ff_projection.cpp File Reference

#include "plane.hpp"
#include "grid_node_tr_mapping.hpp"
#include "particles.hpp"
#include "xgc_io.hpp"
#include "my_subview.hpp"
#include "send_recv_potential.hpp"
#include "ff_projection.hpp"

Include dependency graph for ff_projection.cpp:

Functions
KOKKOS_INLINE_FUNCTION void	use_nearest_triangle (const Plane< DeviceType > &plane, const View< int **, CLayout, DeviceType > &tr_node, const SimdVector2D &x, bool weight_p_on_boundary, int i_simd, Simd< int > &itr, SimdGridVec &p_loc)
void	project (const Plane< DeviceType > &plane_start, const Plane< DeviceType > &plane_dest, double phi_start, double phi_dest, const MagneticField< DeviceType > &magnetic_field, const View< int **, CLayout, DeviceType > &tr_node, bool weight_p_on_boundary, const View< int *, CLayout, DeviceType > &tr_d, const View< int **, CLayout, DeviceType > &nd_d, const View< double **, CLayout, DeviceType > &p_d, const View< double *, CLayout, DeviceType > &dx_d, bool use_two_step_method)
static void	write_ff (const Plane< HostType > &plane_start, const Projection< HostType > &projection, const std::string stream_name, const std::string filename, bool reverse)
void	ff_write_projections_to_file (const Plane< HostType > &midplane, const Plane< HostType > &lplane, const Plane< HostType > &rplane, const Projection< HostType > &half_plane_ff, const Projection< HostType > &ff_lplane_to_neighbors, const Projection< HostType > &ff_to_midplane)
void	check_ff_property (const Projection< HostType > &projection, const std::string label)

Function Documentation

void check_ff_property	(	const Projection< HostType > &	projection,
		const std::string	label
	)

Here is the caller graph for this function:

void ff_write_projections_to_file	(	const Plane< HostType > &	midplane,
		const Plane< HostType > &	lplane,
		const Plane< HostType > &	rplane,
		const Projection< HostType > &	half_plane_ff,
		const Projection< HostType > &	ff_lplane_to_neighbors,
		const Projection< HostType > &	ff_to_midplane
	)

Here is the call graph for this function:

Here is the caller graph for this function:

void project	(	const Plane< DeviceType > &	plane_start,
		const Plane< DeviceType > &	plane_dest,
		double	phi_start,
		double	phi_dest,
		const MagneticField< DeviceType > &	magnetic_field,
		const View< int **, CLayout, DeviceType > &	tr_node,
		bool	weight_p_on_boundary,
		const View< int *, CLayout, DeviceType > &	tr_d,
		const View< int **, CLayout, DeviceType > &	nd_d,
		const View< double **, CLayout, DeviceType > &	p_d,
		const View< double *, CLayout, DeviceType > &	dx_d,
		bool	use_two_step_method
	)

Initialize the field-aligned mapping between adjacent planes. The routine traces magnetic field lines from each vertex of a plane in both toroidal directions to the adjacent mid-planes and full-step planes, records the triangle in which the field line intersects these planes and the corresponding interpolation weights.

Parameters

[in]	plane_start	the plane being projected from
[in]	plane_dest	the plane being projected to
[in]	phi_start	the toroidal angle of the plane being projected from
[in]	phi_dest	the toroidal angle of the plane being projected to
[in]	magnetic_field	the magnetic field for field following
[in]	tr_node	the triangle-to-node mapping of the destination plane
[in]	weight_p_on_boundary	whether to weight the nearest node if the projection falls outside the grid
[out]	tr_d	the triangles on the destination plane
[out]	nd_d	the nodes on the destination plane
[out]	p_d	the barycentric weights on the destination plane
[out]	dx_d	distance traveled along the field line
[in]	use_two_step_method	whether to use an intermediate value to increase accuracy of dx

Here is the call graph for this function:

Here is the caller graph for this function:

KOKKOS_INLINE_FUNCTION void use_nearest_triangle	(	const Plane< DeviceType > &	plane,
		const View< int **, CLayout, DeviceType > &	tr_node,
		const SimdVector2D &	x,
		bool	weight_p_on_boundary,
		int	i_simd,
		Simd< int > &	itr,
		SimdGridVec &	p_loc
	)

Find the nearest triangle and use it. This algorithm loops over ALL nodes - very inefficient! Should use wall nodes - or last surface if no wall is present

Parameters

[in]	plane	the plane
[in]	tr_node	the triangle-to-node mapping of the plane
[in]	x	the coordinates on the plane
[in]	weight_p_on_boundary	whether to weight the nearest node if the projection falls outside the grid
[in]	i_simd	the simd index since this is a vectorized operation
[out]	itr	the index of the nearest triangle
[out]	p_loc	the barycentric weights on the nearest triangle

Here is the call graph for this function:

Here is the caller graph for this function:

static void write_ff ( const Plane< HostType > &  plane_start, const Projection< HostType > &  projection, const std::string  stream_name, const std::string  filename, bool  reverse  )

static

Here is the call graph for this function:

Here is the caller graph for this function: