poisson_lpmf.hpp

Go to the documentation of this file.

#ifndef STAN_MATH_OPENCL_PRIM_POISSON_LPMF_HPP
#define STAN_MATH_OPENCL_PRIM_POISSON_LPMF_HPP
#ifdef STAN_OPENCL
 
#include <stan/math/prim/meta.hpp>
#include <stan/math/prim/err.hpp>
#include <stan/math/prim/fun/constants.hpp>
#include <stan/math/prim/fun/log.hpp>
#include <stan/math/opencl/kernel_generator.hpp>
#include <stan/math/prim/functor/partials_propagator.hpp>
 
namespace stan {
namespace math {
 
template <
    bool propto, typename T_n_cl, typename T_rate_cl,
    require_all_prim_or_rev_kernel_expression_t<T_n_cl, T_rate_cl>* = nullptr,
    require_any_not_stan_scalar_t<T_n_cl, T_rate_cl>* = nullptr>
return_type_t<T_rate_cl> poisson_lpmf(const T_n_cl& n,
                                      const T_rate_cl& lambda) {
  static constexpr const char* function = "poisson_lpmf(OpenCL)";
  using T_partials_return = partials_return_t<T_rate_cl>;
  using std::isinf;
  constexpr bool is_n_vector = !is_stan_scalar<T_n_cl>::value;
 
  check_consistent_sizes(function, "Random variable", n, "Rate parameter",
                         lambda);
  const size_t N = is_n_vector ? math::size(n) : math::size(lambda);
  if (N == 0) {
    return 0.0;
  }
  if (!include_summand<propto, T_rate_cl>::value) {
    return 0.0;
  }
 
  const auto& lambda_col = as_column_vector_or_scalar(lambda);
  const auto& lambda_val = value_of(lambda_col);
 
  T_partials_return logp(0.0);
  auto ops_partials = make_partials_propagator(lambda_col);
 
  auto check_n_nonnegative
      = check_cl(function, "Random variable", n, "nonnegative");
  auto n_nonnegative = 0 <= n;
  auto check_lambda_nonnegative
      = check_cl(function, "Log rate parameter", lambda_val, "nonnegative");
  auto lambda_nonnegative = 0.0 <= lambda_val;
 
  auto return_log_zero = colwise_max(
      cast<char>(isinf(lambda_val) || (lambda_val == 0 && n != 0)));
 
  auto logp1 = multiply_log(n, lambda_val);
  auto logp2 = static_select<include_summand<propto, T_rate_cl>::value>(
      logp1 - lambda_val, logp1);
  auto logp_expr = colwise_sum(static_select<include_summand<propto>::value>(
      logp2 - lgamma(n + 1.0), logp2));
 
  auto deriv = elt_divide(n, lambda_val) - 1.0;
 
  matrix_cl<char> return_log_zero_cl;
  matrix_cl<double> logp_cl;
  matrix_cl<double> deriv_cl;
 
  results(check_n_nonnegative, check_lambda_nonnegative, return_log_zero_cl,
          logp_cl, deriv_cl)
      = expressions(n_nonnegative, lambda_nonnegative, return_log_zero,
                    logp_expr,
                    calc_if<!is_constant_all<T_rate_cl>::value>(deriv));
 
  if (from_matrix_cl(return_log_zero_cl).any()) {
    return LOG_ZERO;
  }
 
  logp = sum(from_matrix_cl(logp_cl));
 
  if (!is_constant_all<T_rate_cl>::value) {
    partials<0>(ops_partials) = deriv_cl;
  }
 
  return ops_partials.build(logp);
}
 
}  // namespace math
}  // namespace stan
#endif
#endif