Dear all, Thanks to enlightening advice from Tom Nielsen and Zoltan Novak normaldistribution now generates normally distributed values using the Box-Muller method instead of the Central Limit Theorem. This change provides a 5-10x speedup. The only API impact is that the random values must now be instances of Floating rather than just Fractional. (The pre-canned Random instances that satisfy Fractional, i.e. Double and Float, also satisfy Floating.) Cheers, Bjorn ---------- Forwarded message ---------- From: Bjorn Buckwalter Date: Sat, Apr 9, 2011 at 15:21 Subject: ANNOUNCE: normaldistribution-1.0 – Minimum fuss normally distributed random values. Dear all, I'm pleased to announce the immediate availability of the normaldistribution library on hackage:  http://hackage.haskell.org/package/normaldistribution This purpose of this library is to have a simple API and no dependencies beyond Haskell 98 in order to let you produce normally distributed random values with a minimum of fuss. This library does not attempt to be blazingly fast nor to pass stringent tests of randomness. It attempts to be very easy to install and use while being "good enough" for many applications (simulations, games, etc.).  The API builds upon and is largely analogous to that of the Haskell 98 Random module (more recently System.Random). Pure:

(sample,g) = normal  myRandomGen  -- using a Random.RandomGen samples    = normals myRandomGen  -- infinite list samples2   = mkNormals 10831452   -- infinite list using a seed

In the IO monad:

sample    <- normalIO samples   <- normalsIO  -- infinite list

With custom mean and standard deviation:

(sample,g) = normal'    (mean,sigma) myRandomGen samples    = normals'   (mean,sigma) myRandomGen samples2   = mkNormals' (mean,sigma) 10831452

sample    <- normalIO'  (mean,sigma) samples   <- normalsIO' (mean,sigma)

Internally the library uses the Central Limit Theorem to approximate normally distributed values from multiple uniformly distributed random values. Enjoy, Bjorn