msmu.dt.assign

Assign a column in place and record its values for replay.

Equivalent to mdata[modality].var[column] = values (or .obs), or to mdata[modality].uns["decoy"][column] = values with on="decoy". pandas handles scalar broadcasting, positional arrays, and Series index alignment. Returns the same MuData. Values are stored in full in provenance; the computation that produced them is not recorded.

Parameters:

Name	Type	Description	Default
mdata	MuData	MuData to modify.	required
column	str	Destination column name; an existing column is overwritten.	required
values	Any	Scalar, positional array, or pandas Series. Series align by index. Values are captured in full for replay.	required
modality	str	Modality containing the target table.	required
on	Literal['obs', 'var', 'decoy']	"var", "obs", or "decoy" for the decoy table in .uns["decoy"], which add_filter requires to carry every column a variable filter uses.	'var'

Returns:

Type	Description
MuData	The same MuData, modified in place.

Examples:

import msmu as mm
mm.dt.assign(mdata, "reviewed", True, modality="protein")
mm.dt.assign(mdata, "n_missing", 0, modality="protein", on="decoy")