mt.pandas.dataclass

Functions dealing with dataclass columns.

Functions

• pack_dataclass(): Packs a list of component fields of a dataframe as a dataclass field.

• unpack_dataclass(): Unpacks a dataclass field of a dataframe as a list of component fields.

mt.pandas.dataclass.pack_dataclass(df, output_col: str, klass, field_map: dict | None = None, drop: bool = False)

Packs a list of component fields of a dataframe as a dataclass field.

Parameters:

• df (pandas.DataFrame) – the input dataframe to work on

• output_col (str) – the name of the output dataclass field

• klass (object) – the dtype of the output dataclass field

• field_map (dict, optional) – if provided, a mapping that maps each component field’s name into a field of the dataclass

• drop (bool) – whether or not (default) to drop the component fields in the dataframe

Returns:

the output dataframe, where the dataclass field is appended to the input dataframe, and optionally the component fields dropped

Return type:

pandas.DataFrame

mt.pandas.dataclass.unpack_dataclass(df, input_col: str, klass: object | None = None, field_map: dict | None = None, drop: bool = False)

Unpacks a dataclass field of a dataframe as a list of component fields.

Parameters:

• df (pandas.DataFrame) – the input dataframe to work on

• input_col (str) – the name of the input dataclass field

• klass (object, optional) – the dtype of the input dataclass field. To be detected if not provided.

• field_map (dict, optional) – if provided, a mapping that maps each component field’s name into a field of the dataclass

• drop (bool) – whether or not (default) to drop the dataclass field after unpacking

Returns:

the output dataframe, where new component fields are generated, and optionally the dataclass field dropped

Return type:

pandas.DataFrame