kQjfddlmZmZmZddlZddlZddlm Z ddl m Z erddl m Z Gdd e Zy) )UnionList TYPE_CHECKINGN) defaultdict)Reranker)LanceVectorQueryBuilderceZdZdZ d dedeffd ZdZdedejdejfd Z dde e eje d fded e fd ZxZS) MRRRerankera Reranks the results using Mean Reciprocal Rank (MRR) algorithm based on the scores of vector and FTS search. Algorithm reference - https://en.wikipedia.org/wiki/Mean_reciprocal_rank MRR calculates the average of reciprocal ranks across different search results. For each document, it computes the reciprocal of its rank in each system, then takes the mean of these reciprocal ranks as the final score. Parameters ---------- weight_vector : float, default 0.5 Weight for vector search results (0.0 to 1.0) weight_fts : float, default 0.5 Weight for FTS search results (0.0 to 1.0) Note: weight_vector + weight_fts should equal 1.0 return_score : str, default "relevance" Options are "relevance" or "all" The type of score to return. If "relevance", will return only the relevance score. If "all", will return all scores from the vector and FTS search along with the relevance score. weight_vector weight_ftscd|cxkrdkstdtdd|cxkrdkstdtdt||zdz dkDr tdt| |||_||_y)N?z)weight_vector must be between 0.0 and 1.0z&weight_fts must be between 0.0 and 1.0gư>z)weight_vector + weight_fts must equal 1.0) ValueErrorabssuper__init__r r)selfr r return_score __class__s Z/home/oem/CyberBrain26-unified/.venv/lib/python3.12/site-packages/lancedb/rerankers/mrr.pyrzMRRReranker.__init__(s }++HI I,HI Iz(S(EF F)EF F }z)C/ 04 7HI I &*$c<d|jd|jdS)NzMRRReranker(weight_vector=z , weight_fts=))r r)rs r__str__zMRRReranker.__str__9s+(););(<=//*! - rqueryvector_results fts_resultsc|r|djng}|r|djng}tt}|r/t|dD] \}}d|z } ||j d| f"|r/t|dD] \}}d|z } ||j d| f"i} |j D]G\}} d} d} | D]\}} |dk(r| } |dk(s| } |j | z|j| zz}|| |<I|jdk(r|j||}n|j||}|dj}|Dcgc]}| | }}|jdtj|tj }|jd g}|jd k(r|j!|}|Scc}w) N_rowidrrvectorftsrall_relevance_scoretyper& descending relevance) to_pylistrlist enumerateappenditemsr rscore_merge_and_keep_scores merge_results append_columnpaarrayfloat32sort_by_keep_relevance_score)rrrr vector_idsfts_ids mrr_score_maprank result_idreciprocal_rankfinal_mrr_scoresscores vector_rrfts_rr score_type weighted_mrrcombined_resultscombined_row_idsrow_idrelevance_scoress r rerank_hybridzMRRReranker.rerank_hybrid?s >L^H-779QS 7B+h'113$D) #,Z#; Mi"%*i(//?0KL M #,Wa#8 Ji"%*i(//0HI J!.!4!4!6 7 IvIF/5 -+ O) /I5(,F  - -- 9DOOf;W #11.+N +H5??ACST,V4TT+99 )9 M ,33 / 0  :: $#99:JK Us Gr deduplicatec6s tdtfdDs tdtdjdk(rDcgc]}|j c}n(t dt js tdtdDs tdtt}D]B}|d j}t|d D]\}} d |z } || j|  Dt} i} |jD]+\} } tt!j"| | z }|| | <-t j$fi|j&}|j)|}|d j}|Dcgc]}| | }}|j+d t j,|t j. }|j1dg}|j2dk(r|j5|}|Scc}wcc}w)aw Reranks the results from multiple vector searches using MRR algorithm. Each vector search result is treated as a separate ranking system, and MRR calculates the mean of reciprocal ranks across all systems. This cannot reuse rerank_hybrid because MRR semantics require treating each vector result as a separate ranking system. z vector_results must not be emptyc3NK|]}t|tdyw)rN) isinstancer().0vrs r z1MRRReranker.rerank_multivector..s"Ra:anQ&7!89Rs"%z9All elements in vector_results should be of the same typerr zFvector_results should be a list of pa.Table or LanceVectorQueryBuilderc38K|]}d|jvyw)r"N) column_names)rOresults rrQz1MRRReranker.rerank_multivector..sPv8v222Psz'_rowid' is required for deduplication. add _rowid to search results like this: `search().with_row_id(True)`r"rrr&r'r)r+)rr%r(__name__to_arrowrNr5Tablerr-r,r.r/lenr0floatnpsum concat_tables_concat_tables_args _deduplicater4r6r7r8r1r9)rrrrKrTr< result_table result_idsr=r>r? num_systemsr@reciprocal_ranksmean_rrcombinedrGrHrIs ` rrerank_multivectorzMRRReranker.rerank_multivectorxs,?@ @R>RRK  q! " + +/H H>LMFfoo/MNN1-rxx8X PPP2  $D) * AL%h/99;J#,Z#; Ai"%*i(//@ A A.) +8+>+>+@ 2 'I'BFF#345 CG*1 Y ' 2##NOd6N6NO$$X.#H-779CST,V4TT)) )9 M ##%G$HI :: $11(;H_NLUs H H)?rfr+)NT)rU __module__ __qualname____doc__rYrrstrr5rWrJrrboolre __classcell__)rs@rr r s2 # %%%" 7 7 7 XX 7 x Fd288nd3L.MMNFF Frr )typingrrrpyarrowr5numpyrZ collectionsrbasertabler r rrrts- .-#/n(nr