www.pudn.com > lrr.rar > sr2.m, change:2014-12-16,size:1589b

```function [Z] = sr2(X,lambda)
%This routine solves the following l1-norm
% optimization problem with l21-error
% min |Z|_1+lambda*|E|_2,1
% s.t., X = XZ+E
%       Zii = 0
% inputs:
%        X -- D*N data matrix, D is the data dimension, and N is the number
%             of data vectors.
if nargin<2
lambda = 1;
end
tol = 1e-8;
maxIter = 1e6;
[d n] = size(X);
rho = 1.1;
max_mu = 1e30;
mu = 1e-6;
xtx = X'*X;
inv_x = inv(xtx+eye(n));
%% Initializing optimization variables
% intialize
J = zeros(n);
E = sparse(d,n);
Z = J;

Y1 = zeros(d,n);
Y2 = zeros(n);
%% Start main loop
iter = 0;
while iter<maxIter
iter = iter + 1;

temp = Z + Y2/mu;
J = max(0,temp - 1/mu)+min(0,temp + 1/mu);
J = J - diag(diag(J)); %Jii = 0

Z = inv_x*(xtx-X'*E+J+(X'*Y1-Y2)/mu);

xmaz = X-X*Z;
temp = X-X*Z+Y1/mu;
%E = max(0,temp - lambda/mu)+min(0,temp + lambda/mu);
E = solve_l1l2(temp,lambda/mu);

leq1 = xmaz-E;
leq2 = Z-J;
stopC = max(max(max(abs(leq1))),max(max(abs(leq2))));
if iter==1 || mod(iter,50)==0 || stopC<tol
disp(['iter ' num2str(iter) ',mu=' num2str(mu,'%2.1e') ',stopALM=' num2str(stopC,'%2.3e')]);
end
if stopC<tol
break;
else
Y1 = Y1 + mu*leq1;
Y2 = Y2 + mu*leq2;
mu = min(max_mu,mu*rho);
end
end

function [E] = solve_l1l2(W,lambda)
n = size(W,2);
E = W;
for i=1:n
E(:,i) = solve_l2(W(:,i),lambda);
end

function [x] = solve_l2(w,lambda)
% min lambda |x|_2 + |x-w|_2^2
nw = norm(w);
if nw>lambda
x = (nw-lambda)*w/nw;
else
x = zeros(length(w),1);
end```