Nothing is displayed in the image, but I can display it normally using official data in the same environment.

%reset -f
import scanpy as sc
import squidpy as sq
adata = sc.read("data/MCAO_1d.h5ad")
sq.pl.spatial_scatter(
    adata,
    color="Gfap",
    library_id="MCAO_1d",
    img=True,                 # 打开底图
    img_res_key="lowres",     # 指定用 lowres
    spatial_key="spatial",    # 你的坐标键
    # size=1.0,                 # 点大小（可调）
    cmap="magma",
    frameon=False
)

What puzzles me most is that I can display spot very smoothly on Squidpy's official data.

%reset -f
import numpy as np
import pandas as pd

import anndata as ad
import scanpy as sc
import squidpy as sq

display(sc.logging.print_header())
print(f"squidpy=={sq.__version__}")

# load the pre-processed dataset
img = sq.datasets.visium_hne_image(path="data/hne_image")
adata = sq.datasets.visium_hne_adata(path="data/hne_adata")

sq.pl.spatial_scatter(
    adata,
    color="Gfap",
    use_raw=False,
    library_id="V1_Adult_Mouse_Brain",
    img=True,                 # 打开底图
    img_res_key="lowres",     # 指定用 lowres
    spatial_key="spatial",    # 你的坐标键
    size=1.0,                 # 点大小（可调）
    cmap="magma",
    frameon=False
)

Reproduce the report with data/MCAO_1d.h5ad using sq.pl.spatial_scatter, then compare its spatial metadata and image entries with the official visium_hne_image and visium_hne_adata datasets. Check the img_res_key, library_id, spatial_key, and coordinate/image metadata first; done means identifying the data difference that prevents the background image from rendering and confirming the behavior with the supplied examples.