scanpy.read_loom

Contents

scanpy.read_loom#

scanpy.read_loom(filename, *, sparse=True, cleanup=False, X_name='spliced', obs_names='CellID', obsm_names=None, var_names='Gene', varm_names=None, dtype='float32', obsm_mapping=mappingproxy({}), varm_mapping=mappingproxy({}), **kwargs)[source]#

Read .loom-formatted hdf5 file.

This reads the whole file into memory.

Beware that you have to explicitly state when you want to read the file as sparse data.

Parameters:
filename PathLike[str] | str

The filename.

sparse bool (default: True)

Whether to read the data matrix as sparse.

cleanup bool (default: False)

Whether to collapse all obs/var fields that only store one unique value into .uns['loom-.'].

X_name str (default: 'spliced')

Loompy key with which the data matrix X is initialized.

obs_names str (default: 'CellID')

Loompy key where the observation/cell names are stored.

obsm_mapping Mapping[str, Iterable[str]] (default: mappingproxy({}))

Loompy keys which will be constructed into observation matrices

var_names str (default: 'Gene')

Loompy key where the variable/gene names are stored.

varm_mapping Mapping[str, Iterable[str]] (default: mappingproxy({}))

Loompy keys which will be constructed into variable matrices

**kwargs

Arguments to loompy.connect

Return type:

AnnData

Example

pbmc = anndata.io.read_loom(
    "pbmc.loom",
    sparse=True,
    X_name="lognorm",
    obs_names="cell_names",
    var_names="gene_names",
    obsm_mapping={
        "X_umap": ["umap_1", "umap_2"]
    }
)
Contents