gp_matern52_cov_cross()

__kernel void stan::math::opencl_kernels::gp_matern52_cov_cross	(	const __global double *	x1,
		const __global double *	x2,
		__global double *	res,
		const double	sigma_sq,
		const double	root_5_inv_l,
		const double	inv_l_sq_5_3,
		const int	size1,
		const int	size2,
		const int	element_size
	)

GPU part of calculation of Matern 5/2 kernel.

This function is for the cross covariance matrix needed to compute the posterior predictive density.

Parameters

[in]	x1	first input vector or matrix
[in]	x2	second input vector or matrix
[out]	res	squared distances between elements of x
	sigma_sq	squared standard deviation
	root_5_inv_l	sqrt(5.0) / length_scale
	inv_l_sq_5_3	5.0 / 3.0 / square(length_scale)
	size1	number of elements in x1
	size2	number of elements in x2
	element_size	the number of doubles that make one element of x and y

Definition at line 78 of file gp_matern52_cov.hpp.