Multinomial (VSL_RNG_METHOD_MULTINOMIAL_MULTPOISSON)
Multinomial distribution with parameters
m
, k
, and a probability vector p
. Random numbers of the multinomial distribution are generated by Poisson Approximation method (see [Charles93] for details).- In the first stage,kindependent Poisson values(Xare generated by the POISSNORM method.1...Xk)
- Letm*denote sum of the generatedkPoisson variates:
- Ifm*=m, the first-stage sample has the required distribution.
- Ifm*>m, the sample is discarded and the first stage is repeated.
- m*-muniformly distributed independent random variatesUare generated on the interval (0, 1).i
- The componentXis incremented by 1 ifi
See Intel® oneAPI Math Kernel Library Vector Statistics Random Number Generator Performance Data for test results summary and performance graphs.