I have a huge graph of the form
{"john" {1 {"alice" 1, "bob" 3}, 4 {"alice" 2, "stan" 1}, "alice" {1
{"john" 1, "mandy" 2}, 2 {"john" 3, "stan" 2}}}
It shows for each user days on which he communicated, and for each
day, with whom and how many times he addressed that person.
Essentially this is an adjacency list -- here it's a map --
representation of a graph.
I need to verify that for any top-level user, the days in his
adjacency list, as returned by seq'ing the adjacency map. I've
written these functions to do it, with the help of the #clojure folks:
(defn reps-sorted1? [dreps & [progress]]
(loop [[[user ureps :as userdays] :as users] (seq dreps) i 0]
(if ureps
(if (loop [prev 0 [[day _] :as days] (seq ureps)]
(when (and progress (= 0 (mod i progress))) (print ".")
(.flush System/out))
(if day (if (<= prev day)
(recur day (next days))
(do (println userdays) false)) true))
(recur (next users) (inc i)) false) true)))
(defn reps-sorted2? [dreps]
(->> dreps (pmap (fn [[_ days]]
(let [s (map first days)]
(every? true? (map >= s (cons (first s) s))))))
(every? true?)))
user=> (time (reps-sorted2?
dreps))
"Elapsed time: 127606.905 msecs"
true
user=> (time (reps-sorted1? dreps))
"Elapsed time: 8368.616 msecs"
true
As you can see, the imperative solution vastly outperforms the
functional one, even though the functional allows for pmap! Without
the pmap, it actually did faster, in 40 seconds. Why is FP slower
here?
-- Alexy
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your
first post.
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
