gradparx2.cpp File Reference

#include "globals.hpp"
#include "my_subview.hpp"
#include "gradparx2.hpp"

Include dependency graph for gradparx2.cpp:

Functions
void	neg_gradpar_single_plane (const View< int *, CLayout, DeviceType > &is_inside, const Projection< DeviceType > &projection, const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &input_left, const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &input_center, const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &input_right, const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &output, double sgn, Order derivative_order)
void	negative_gradparx2 (const GradParXTmp &gpxt, const Kokkos::View< double **, Kokkos::LayoutRight, DeviceType > &input, const Kokkos::View< double **, Kokkos::LayoutRight, DeviceType > &output, Order derivative_order)

Function Documentation

void neg_gradpar_single_plane	(	const View< int *, CLayout, DeviceType > &	is_inside,
		const Projection< DeviceType > &	projection,
		const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &	input_left,
		const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &	input_center,
		const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &	input_right,
		const Kokkos::View< double *, Kokkos::LayoutRight, DeviceType > &	output,
		double	sgn,
		Order	derivative_order
	)

Calculates the gradient of the input array. Includes a precomputation to check which nodes are inside the boundary provided. This could be done once in grid setup and stored. This code is called from Fortran and is self-contained. Thus, everything needed is passed from Fortran to C++ even though grid information is already in the C++ grid object. It will eventually be refactored accordingly.

Parameters

[in]	nseg	is the number of boundary segments(?)
[in]	sml_bt_sign	is the user-input direction of the toroidal magnetic field
[in]	tr_fort	is an array of the triangles the corresponding to the field-following path through the vertex
[in]	p_fort	is an array of the weights of the three triangle vertices for tr_fort triangles
[in]	dx_fort	is an array of distances
[in]	bd_fort	is an array of the boundary segments
[in]	input	is the potential
[out]	output	is the gradient of the potential
[in]	derivative_order	is which derivative to take

Returns

void

Here is the call graph for this function:

Here is the caller graph for this function:

void negative_gradparx2	(	const GradParXTmp &	gpxt,
		const Kokkos::View< double **, Kokkos::LayoutRight, DeviceType > &	input,
		const Kokkos::View< double **, Kokkos::LayoutRight, DeviceType > &	output,
		Order	derivative_order
	)

Calculates the gradient of the input array. Includes a precomputation to check which nodes are inside the boundary provided. This could be done once in grid setup and stored. This code is called from Fortran and is self-contained. Thus, everything needed is passed from Fortran to C++ even though grid information is already in the C++ grid object. It will eventually be refactored accordingly.

Parameters

[in]	nseg	is the number of boundary segments(?)
[in]	sml_bt_sign	is the user-input direction of the toroidal magnetic field
[in]	tr_fort	is an array of the triangles the corresponding to the field-following path through the vertex
[in]	p_fort	is an array of the weights of the three triangle vertices for tr_fort triangles
[in]	dx_fort	is an array of distances
[in]	bd_fort	is an array of the boundary segments
[in]	input	is the potential
[out]	output	is the gradient of the potential
[in]	derivative_order	is which derivative to take

Returns

void

Here is the call graph for this function:

Here is the caller graph for this function: