Re: [Boost-users] Random library, discrete distribution

Steven Watanabe wrote:

Your test isn't quite right. You're comparing the speed of categorical to itself for sampling.

Here's a correction: --initialize, n = 10000 categorical : t = 0.013616 discrete : t = 0.001571 --sample, m = 100000 categorical : t = 0.029282 discrete : t = 0.014483