bvls_module Module

Module for bvls.

History

Original code from http://www.netlib.org/lawson-hanson/all
Jacob Williams, Nov 2018 : refactored into a module. some code cleanup.

Uses

- slsqp_kinds
- slsqp_support
Help

Graph Key

Nodes of different colours represent the following:

Solid arrows point from a submodule to the (sub)module which it is descended from. Dashed arrows point from a module or program unit to modules which it uses.

Used by

Help

Graph Key

Nodes of different colours represent the following:

Solid arrows point from a submodule to the (sub)module which it is descended from. Dashed arrows point from a module or program unit to modules which it uses.

Subroutines

public subroutine bvls_wrapper(a, mda, m, n, b, x, rnorm, w, zz, mode, max_iter)

Call bvls, but matching the interface of the old nnls.

Arguments

Type	Intent	Optional	Attributes		Name
real(kind=wp),	intent(inout),		dimension(mda,n)	::	a
integer,	intent(in)			::	mda
integer,	intent(in)			::	m
integer,	intent(in)			::	n
real(kind=wp),	intent(inout),		dimension(m)	::	b
real(kind=wp),	intent(out),		dimension(n)	::	x
real(kind=wp),	intent(out)			::	rnorm
real(kind=wp),	intent(inout),		dimension(n)	::	w
real(kind=wp),	intent(inout),		dimension(m)	::	zz
integer,	intent(out)			::	mode
integer,	intent(in)			::	max_iter	maximum number of iterations (if <=0, then `3*n` is used)

public subroutine bvls(a, b, bnd, x, rnorm, nsetp, w, index, ierr, max_iter)

Given an m by n matrix, $\mathbf{A}$ , and an m-vector, $\mathbf{b}$ , compute an n-vector, $\mathbf{x}$ , that solves the least squares problem:

Arguments

Type	Intent	Optional	Attributes		Name
real(kind=wp),	intent(inout),		dimension(:,:)	::	a	On entry A() contains the M by N matrix, A. On return A() contains the product matrix, Q*A, where Q is an M by M orthogonal matrix generated by this subroutine. The dimensions are M=size(A,1) and N=size(A,2).
real(kind=wp),	intent(inout),		dimension(:)	::	b	On entry B() contains the M-vector, B. On return, B() contains Q*B. The same Q multiplies A.
real(kind=wp),	intent(in),		dimension(:,:)	::	bnd	BND(1,J) is the lower bound for X(J). BND(2,J) is the upper bound for X(J). Read more…
real(kind=wp),	intent(out),		dimension(:)	::	x	On entry X() need not be initialized. On return, X() will contain the solution N-vector.
real(kind=wp),	intent(out)			::	rnorm	Euclidean norm of the residual vector, b - A*X.
integer,	intent(out)			::	nsetp	Indicates the number of components of the solution vector, X(), that are not at their constraint values.
real(kind=wp),	intent(out),		dimension(:)	::	w	An N-array. On return, W() will contain the dual solution vector. Using Set definitions below: Read more…
integer,	intent(out),		dimension(:)	::	index	An INTEGER working array of size N. On exit the contents of this array define the sets P, Z, and F as follows: Read more…
integer,	intent(out)			::	ierr	Indicates status on return: Read more…
integer,	intent(in)			::	max_iter	maximum number of iterations (if <=0, then `3*n` is used)