I am not sure why the solution of CVX is transpose

273927923_486570796422332_2821828847333166113_n864×1152 60.6 KB

It’s my first time using CVX.
Although I read the documentation, I still confuse why the solution of pc_u looks so weird especially the
matrix size of pc_u

There is my code
for u=1:N
cvx_begin
size(pc_user)
variable pc_u(u,m)
minimize( -(sum(pc_u(u,m).*F_u(u,:)))/numactive)
subject to
sum(pc_u(u,:)) == S
0<=pc_u(:,m)<=1

cvx_end
pc_u

end

the value of parameter respectively
N=3 M=10
m had been equal to 10

This might be variable pc_u(1,M). In your above code, you define a new u-by-m variable array by variable pc_u(u,m) for every iteration (also u is changing), which might not be what you wanted. If you need to save your solution of every iteration to an array, you can define another M-by-N array outside the loop rather than to define it CVX variable.

Really appreciate your help !!!