oneapi::mkl::rng::multinomial¶
Generates multinomially distributed random numbers.
Description¶
The oneapi::mkl::rng::multinomial
class object is used in the oneapi::mkl::rng::generate
function to provide multinomially distributed random numbers with ntrial
independent trials and k
possible mutually exclusive outcomes,
with corresponding probabilities p
i, where p
i∈R
; 0 ≤p
i≤ 1, m∈N, k∈N
.
The probability distribution is given by:
Product and Performance Information |
---|
Performance varies by use, configuration, and other factors. Learn more at https://www.intel.com/PerformanceIndex. Notice revision #20201201 |
API¶
Syntax¶
template<typename IntType = std::int32_t, typename Method = multinomial_method::by_default>
class multinomial {
public:
using method_type = Method;
using result_type = IntType;
explicit multinomial(double ntrial, std::vector<double> p); // deprecated since oneMKL 2023.0
explicit multinomial(double ntrial, sycl::span<double> p);
explicit multinomial(const param_type& pt);
std::int32_t ntrial() const;
std::vector<double> p() const;
param_type param() const;
void param(const param_type& pt);
};
Devices supported: CPU and GPU
Note
This distribution supports only input parameters (on GPU) that follow
the condition \((k \geq ntrial * 16 \land ntrial \leq 16)\),
where k
is the probability vector length.
Include Files¶
oneapi/mkl/rng.hpp
Template Parameters¶
|
Type of the produced values. The specific values are as follows:
|
|
Generation method. The specific values are as follows:
See brief descriptions of the methods in Distributions Template Parameter Method. |
Input Parameters¶
Name |
Type |
Description |
---|---|---|
ntrial |
|
Number of independent trials |
p |
|
Probability vector of possible outcomes (length is |
Note
explicit multinomial(double ntrial, std::vector<double> p);
is deprecated and will be removed in one of the next releases. Useexplicit multinomial(double ntrial, sycl::span<double> p);
instead.When passing a
sycl::span
that is constructed over a user’s memory to the constructor, users must manage the memory under sycl::span by themselves. They must not destroy the memory while data are processed.