data_designer.config.custom_column

User-facing utilities for custom column generation.

Module Contents

Functions

Name	Description
`validate_generator_signature`	Validate generator function signature. Returns positional params if valid.
`custom_column_generator`	Decorator to define metadata for a custom column generator function.

Data

F EXPECTED_PARAMS

API

1 F = TypeVar(...)

1 EXPECTED_PARAMS = ()

1 data_designer.config.custom_column.validate_generator_signature(fn: typing.Callable[..., typing.Any]) -> list[inspect.Parameter]

Validate generator function signature. Returns positional params if valid.

1 data_designer.config.custom_column.custom_column_generator(
2     required_columns: list[str] | None = None,
3     side_effect_columns: list[str] | None = None,
4     model_aliases: list[str] | None = None
5 ) -> typing.Callable[[data_designer.config.custom_column.F], data_designer.config.custom_column.F]) -> typing.Callable[[data_designer.config.custom_column.F], data_designer.config.custom_column.F]

Decorator to define metadata for a custom column generator function.

Parameters:

required_columns

list[str] | NoneDefaults to None

Columns that must exist before this column runs (DAG ordering).

side_effect_columns

list[str] | NoneDefaults to None

Additional columns the function will create.

model_aliases

list[str] | NoneDefaults to None

Model aliases to include in the models dict (required for LLM access).

1	data_designer.config.custom_column.custom_column_generator(
2	required_columns: list[str] \| None = None,
3	side_effect_columns: list[str] \| None = None,
4	model_aliases: list[str] \| None = None
5	) -> typing.Callable[[data_designer.config.custom_column.F], data_designer.config.custom_column.F]) -> typing.Callable[[data_designer.config.custom_column.F], data_designer.config.custom_column.F]