1function x = cs_lusol (A,b,order,tol) %#ok 2%CS_LUSOL solve Ax=b using LU factorization. 3% x = cs_lusol(A,b) computes x = A\b, where A is sparse and square, and b is a 4% full vector. The ordering cs_amd(A,2) is used. 5% 6% x = cs_lusol(A,b,1) also computes x = A\b, but uses the cs_amd(A) ordering 7% with diagonal preference (tol=0.001). 8% 9% x = cs_lusol(A,b,order,tol) allows both the ordering and tolerance to be 10% defined. The ordering defaults to 1, and tol defaults to 1. 11% ordering: 0: natural, 1: amd(A+A'), 2: amd(S'*S) where S=A except with no 12% dense rows, 3: amd(A'*A). 13% 14% Example: 15% Prob = ssget ('HB/arc130') ; A = Prob.A ; n = size(A,1) ; 16% b = rand (n,1) ; x = cs_lusol (A,b) ; norm (A*x-b) 17% 18% See also CS_LU, CS_AMD, CS_CHOLSOL, CS_QRSOL, MLDIVIDE. 19 20% Copyright 2006-2012, Timothy A. Davis, http://www.suitesparse.com 21 22error ('cs_lusol mexFunction not found') ; 23