Microenvironment modeling analysis (patsy.PatsyError: categorical data cannot be >1-dimensional)

[heziqing]

Hello everyone,

I re-run the analysis https://spateo-release.readthedocs.io/en/latest/tutorials/notebooks/3_cci/2_microenvironment_modeling_analysis.html#, but I got an error when running the following code

niche_interp = st.tl.Niche_Model(
    adata,
    group_key='Annotation',
    genes=targets,
    smooth=False,
    normalize=False,
    log_transform=False,
    weights_mode="knn",
    distr="poisson",
    data_id="axolotl",
)

File "/share/home/heziqing/miniconda3/envs/Spateo/lib/python3.9/site-packages/patsy/categorical.py", line 157, in _categorical_shape_fix raise PatsyError("categorical data cannot be >1-dimensional") patsy.PatsyError: categorical data cannot be >1-dimensional

I also check the scripts, and I found the main steps are the following codes:

group_name = adata.obs['Annotation']
db = pd.DataFrame({"group": group_name})
categories = np.array(adata.obs['Annotation'].unique().tolist())
db["group"] = pd.Categorical(db["group"], categories=categories)
X = pd.get_dummies(data=db, drop_first=False)
X = X.reindex(sorted(X.columns), axis=1)
adata.obsm["adj"] = pd.read_csv("/neighbors/axolotl_neighbors.csv",index_col=0).values
adj = adata.obsm["adj"]
dmat_neighbors = (adj > 0).astype("int").dot(X.values)
data = {"categories": X, "dmat_neighbours": dmat_neighbors}
connections = np.asarray(dmatrix("categories:dmat_neighbours-1", data))

Could anyone help me how to fix this error? Thanks in advance!

[github-actions[bot]]

This issue is stale because it has been open 90 days with no activity. Remove stale label or comment or this will be closed in 14 days

[Sichao25]

Hi, thanks for reporting this issue. The Niche_Model and the related tutorial have been deprecated, so it could be difficult to debug. We’ve integrated the related functionality into a new class called MuSIC. Feel free to try it out for cell-cell interaction analysis. Let me know if you meet the same issue.