadata in the script (fresh, isolated environment; spatialdata built from main @ ccf1ea0 (2026-08-28); Python 3.13, latest releases of the dependencies at run time: pandas 3.0, anndata 0.13, zarr 3.3, dask 2026.8, numpy 2.5, geopandas 1.1, shapely 2.1). macOS (arm64). Also reproduced in a second environment with pandas 2.3.3 / anndata 0.12.11 / numpy 2.4.4 / zarr 3.2.1.

Automatically generated; discovered by an AI agent (Claude) and not yet reviewed by a human.