I am using CVX to solve a sequence of large-scale problems whose feasible sets are identical to each other. Currently, I use a for loop over the entire problems which is computationally inefficient. Does anyone know a better practical way so that I can form the problem once and then use it in every iteration of the for loop?