rows_dot_product.hpp

Go to the documentation of this file.

#ifndef STAN_MATH_REV_FUN_ROWS_DOT_PRODUCT_HPP
#define STAN_MATH_REV_FUN_ROWS_DOT_PRODUCT_HPP
 
#include <stan/math/rev/meta.hpp>
#include <stan/math/rev/core.hpp>
#include <stan/math/rev/core/typedefs.hpp>
#include <stan/math/rev/fun/dot_product.hpp>
#include <stan/math/prim/err.hpp>
#include <stan/math/prim/fun/Eigen.hpp>
#include <stan/math/prim/fun/rows_dot_product.hpp>
#include <type_traits>
 
namespace stan {
namespace math {
 
template <typename Mat1, typename Mat2,
          require_all_eigen_t<Mat1, Mat2>* = nullptr,
          require_any_eigen_vt<is_var, Mat1, Mat2>* = nullptr>
inline Eigen::Matrix<var, Mat1::RowsAtCompileTime, 1> rows_dot_product(
    const Mat1& v1, const Mat2& v2) {
  check_matching_sizes("dot_product", "v1", v1, "v2", v2);
  Eigen::Matrix<var, Mat1::RowsAtCompileTime, 1> ret(v1.rows(), 1);
  for (size_type j = 0; j < v1.rows(); ++j) {
    ret.coeffRef(j) = dot_product(v1.row(j), v2.row(j));
  }
  return ret;
}
 
template <typename Mat1, typename Mat2,
          require_all_matrix_t<Mat1, Mat2>* = nullptr,
          require_any_var_matrix_t<Mat1, Mat2>* = nullptr>
inline auto rows_dot_product(const Mat1& v1, const Mat2& v2) {
  check_matching_sizes("rows_dot_product", "v1", v1, "v2", v2);
 
  using return_t = return_var_matrix_t<
      decltype((v1.val().array() * v2.val().array()).rowwise().sum().matrix()),
      Mat1, Mat2>;
 
  if (!is_constant<Mat1>::value && !is_constant<Mat2>::value) {
    arena_t<promote_scalar_t<var, Mat1>> arena_v1 = v1;
    arena_t<promote_scalar_t<var, Mat2>> arena_v2 = v2;
 
    return_t res
        = (arena_v1.val().array() * arena_v2.val().array()).rowwise().sum();
 
    reverse_pass_callback([arena_v1, arena_v2, res]() mutable {
      if (is_var_matrix<Mat1>::value) {
        arena_v1.adj().noalias() += res.adj().asDiagonal() * arena_v2.val();
      } else {
        arena_v1.adj() += res.adj().asDiagonal() * arena_v2.val();
      }
      if (is_var_matrix<Mat2>::value) {
        arena_v2.adj().noalias() += res.adj().asDiagonal() * arena_v1.val();
      } else {
        arena_v2.adj() += res.adj().asDiagonal() * arena_v1.val();
      }
    });
 
    return res;
  } else if (!is_constant<Mat2>::value) {
    arena_t<promote_scalar_t<double, Mat1>> arena_v1 = value_of(v1);
    arena_t<promote_scalar_t<var, Mat2>> arena_v2 = v2;
 
    return_t res = (arena_v1.array() * arena_v2.val().array()).rowwise().sum();
 
    reverse_pass_callback([arena_v1, arena_v2, res]() mutable {
      if (is_var_matrix<Mat2>::value) {
        arena_v2.adj().noalias() += res.adj().asDiagonal() * arena_v1;
      } else {
        arena_v2.adj() += res.adj().asDiagonal() * arena_v1;
      }
    });
 
    return res;
  } else {
    arena_t<promote_scalar_t<var, Mat1>> arena_v1 = v1;
    arena_t<promote_scalar_t<double, Mat2>> arena_v2 = value_of(v2);
 
    return_t res = (arena_v1.val().array() * arena_v2.array()).rowwise().sum();
 
    reverse_pass_callback([arena_v1, arena_v2, res]() mutable {
      if (is_var_matrix<Mat2>::value) {
        arena_v1.adj().noalias() += res.adj().asDiagonal() * arena_v2;
      } else {
        arena_v1.adj() += res.adj().asDiagonal() * arena_v2;
      }
    });
 
    return res;
  }
}
 
}  // namespace math
}  // namespace stan
#endif