msmu.dt.replace

Replace column values in place using pandas-style replacement mappings.

columns maps column names to {old: new} rules. Unmatched values (including missing values unless explicitly mapped) remain unchanged. Restricted dtypes such as nullable booleans and categories are converted to object when needed to accommodate new values. Only rules are recorded. Table paths follow map. Returns the same MuData.

Parameters:

Name	Type	Description	Default
mdata	MuData	MuData to modify.	required
target	str	Table path as described by map.	required
columns	dict[str, dict]	Nonempty mapping from each column name to an {old_value: new_value} dictionary.	required

Returns:

Type	Description
MuData	The same MuData, modified in place.

Examples:

import msmu as mm
mm.dt.replace(mdata, target="protein.obs", columns={"condition": {"ctrl": "control"}})