matrix_utils.f90 File Reference

Lower and upper triangular decomposition. More...

Go to the source code of this file.

Functions/Subroutines
subroutine	lubksb (a, n, np, indx, b)
	Lower and upper triangular back substitution.
subroutine	ludcmp (a, n, np, indx, d)
	This subprogram decomposes a matrix into a product of lower and upper triangular matrices.

Detailed Description

Lower and upper triangular decomposition.

Author

Mark Iredell

Date

2008-08-01

Definition in file matrix_utils.f90.

Function/Subroutine Documentation

lubksb()

subroutine lubksb	(	real, dimension(np,np), intent(in)	a,
		integer, intent(in)	n,
		integer, intent(in)	np,
		integer, dimension(n), intent(in)	indx,
		real, dimension(n), intent(inout)	b )

Lower and upper triangular back substitution.

Author

Iredell

Date

2008-08-01

This subprogram back substitutes to solve decomposed lower and upper triangular matrices as outputted by ludcmp.

Parameters

[in]	a	real(np,np) LU-decomposed matrix (from ludcmp)
[in]	n	integer order of matrix
[in]	np	integer dimension of matrix
[in]	indx	integer(n) pivot indices (from ludcmp)
[in,out]	b	- input real(n) rhs vector of linear problem (will be overwritten) output real(n) solution of linear problem (original A times output B equals original B)

Author

Mark Iredell

Date

2008-08-01

Definition at line 100 of file matrix_utils.f90.

Referenced by vcoord_gen().

ludcmp()

subroutine ludcmp	(	real, dimension(np,np), intent(inout)	a,
		integer, intent(in)	n,
		integer, intent(in)	np,
		integer, dimension(n), intent(out)	indx,
		real, intent(out)	d )

This subprogram decomposes a matrix into a product of lower and upper triangular matrices.

Parameters

[in,out]	a	- input: a real(np,np) matrix (will be overwritten) output: output real(np,np) LU-decomposed matrix (U is upper part of A, including diagonal; L is lower part of A, with 1 as diagonal; L*U equals original A after permuting)
[in]	n	integer order of matrix
[in]	np	integer dimension of matrix
[out]	indx	integer(n) pivot indices (original A rows are permuted in order i with indx(i))
[out]	d	real determinant permutation (1 or -1, or 0 if singular) (determinant is output diagonal product times d)

Author

Mark Iredell

Date

2008-08-01

Definition at line 20 of file matrix_utils.f90.

Referenced by vcoord_gen().