Merging multiple samples into a single SpatialData object + single merged table

Hello!

I am trying to merge several samples into a single spatialdata object, and currently this is what I'm doing:

adatas = {
    "square_002um": [],
    "square_008um": [],
    "square_016um": [],
}
images = {}

for sample in sample_folders:
    sdata = spatialdata_io.visium_hd(
        f"{outs}/{sample}/outs",
        dataset_id=sample,
        fullres_image_file=id_to_image[sample],
    )
    # Collect tables
    for size, table in sdata.tables.items():
        table.var_names_make_unique()
        table.obs[f"sample_id"] = sample
        #Add mapping from dictionary above:
        table.obs["sfh_id"] = sample_mapping[sample]
        adatas.setdefault(size, []).append(table)

    # Collect images
    for name, img in sdata.images.items():
        images[f"{sample}_{name}"] = img

# Build one SpatialData object with merged table + all images
merged_bin_sizes = {}
for size, adata in adatas.items():
    temp_merged = ad.concat(adata, join="outer", label="sample_id", fill_value=0)
    merged_bin_sizes[size] = temp_merged
sdata_merged = SpatialData(
    images=images,
    tables=merged_bin_sizes
)

Which ultimately gives me this:

To me it seems like the way to go, but I was wondering if this is recommended or if there is another specific way of doing things to easy the downstream analysis? Perhaps even split things a bit further with the tables between 02,08 and 16 mu? Thanks

Review the spatialdata_io.visium_hd entry point and the SpatialData construction shown in the issue, focusing on how tables at different bin sizes and images are represented. Determine whether the proposed merged structure is supported for downstream analysis and document the recommended organization, including what “done” means for the 02, 08, and 16 μm tables.