giles
a3f9d4f6c9
Some checks failed
Test, Build, and Deploy / test-build-deploy (push) Failing after 37s
rankTfIdf and rankBm25 (configurable k1/b) over the candidate set, float scores with deterministic DocId tiebreak; topNTfIdf/topNBm25. df/idf derived from posting-list length. Tests cover tf/idf behavior, a BM25-vs-TF-IDF flip from length-norm + tf-saturation, the b-parameter effect, tiebreak stability. 101/101. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
91 lines
2.8 KiB
Plaintext
91 lines
2.8 KiB
Plaintext
;; Phase 3 — ranking (TF-IDF, BM25, top-N). Deterministic: ties broken by DocId.
|
|
;; Corpora:
|
|
;; idx1: 1 "alpha alpha alpha gamma" 2 "alpha" 3 "beta"
|
|
;; idx2: 1 "cat" 2 "cat cat dog elephant frog grape" 3 "zzz"
|
|
;; idx3: 1 "kite" 2 "kite" (identical docs -> tiebreak)
|
|
|
|
(define
|
|
rank-setup
|
|
"idx1 = indexDoc 3 \"beta\" (indexDoc 2 \"alpha\" (indexDoc 1 \"alpha alpha alpha gamma\" emptyIndex))\nidx2 = indexDoc 3 \"zzz\" (indexDoc 2 \"cat cat dog elephant frog grape\" (indexDoc 1 \"cat\" emptyIndex))\nidx3 = indexDoc 2 \"kite\" (indexDoc 1 \"kite\" emptyIndex)\n")
|
|
|
|
(define
|
|
rank-cases
|
|
(list
|
|
(list
|
|
"tfidf tf ordering"
|
|
"rankTfIdf [\"alpha\"] idx1"
|
|
(list 1 2))
|
|
(list
|
|
"tfidf rare term boosts"
|
|
"rankTfIdf [\"alpha\", \"beta\"] idx1"
|
|
(list 1 3 2))
|
|
(list
|
|
"tfidf single-doc term"
|
|
"rankTfIdf [\"gamma\"] idx1"
|
|
(list 1))
|
|
(list "tfidf absent term empty" "rankTfIdf [\"nope\"] idx1" (list))
|
|
(list "tfidf empty query empty" "rankTfIdf [] idx1" (list))
|
|
(list
|
|
"tfidf candidate union tie by docid"
|
|
"rankTfIdf [\"beta\", \"gamma\"] idx1"
|
|
(list 1 3))
|
|
(list
|
|
"tfidf tf ordering idx2"
|
|
"rankTfIdf [\"cat\"] idx2"
|
|
(list 2 1))
|
|
(list "topN tfidf 1" "topNTfIdf 1 [\"alpha\"] idx1" (list 1))
|
|
(list
|
|
"topN tfidf 2"
|
|
"topNTfIdf 2 [\"alpha\", \"beta\"] idx1"
|
|
(list 1 3))
|
|
(list
|
|
"topN exceeds results"
|
|
"topNTfIdf 10 [\"gamma\"] idx1"
|
|
(list 1))
|
|
(list "topN zero" "topNTfIdf 0 [\"alpha\"] idx1" (list))
|
|
(list
|
|
"bm25 tf+length flips tfidf"
|
|
"rankBm25 1.5 0.75 [\"cat\"] idx2"
|
|
(list 1 2))
|
|
(list
|
|
"bm25 b=0 ignores length"
|
|
"rankBm25 1.5 0.0 [\"cat\"] idx2"
|
|
(list 2 1))
|
|
(list
|
|
"bm25 alpha idx1"
|
|
"rankBm25 1.5 0.75 [\"alpha\"] idx1"
|
|
(list 1 2))
|
|
(list "bm25 absent empty" "rankBm25 1.5 0.75 [\"nope\"] idx1" (list))
|
|
(list
|
|
"bm25 single-doc term"
|
|
"rankBm25 1.5 0.75 [\"gamma\"] idx1"
|
|
(list 1))
|
|
(list "bm25 topN 1" "topNBm25 1 1.5 0.75 [\"cat\"] idx2" (list 1))
|
|
(list
|
|
"bm25 same candidate set"
|
|
"sort (rankBm25 1.5 0.75 [\"alpha\", \"beta\"] idx1)"
|
|
(list 1 2 3))
|
|
(list
|
|
"tfidf stable tiebreak"
|
|
"rankTfIdf [\"kite\"] idx3"
|
|
(list 1 2))
|
|
(list
|
|
"bm25 stable tiebreak"
|
|
"rankBm25 1.5 0.75 [\"kite\"] idx3"
|
|
(list 1 2))
|
|
(list "numDocs" "[numDocs idx1]" (list 3))
|
|
(list "docLen counts tokens" "[docLen 1 idx1]" (list 4))
|
|
(list "docFreq via index" "[docFreq \"alpha\" idx1]" (list 2))))
|
|
|
|
(define
|
|
rank-results
|
|
(search-batch rank-setup (map (fn (c) (nth c 1)) rank-cases)))
|
|
|
|
(map-indexed
|
|
(fn
|
|
(i c)
|
|
(hk-test (nth c 0) (nth rank-results i) (nth c 2)))
|
|
rank-cases)
|
|
|
|
{:fail hk-test-fail :pass hk-test-pass :fails hk-test-fails}
|