msmu.pp.add_meta

Attach metadata by exact matching of one metadata key to one observation key.

Parameters:

Name	Type	Description	Default
mdata	MuData	MuData whose modality and global .obs annotations will be updated in a copy.	required
metadata	DataFrame | str | PathLike[str]	DataFrame, local file, or URL containing metadata.	required
format	Literal['dataframe', 'df', 'generic', 'csv', 'tsv', 'parquet', 'sdrf'] | None	dataframe/df/generic, csv, tsv, parquet, or sdrf. None accepts a DataFrame or infers the file format from its name.	None
metadata_on	str | Sequence[str] | None	Metadata join column; None uses its index. A sequence must contain exactly one key.	None
match_columns	str | Sequence[str] | None	Alias of metadata_on; cannot be supplied together with it.	None
obs_columns	str | Sequence[str] | None	Join column in each modality .obs; None uses each observation index. A sequence must contain exactly one key.	None
validate_sdrf	bool	Validate SDRF input when using SDRF format.	True
skip_ontology	bool	Skip ontology checks during SDRF validation.	True

Returns:

Type	Description
MuData	A new MuData with metadata merged into each modality and global .obs; the original is unchanged.

Notes

Metadata must be nonempty with unique columns and unique non-null join keys. Each modality must have at least one matching row; unmatched rows receive missing metadata. Existing non-null annotations take precedence; only gaps are filled. For separate SDRF storage/application, see attach_sdrf and apply_sdrf_to_obs.

Examples:

import msmu as mm
import pandas as pd
metadata = pd.DataFrame({"condition": ["control"] * mdata.n_obs}, index=mdata.obs_names)
mdata = mm.pp.add_meta(mdata, metadata)