function [x, resvec, lsvec] = dense_overdetermined_lsqr(A, b, R, tol, maxit)
% x = dense_overdetermined_lsqr(A, b, R, tol, maxit)
% [x, resvec, lsvec] = dense_overdetermined_lsqr(A, b, R, tol, maxit)
%
% LSQR on a dense overdetermined matrix with a dense preconditioner R.
% Solves x = arg min (A*x - b)
% Inputs:
% A, b
% R - Upper triangular preconditioner. kappa(A * inv(R)) governs
% the convergence.
% tol - tolerance. Stop when
% norm(inv(R') * A' * r) / (norm(A * inv(R), 'fro') * norm(r)) < tol
% maxit - Maximum number of iterations
%
% Outputs:
% x
% optional: resvec, lsvec - values of norm(A' * r) and norm(r).
%
% 6-December 2009, Version 1.3
% Copyright (C) 2009, Haim Avron and Sivan Toledo.
