I am experimenting with the Mersenne Twister under CUDA SDK and would like to know if the samples are uniform over [0,1) or [0,1].
Is there a way to modify the code so that I can get all 4 ([0,1], (0,1), [0,1), (0,1]) varieties ?
Regards
You should take a look at the CUDA implementation by the original authors of MT:
http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/MTGP/index.html