v2Gv.m

function [Gv] = v2Gv(v,g0,g1,g2,h1,W1,W2,indices,regularization)
[subsetSize,backup] = size(v);
if (backup > subsetSize)
    subsetSize = backup;
end
v = decompress(v,indices);

[rows1,cols1] = size(W1);
[rows2,cols2] = size(W2);

W1_1 = zeros(rows1,cols1);
W2_1 = zeros(rows2,cols2);

% matricize
for i = 1:rows1
    W1_1(i,:) = v((i-1)*cols1 + 1:i*cols1);    
end
count = rows1*cols1;

for i = 1:rows2
    W2_1(i,:) = v(count + (i-1)*cols2 + 1:count + i*cols2);
end
[W1_2,W2_2] = v2Gv_average(g0,g1,g2,h1,W1_1,W2_1,W1,W2);

% vectorize

for i = 1:rows1
    Gv((i-1)*cols1 + 1:i*cols1) = W1_2(i,:);    
end
count = rows1*cols1;

for i = 1:rows2
    Gv(count + (i-1)*cols2 + 1:count + i*cols2) = W2_2(i,:);
end

Gv = Gv.';

Gv = Gv + regularization*v;

Gv = compress(Gv,indices,subsetSize);   