sparsematrixmodule Module Reference

A module for handling locally stored CSR matrices. More...

Data Types
type	sparsematrix_t
	A datatype for storing a CSR matrix. More...

Functions/Subroutines
pure subroutine, public	constructemptysparsematrix (this, columns, rows)
	Create a sparse matrix with a certain number of columns and rows. Will allocate storage for the outer values. More...
subroutine, public	constructsparsematrixfromfile (this, file_name)
	Create a sparse matrix by reading in a matrix market file. More...
pure subroutine, public	constructfromtripletlist (this, triplet_list, rows, columns)
	Construct a sparse matrix from a SORTED triplet list. The triplet list must be sorted to efficiently fill in the matrix. This constructor assumes you have already sorted the triplet list. More...
pure subroutine, public	destructsparsematrix (this)
	Explicitly destruct a sparse matrix. This will always check if arrays are actually allocated, so you can feel free to destruct a matrix even if it has no data. More...
pure subroutine, public	copysparsematrix (matA, matB)
	Copy a sparse matrix in a safe way. More...
pure integer function, public	getrows (this)
	Get the number of rows of a matrix. More...
pure integer function, public	getcolumns (this)
	Get the number of columns of a matrix. More...
pure subroutine, public	transposesparsematrix (this, matT)
	Transpose a sparse matrix and return it in a separate matrix. The current implementation has you go from matrix to triplet list, triplet list to transposed triplet list. The triplet list must then be sorted and then the return matrix is constructed. More...
pure subroutine, public	composesparsematrixcolumns (mat_list, out_matrix)
	Create a big Matrix C = [Matrix 1 | Matrix 1, ...] where the columns of the first matrix are followed by the columns of the matrices in the list. More...
pure subroutine, public	splitsparsematrixcolumns (this, num_blocks, split_list, block_offsets_out)
	Take a matrix, and split into into small blocks. More...
pure subroutine, public	matrixtotripletlist (this, triplet_list)
	Construct a triplet list from a matrix. More...
subroutine, public	printsparsematrix (this, file_name_in)
	Print out a sparse matrix. We first create a triplet list, and then call the print triplet list function. More...
pure logical function, public	checkifidentity (this, threshold_in)
	Check if a matrix is equal to the identity matrix. This routine is really just for testing. You can multiply a matrix and its inverse, and then call this routine to make sure multiplication is correct. More...

Detailed Description

A module for handling locally stored CSR matrices.

Function/Subroutine Documentation

checkifidentity()

pure logical function, public sparsematrixmodule::checkifidentity	(	type(sparsematrix_t), intent(in)	this,
		real(ntreal), intent(in), optional	threshold_in
	)

Check if a matrix is equal to the identity matrix. This routine is really just for testing. You can multiply a matrix and its inverse, and then call this routine to make sure multiplication is correct.

Parameters

[in]	this	the matrix to check.
[in]	threshold_in	for flushing values to zero. Default value is 10^-8.

Returns

true if the matrix is equal to the identity matrix.

composesparsematrixcolumns()

pure subroutine, public sparsematrixmodule::composesparsematrixcolumns	(	type(sparsematrix_t), dimension(:), intent(in)	mat_list,
		type(sparsematrix_t), intent(inout)	out_matrix
	)

Create a big Matrix C = [Matrix 1 | Matrix 1, ...] where the columns of the first matrix are followed by the columns of the matrices in the list.

Parameters

[in]	mat_list	list of matrices to compose.
[out]	out_matrix	= [Matrix 1 | Matrix 2, ...] .

constructemptysparsematrix()

pure subroutine, public sparsematrixmodule::constructemptysparsematrix	(	type(sparsematrix_t), intent(out)	this,
		integer, intent(in)	columns,
		integer, intent(in)	rows
	)

Create a sparse matrix with a certain number of columns and rows. Will allocate storage for the outer values.

Parameters

[out]	this	the matrix being created. It will have the outer index allocated, but nothing else.
[in]	columns	number of matrix columns.
[in]	rows	number of matrix rows.

constructfromtripletlist()

pure subroutine, public sparsematrixmodule::constructfromtripletlist	(	type(sparsematrix_t), intent(out)	this,
		type(tripletlist_t), intent(in)	triplet_list,
		integer, intent(in)	rows,
		integer, intent(in)	columns
	)

Construct a sparse matrix from a SORTED triplet list. The triplet list must be sorted to efficiently fill in the matrix. This constructor assumes you have already sorted the triplet list.

Parameters

[out]	this	the matrix being constructed
[in]	triplet_list	a list of triplet values. They must be sorted.
[in]	rows	number of matrix rows
[in]	columns	number of matrix columns

constructsparsematrixfromfile()

subroutine, public sparsematrixmodule::constructsparsematrixfromfile	(	type(sparsematrix_t), intent(out)	this,
		character(len=*), intent(in)	file_name
	)

Create a sparse matrix by reading in a matrix market file.

Parameters

[out]	this	the matrix being constructed.
[in]	file_name	name of the file.

copysparsematrix()

pure subroutine, public sparsematrixmodule::copysparsematrix	(	type(sparsematrix_t), intent(in)	matA,
		type(sparsematrix_t), intent(inout)	matB
	)

Copy a sparse matrix in a safe way.

Parameters

[in]	matA	matrix to copy
[in,out]	matB	= matA

destructsparsematrix()

pure subroutine, public sparsematrixmodule::destructsparsematrix

(

type(sparsematrix_t), intent(inout)

this

)

Explicitly destruct a sparse matrix. This will always check if arrays are actually allocated, so you can feel free to destruct a matrix even if it has no data.

Parameters

[in,out]

this

the matrix to free up

getcolumns()

pure integer function, public sparsematrixmodule::getcolumns

(

type(sparsematrix_t), intent(in)

this

)

Get the number of columns of a matrix.

Parameters

[in]

this

the matrix.

Returns

number of columns.

getrows()

pure integer function, public sparsematrixmodule::getrows

(

type(sparsematrix_t), intent(in)

this

)

Get the number of rows of a matrix.

Parameters

[in]

this

the matrix.

Returns

number of rows.

matrixtotripletlist()

pure subroutine, public sparsematrixmodule::matrixtotripletlist	(	type(sparsematrix_t), intent(in)	this,
		type(tripletlist_t), intent(inout)	triplet_list
	)

Construct a triplet list from a matrix.

Parameters

[in]	this	the matrix to construct the triplet list from.
[out]	triplet_list	the triplet list we created.

printsparsematrix()

subroutine, public sparsematrixmodule::printsparsematrix	(	type(sparsematrix_t), intent(in)	this,
		character(len=*), intent(in), optional	file_name_in
	)

Print out a sparse matrix. We first create a triplet list, and then call the print triplet list function.

Parameters

[in]	this	the matrix to be printed.
[in]	file_name_in	optionally, you can pass a file to print to.

splitsparsematrixcolumns()

pure subroutine, public sparsematrixmodule::splitsparsematrixcolumns	(	type(sparsematrix_t), intent(in)	this,
		integer, intent(in)	num_blocks,
		type(sparsematrix_t), dimension(num_blocks), intent(out)	split_list,
		integer, dimension(num_blocks+1), intent(out), optional	block_offsets_out
	)

Take a matrix, and split into into small blocks.

Parameters

[in]	this	matrix to perform this operation on.
[in]	num_blocks	number of blocks to split into.
[out]	split_list	1D array of blocks.
[out]	block_offsets_out	the offsets used for splitting.

transposesparsematrix()

pure subroutine, public sparsematrixmodule::transposesparsematrix	(	type(sparsematrix_t), intent(in)	this,
		type(sparsematrix_t), intent(inout)	matT
	)

Transpose a sparse matrix and return it in a separate matrix. The current implementation has you go from matrix to triplet list, triplet list to transposed triplet list. The triplet list must then be sorted and then the return matrix is constructed.

Parameters

[in]	this	the matrix to be transposed.
[out]	matT	the input matrix transposed.