utils.h File Reference

#include "output.h"

Include dependency graph for utils.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Data Structures
struct	timing
struct	norm
struct	stats
	The statsf() function returns the minimum, maximum, volume sum, standard deviation and volume for field f. More...

Functions
void	update_perf ()
	Performance statistics are gathered by this function, which is typically called by the run() loop.
timing	timer_timing (timer t, int i, size_t tnc, double *mpi)
	Given a timer, iteration count i, total number of cells tnc and array of MPI timings mpi (with a size equal to the number of processes), this function returns the statistics above.
void	timer_print (timer t, int i, size_t tnc)
	This function writes timing statistics on standard output.
norm	normf (scalar f)
stats	statsf (scalar f)
static double	generic_limiter (double r, double beta)
double	minmod (double s0, double s1, double s2)
double	superbee (double s0, double s1, double s2)
double	sweby (double s0, double s1, double s2)
double	minmod2 (double s0, double s1, double s2)
void	gradients (scalar *f, vector *g)
	Given a list of scalar fields f, this function fills the gradient fields g with the corresponding gradients.
void	vorticity (const vector u, scalar omega)
double	change (scalar s, scalar sn)
	Given two scalar fields s and sn this function returns the maximum of their absolute difference.
scalar	lookup_field (const char *name)
	These functions return the scalar/vector fields called name, or -1 if they don't exist.
vector	lookup_vector (const char *name)
macro	foreach_segment (coord S[2], coord p[2], Reduce reductions=None)
	The macro below traverses the set of sub-segments intersecting the mesh and spanning the [A:B] segment.
void	fields_stats (scalar *list=all)
	This function returns a summary of the currently-defined fields.

Variables
double	DT = HUGE [0,1]
double	CFL = 0.5 [0]
struct {
long   nc
long   tnc
double   t
double   tp
double   speed
double   ispeed
timer   gt
}	perf = {0}
	Performance statistics are stored in this structure.
double	theta = 1.3 [0]
	This is the generalised minmod limiter.

Function Documentation

change()

double change	(	scalar	s,
		scalar	sn
	)

Given two scalar fields s and sn this function returns the maximum of their absolute difference.

Definition at line 303 of file utils.h.

References dv, max, s, and x.

fields_stats()

void fields_stats

(

scalar *

list = all

)

This function returns a summary of the currently-defined fields.

Definition at line 382 of file utils.h.

References s, statsf(), t, and x.

Referenced by main(), and process_line().

Here is the call graph for this function:

Here is the caller graph for this function:

foreach_segment()

macro foreach_segment	(	coord	S[2],
		coord	p[2],
		Reduce	reductions = None
	)

The macro below traverses the set of sub-segments intersecting the mesh and spanning the [A:B] segment.

The pair of coordinates defining the sub-segment contained in each cell are defined by p[0] and p[1].

Definition at line 349 of file utils.h.

References _i, a, alpha, clamp(), dimension, p, reductions, S, sq(), t, vector::x, x, coord::x, vector::y, and y.

Referenced by segment_flux().

Here is the call graph for this function:

Here is the caller graph for this function:

generic_limiter()

static double generic_limiter ( double  r, double  beta  )

static

Slope limiters

Given three values, these slope limiters return the corresponding slope-limited gradient.

Definition at line 197 of file utils.h.

References beta, max, min, and x.

Referenced by minmod(), superbee(), and sweby().

Here is the caller graph for this function:

gradients()

void gradients	(	scalar *	f,
		vector *	g
	)

Given a list of scalar fields f, this function fills the gradient fields g with the corresponding gradients.

If the gradient attribute of a field is set (typically to one of the limiting functions above), it is used to compute the gradient, otherwise simple centered differencing is used.

Definition at line 247 of file utils.h.

References _i, assert, dimension, f, fs, g, list_len(), s, v, vectors_len(), vector::x, and x.

Referenced by tracer_fluxes(), update_conservation(), and update_saint_venant().

Here is the call graph for this function:

Here is the caller graph for this function:

lookup_field()

scalar lookup_field

(

const char *

name

)

These functions return the scalar/vector fields called name, or -1 if they don't exist.

Definition at line 321 of file utils.h.

References s, and x.

Referenced by assemble_node(), and slave_interpolate().

Here is the caller graph for this function:

lookup_vector()

vector lookup_vector

(

const char *

name

)

Definition at line 330 of file utils.h.

References s, and x.

minmod()

double minmod	(	double	s0,
		double	s1,
		double	s2
	)

Definition at line 204 of file utils.h.

References generic_limiter(), and x.

Here is the call graph for this function:

minmod2()

double minmod2	(	double	s0,
		double	s1,
		double	s2
	)

Definition at line 225 of file utils.h.

References max, min, theta, and x.

Referenced by event_defaults().

Here is the caller graph for this function:

normf()

norm normf

(

scalar

f

)

Definition at line 142 of file utils.h.

References norm::avg, dv, f, max, n, nodata, sq(), v, and x.

Here is the call graph for this function:

statsf()

stats statsf

(

scalar

f

)

Definition at line 170 of file utils.h.

References dv, f, max, min, nodata, s, sq(), sum, and x.

Referenced by event_init(), fields_stats(), RPE(), and zarea().

Here is the call graph for this function:

Here is the caller graph for this function:

superbee()

double superbee	(	double	s0,
		double	s1,
		double	s2
	)

Definition at line 208 of file utils.h.

References generic_limiter(), and x.

Here is the call graph for this function:

sweby()

double sweby	(	double	s0,
		double	s1,
		double	s2
	)

Definition at line 212 of file utils.h.

References generic_limiter(), and x.

Here is the call graph for this function:

timer_print()

void timer_print	(	timer	t,
		int	i,
		size_t	tnc
	)

This function writes timing statistics on standard output.

Definition at line 110 of file utils.h.

References datasize, fflush(), glFinish, GRIDNAME, i, npe, s, t, timer_timing(), tnc, and x.

Referenced by run(), and run().

Here is the call graph for this function:

Here is the caller graph for this function:

timer_timing()

timing timer_timing	(	timer	t,
		int	i,
		size_t	tnc,
		double *	mpi
	)

Given a timer, iteration count i, total number of cells tnc and array of MPI timings mpi (with a size equal to the number of processes), this function returns the statistics above.

Definition at line 64 of file utils.h.

References i, mpi, n, npe, s, t, timer_elapsed(), tnc, and x.

Referenced by timer_print().

Here is the call graph for this function:

Here is the caller graph for this function:

update_perf()

void update_perf

(

)

Performance statistics are gathered by this function, which is typically called by the run() loop.

Definition at line 34 of file utils.h.

References grid, Grid::n, perf, timer_elapsed(), and Grid::tn.

Referenced by run(), and run().

Here is the call graph for this function:

Here is the caller graph for this function:

vorticity()

void vorticity	(	const vector	u,
		scalar	omega
	)

Other functions

Given a velocity field \(\mathbf{u}\), this function fills a scalar field \(\omega\) with the vorticity field

\[ \omega = \partial_x u_y - \partial_y u_x \]

To properly take the metric into account, \(\omega\) is computed as an average over the (surface) element, which is easily obtained as the circulation of \(\mathbf{u}\) along the boundary of the element. Using the definitions of the metric factors fm and cm, this gives the expression below.

Definition at line 290 of file utils.h.

References _i, cm, fm, omega, SEPS, u, vector::x, x, and vector::y.

Variable Documentation

CFL

double CFL = 0.5 [0]

Definition at line 10 of file utils.h.

Referenced by advect(), event_defaults(), event_face_fields(), event_stability(), hllc(), kinetic(), kurganov(), riemann(), timestep(), and timestep().

DT

double DT = HUGE [0,1]

Various utility functions

Default parameters and variables

The default maximum timestep and CFL number.

Definition at line 10 of file utils.h.

Referenced by event_init(), event_velocity(), run(), and timestep().

gt

timer gt

Definition at line 27 of file utils.h.

ispeed

double ispeed

Definition at line 25 of file utils.h.

nc

long nc

Definition at line 17 of file utils.h.

Referenced by args(), embed_fraction_refine(), fine(), for(), and fraction_refine().

[struct]

struct { ... } perf

Performance statistics are stored in this structure.

Referenced by event_perfs(), event_runtime(), run(), run(), and update_perf().

speed

double speed

Definition at line 23 of file utils.h.

t

double t

Definition at line 21 of file utils.h.

Referenced by fields_stats(), foreach_segment(), timer_print(), and timer_timing().

theta

double theta = 1.3 [0]

This is the generalised minmod limiter.

The \(\theta\) global variable can be used to tune the limiting ( \(\theta=1\) gives minmod, the most dissipative limiter and \(\theta=2\) gives superbee, the least dissipative).

Definition at line 223 of file utils.h.

Referenced by eigenvalues(), event_defaults(), event_tracer_diffusion(), if(), minmod2(), and VARIABLES().

tnc

long tnc

Definition at line 19 of file utils.h.

Referenced by timer_print(), and timer_timing().

tp

double tp

Definition at line 21 of file utils.h.

Referenced by poisson_thermal(), relax_thermal(), and residual_thermal().