reciprocal_rank_fusion#

scikitplot.mcp.reciprocal_rank_fusion(ranked_lists, *, k=60, key=<function _chunk_key>)[source]#

Fuse weighted ranked lists into a single key -> score map.

Parameters:

ranked_listssequence of (weight, items)

Each leg’s weight and its best-first ranked items.

kint, optional

RRF constant (default DEFAULT_RRF_K).

keycallable, optional

Maps an item to its fusion identity (default _chunk_key).

Returns:

dict

identity -> fused score (higher is better).

Parameters:

• ranked_lists (Sequence[tuple[float, Sequence[Any]]])

• k (int)

• key (Callable[[Any], str])

Return type:

dict[str, float]