function [ res, jac ] = opt08_rj ( x, flag ) %% OPT08_RJ evaluates RES and JAC for test case #8. % % Discussion: % % This example is known as the extended Powell singular function. % % The problem size is arbitrary, except that it should be a multiple of 4. % % The optimizing value is % % X* = (0,0,0,0,...) % % Modified: % % 06 January 2008 % % Author: % % Jeff Borggaard, % Gene Cliff, % Virginia Tech. % % Reference: % % John Dennis, Robert Schnabel, % Numerical Methods for Unconstrained Optimization % and Nonlinear Equations, % SIAM, 1996, % ISBN13: 978-0-898713-64-0, % LC: QA402.5.D44. % % Parameters: % % Input, real X(N), the evaluation point. % % Input, string FLAG, indicates what must be computed. % 'f' means only the value of RES is needed, % 'g' means only the value of JAC is needed, % 'all' means RES and JAC are needed. % It is acceptable to behave as though FLAG was 'all' % on every call. % % Output, real RES(N,1), the function column vector. % % Output, real JAC(N,N), the Jacobian matrix. % n = length ( x ); if ( mod ( n, 4 ) ~= 0) fprintf ( '\n' ); fprintf ( 'OPT08_RJ - Fatal error!\n' ); fprintf ( ' The input vector X should have length divisible by 4.\n'), fprintf ( ' Instead, it has length = %d.\n', n ); keyboard end res = zeros(n,1); for i=1:n/4 res(4*i-3) = x(4*i-3) + 10*x(4*i-2); res(4*i-2) = sqrt(5)*( x(4*i-1)-x(4*i) ); res(4*i-1) = ( x(4*i-2) - 2 * x(4*i-1) )^2; res(4*i ) = sqrt(10)*( x(4*i-3)-x(4*i) )^2; end jac = zeros(n,n); for i=1:n/4 jac(4*i-3,4*i-3) = 1; jac(4*i-3,4*i-2) = 10; jac(4*i-3,4*i-1) = 0; jac(4*i-3,4*i ) = 0; jac(4*i-2,4*i-3) = 0; jac(4*i-2,4*i-2) = 0; jac(4*i-2,4*i-1) = sqrt ( 5 ); jac(4*i-2,4*i ) = - sqrt ( 5 ); jac(4*i-1,4*i-3) = 0; jac(4*i-1,4*i-2) = 2 * ( x(4*i-2) - 2 * x(4*i-1) ); jac(4*i-1,4*i-1) = - 4 * ( x(4*i-2) - 2 * x(4*i-1) ); jac(4*i-1,4*i ) = 0; jac(4*i ,4*i-3) = 2 * sqrt(10) * ( x(4*i-3) - x(4*i) ); jac(4*i ,4*i-2) = 0; jac(4*i ,4*i-1) = 0; jac(4*i ,4*i ) = - 2 * sqrt(10) * ( x(4*i-3) - x(4*i) ); end