On Jan 18, 2011, at 8:54 PM, Lee Spector wrote:
> Also BTW I saw no clear performance improvement (relative to the overall
> runtime of my system); some numbers on that are also below.
A perf differential will be seen if slot access is a significant portion of the
algorithms in question. IIRC, the difference between struct maps and records
is that between a map lookup + an array dereference vs. a direct field access
(e.g. (.slotname foo)) or a function call that performs a direct field access
(e.g. (:slotname foo)). Irresponsible benchmarks below. So, if your usage
doesn't involve a ton of slot access, you won't see a big perf improvement.
- Chas
-----
Irresponsible microbenchmark:
(defstruct foo [:a :b :c :d])
#'user/foo
(defrecord Bar [a b c d])
user.Bar
(time
(let [s (struct-map foo 1 2 3 4)]
(dotimes [x 1e7]
(:a s)
(:b s)
(:c s)
(:d s))))
"Elapsed time: 4734.852 msecs"
nil
(time
(let [s (Bar. 1 2 3 4)]
(dotimes [x 1e7]
(:a s)
(:b s)
(:c s)
(:d s))))
"Elapsed time: 721.399 msecs"
;; I've not checked this, but I suspect the differential will be even larger
for structs/records with lots of slots
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to [email protected]
Note that posts from new members are moderated - please be patient with your
first post.
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
