Clojure Hash Test

hashtest.clj

(def testset (for [a (range 20)
                   b (range 20)
                   c (range 20)
                   d (range 20)]
               (conj #{[a b]} [c d])))
(def testvec (for [a (range 20)
                   b (range 20)
                   c (range 20)
                   d (range 20)]
               [[a b] [c d]]))
(def testvec2 (for [a (range 1000)
                    b (range 1000)]
               [a b]))

(defrecord Pair [a b])
(def testrec (for [a (range 1000)
                   b (range 1000)]
               (Pair. a b)))
(defn hashtest
  [items & [hash-fn]]
  (let [hash-fn (or hash-fn hash)
        v (into #{} items)
        h (map count (vals (group-by hash v)))
        n (count v)
        m (count h)]
    (prn n m (apply max h) (/ n m 1.0))))

; count-items count-hashes max-items-per-hash mean-items-per-hash

(hashtest testset)
; 80200 1615 200 49.6594427244582

(hashtest testvec)
; 160000 12560 20 12.73885350318471

(hashtest testvec2)
; 1000000 31969 33 31.28030279333104

(hashtest testrec)
; 1000000 2039 976 490.4364884747425