dolfinx Namespace Reference

Top-level namespace. More...

Namespaces
namespace	common
	Miscellaneous classes, functions and types.
namespace	fem
	Finite element method functionality.
namespace	geometry
	Geometry data structures and algorithms.
namespace	graph
	Graph data structures and algorithms.
namespace	io
	Support for file IO.
namespace	la
	Linear algebra interface.
namespace	mesh
	Mesh data structures and algorithms on meshes.
namespace	MPI
	MPI support functionality.
namespace	nls
	Nonlinear solvers.
namespace	refinement
	Mesh refinement algorithms.

Classes
class	Table

Concepts
concept	scalar

Enumerations
enum class	TimingType : int { wall = 0 , user = 1 , system = 2 }

Functions
std::string	version ()
	Return DOLFINx version string.
std::string	ufcx_signature ()
	Return UFC signature string.
std::string	git_commit_hash ()
bool	has_debug ()
bool	has_petsc ()
	Return true if DOLFINx is compiled with PETSc.
bool	has_slepc ()
	Return true if DOLFINx is compiled with SLEPc.
bool	has_scotch ()
	Return true if DOLFINx is compiled with Scotch.
bool	has_parmetis ()
	Return true if DOLFINx is compiled with ParMETIS.
bool	has_kahip ()
	Return true if DOLFINx is compiled with KaHIP.
bool	has_adios2 ()
	Return true if DOLFINX is compiled with ADIOS2.
void	init_logging (int argc, char *argv[])
	Optional initialisation of the logging backend.
template<typename T , int BITS = 8>
void	radix_sort (std::span< T > array)
template<typename T , int BITS = 16>
void	argsort_radix (std::span< const T > array, std::span< std::int32_t > perm)
template<typename T , int BITS = 16>
std::vector< std::int32_t >	sort_by_perm (std::span< const T > x, std::size_t shape1)
	Compute the permutation array that sorts a 2D array by row.
Table	timings (std::set< TimingType > type)
void	list_timings (MPI_Comm comm, std::set< TimingType > type, Table::Reduction reduction=Table::Reduction::max)
std::tuple< std::size_t, double, double, double >	timing (std::string task)

Detailed Description

Top-level namespace.

Enumeration Type Documentation

TimingType

enum class TimingType : int

strong

Timing types:

• TimingType::wall wall-clock time
• TimingType::user user (cpu) time
• TimingType::system system (kernel) time

Function Documentation

argsort_radix()

template<typename T , int BITS = 16>

void argsort_radix	(	std::span< const T >	array,
		std::span< std::int32_t >	perm )

Returns the indices that would sort (lexicographic) a vector of bitsets.

Template Parameters

T	The size of the bitset, which corresponds to the number of bits necessary to represent a set of integers. For example, N = 96 for mapping three std::int32_t.
BITS	The number of bits to sort at a time

Parameters

[in]	array	The array to sort
[in]	perm	FIXME

git_commit_hash()

std::string git_commit_hash

(

)

Return git changeset hash (returns "unknown" if changeset is not known)

has_debug()

bool has_debug

(

)

Return true if DOLFINx is compiled in debugging mode, i.e., with assertions on

init_logging()

void init_logging	(	int	argc,
		char *	argv[] )

Optional initialisation of the logging backend.

The log verbosity can be controlled from the command line using -dolfinx_loglevel <level>, where <level> is an integer. Increasing values increase verbosity.

The full loguru API can be used in applications to control the log system. See https://emilk.github.io/loguru/ for the loguru documentation.

Note

The logging backend is loguru (https://github.com/emilk/loguru).

Parameters

[in]	argc	Number of command line arguments.
[in]	argv	Command line argument vector.

list_timings()

void list_timings	(	MPI_Comm	comm,
		std::set< TimingType >	type,
		Table::Reduction	reduction = Table::Reduction::max )

List a summary of timings and tasks. MPI_AVG reduction is printed.

Parameters

[in]	comm	MPI Communicator
[in]	type	Subset of { TimingType::wall, TimingType::user, TimingType::system }
[in]	reduction	MPI Reduction to apply (min, max or average)

radix_sort()

template<typename T , int BITS = 8>

void radix_sort

(

std::span< T >

array

)

Sort a vector of integers with radix sorting algorithm. The bucket size is determined by the number of bits to sort at a time (2^BITS).

Template Parameters

T	Integral type
BITS	The number of bits to sort at a time.

Parameters

[in,out]

array

The array to sort.

sort_by_perm()

template<typename T , int BITS = 16>

std::vector< std::int32_t > sort_by_perm	(	std::span< const T >	x,
		std::size_t	shape1 )

Compute the permutation array that sorts a 2D array by row.

Parameters

[in]	x	The flattened 2D array to compute the permutation array for.
[in]	shape1	The number of columns of x.

Returns

The permutation array such that x[perm[i]] <= x[perm[i +1]]. @prex.size()must be a multiple ofshape1. @note This function is suitable for small values ofshape1. Each column ofx` is copied into an array that is then sorted.

timing()

std::tuple< std::size_t, double, double, double > timing

(

std::string

task

)

Return timing (count, total wall time, total user time, total system time) for given task.

Parameters

[in]

task

Name of a task

Returns

The (count, total wall time, total user time, total system time) for the task

timings()

Table timings

(

std::set< TimingType >

type

)

Return a summary of timings and tasks in a Table

Parameters

[in]

type

subset of { TimingType::wall, TimingType::user, TimingType::system }

Returns

Table with timings