Can we have all possible x (list of all possible candidates) in the valid search space instead of just one x that satisfies y=Ax while solving minimization of L1 norm of x subject to Ax==y?
No for Cvx. There are other tools that can do it.
Thanks. Which tools exactly?