On Fri, Apr 30, 2010 at 8:56 PM, Lee Spector <lspec...@hampshire.edu> wrote: > > In an earlier thread, in which I learned (from Timothy Pratley) that (. (new > java.util.Random) X) gives an error if X is a bignum, I said that at least > Clojure's rand-int "does the right thing." > > Upon further investigation I see that this is only true in the sense that it > doesn't produce an error.
You're right. Under Clojure 1.1.0, I ran the following: (def count-if (comp count filter)) (count-if #(= 2147483647 %) (take 1000 (repeatedly #(rand-int 10000000000)))) Just running once, it shows that 2147483647 occured 802 times in a thousand, which is... er, not so random. This looks like a side effect of a really screwy coercion to int that just shouldn't happen. And in fact, in versions of clojure more recent than commit d0e1eef2, you get a much more sensible error: user=> (int 10000000000) java.lang.IllegalArgumentException: Value out of range for int: 10000000000 (NO_SOURCE_FILE:0) Of course, you might actually *want* a few really large random numbers. In clojure.contrib.math, we have 'round' which can help you make some big numbers: (take 10 (repeatedly #(round (rand 10000000000000)))) Of course, at some point, the resolution of the Double that Java returns from its built-in 'rand' function will cause some problems for the distribution of your really big numbers, since Java's 'rand' returns a number between 0 and 1 that Clojure multiplies by the parameter. Then, I'd see what the incanter folks would recommend, since they probably know more about big random numbers than I do. Have fun! -- Chris Riddoch -- 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
