Hi,
I'm looking for known libraries that are able to generate non uniformly distributed random numbers for C, C++ and Java.
Thanks
+5
A:
I got some interesting responses in this related question:
http://stackoverflow.com/questions/1396978/biased-random-number-sources
uosɐſ
2009-11-06 23:29:57
+3
A:
For Java, one option is my Uncommons Maths library. It supports Uniform, Gaussian, Binomial, Poisson and Exponential distributions. There is a WebStart demo so you can see what it does.
Dan Dyer
2009-11-06 23:31:48
+2
A:
Have a look at Alglib's implementations, they have a few basic distributions implemented in several languages.
Jonatan Lindén
2009-11-06 23:35:22
This link leads algorithms for the calculation of several distribution functions, but I didn't find any algorithms for the calculation of random numbers drawn from the distributions.
M. S. B.
2009-11-07 06:02:22
Yes, that's true, you have to use a separate RNG. For example invpoissondistribution(k, rng.nextDouble()) would give you a Poisson variate with parameter k.
Jonatan Lindén
2009-11-07 09:44:26
A:
Numerical Recipes discusses a few algorithms for random number generators.
Parappa
2009-11-07 00:11:37
Actually i already wrote some algorithms for random number generators: Box-Muller, Rejection method among others.. So i'm just looking for libraries that has the reputation to generate strong random numbers non uniformly distributed
Amokrane
2009-11-07 00:19:05
+1
A:
Boost has a fairly wide selection of random number generates, plus the ability to filter these through several distributions.
Russell Newquist
2009-11-07 00:14:46
+2
A:
The GNU Scientific Library (GSL), http://www.gnu.org/software/gsl/, provides numerous non-uniform random distributions -- see Chapter 19 of the Manual, "Random Number Distributions". (Uniform random number generators are in Chapter 17, "Random Number Generation"). The implementation is in C.
M. S. B.
2009-11-07 05:38:12