tlc.reduction.reduction_method

Base classes for dimensionality reducers.

Module Contents

Classes

Class	Description
ReducerArgs	Arguments specific to the reduction method.
ReductionMethod	Perform dimensionality reduction on columns of tables.

Data

Data	Description
ReducerArgsType

API

class ReducerArgs

Bases: typing.TypedDict

Arguments specific to the reduction method.

Initialize self. See help(type(self)) for accurate signature.

random_state: int | None = None

The random state to use for the reduction (default None).

retain_source_embedding_column: bool = None

Whether to retain the source embedding column in the output table (default False).

source_embedding_column: str | None = None

The name of the column containing the unreduced embedding in the input table. If None, all columns are considered .

target_embedding_column: str | None = None

The name of the column containing the reduced embedding in the output table. If None, the column is named after the input column.

ReducerArgsType = TypeVar(...)

class ReductionMethod(

reducer_args: tlc.reduction.reduction_method.ReducerArgsType | None = None,

)

Bases: abc.ABC, typing.Generic[tlc.reduction.reduction_method.ReducerArgsType]

Perform dimensionality reduction on columns of tables.

Dimensionality reduction is accomplished by creating derived tables of the appropriate type (e.g., UmapTable) which apply the reduction method to the input tables.

Parameters:

reducer_args – A dictionary of arguments which are specific to the reduction method.

abstract apply_reduction_method(

table: Table,

fit_table_url: Url,

column: str,

) → Url | None

abstract default_args() → tlc.reduction.reduction_method.ReducerArgsType

fit_and_apply_reduction(

producer: Table,

consumers: list[Table] | None = None,

) → dict[Url, Url]

abstract fit_reduction_method(

table: Table,

column: str,

) → Url