On Wed, Dec 30, 2009 at 4:07 AM, Timothy Pratley <timothyprat...@gmail.com> wrote: > 2009/12/30 Rich Hickey <richhic...@gmail.com>: >> This was discussed before, the new version never made it into a patch: >> http://groups.google.com/group/clojure/msg/43de40f078a291cc > > Great! > > The 'old' reductions behaves slightly different from the 'new' > reductions for the 3 argument form: > foo=> (reductions + 0 [3 5 10 1 2 7]) > (0 3 8 18 19 21 28) > foo=> (reductions2 + 0 [3 5 10 1 2 7]) > (3 8 18 19 21 28) > > Which output is more correct?
The former. > On the one hand supplying a preserved > 'initial value' from outside does not make much sense, but on the > other hand perhaps it is more in step with 3 argument reduce. > > Does it matter? Probably not. It does provide an interesting case > study. The 3 argument form could be considered just a helper to > achieve the 2 argument form, but because it is part of the public > interface people may chose to rely on it: > (defn left-total3 [coll] > (map list coll (reductions + 0 coll))) > might have been better written as > (defn left-total3 [coll] > (map list coll (cons 0 (reductions + coll)))) > So is exposing the 3 argument version a bad thing? It is not very > idiomatic to use the alternatives: > (let-fn [(reducer [f acc coll] ...)] > (defn reductions [f coll] ... reducer ...)) > And then again if I can call (reduce + 3 [1 2]) maybe it is reasonable > to call (reductions + 3 [1 2]) > I'm over analyzing what is really a non-issue, but just found it > thought provoking. > > Back to more practical concerns, attached is a modified version for > consideration which preserves the old behavior but is faster. Happy to > supply a git patch if on the right track - let me know what is best :) > > Great - thanks. Rich -- 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
