primal_dual_hybrid_gradient

Primal-dual hybrid gradient (PDHG) algorithm studied by Chambolle and Pock.

The primal-dual hybrid gradient algorithm is a flexible method well suited for non-smooth convex optimization problems in imaging.

Functions

pdhg(x, f, g, L, niter[, tau, sigma])	Primal-dual hybrid gradient algorithm for convex optimization.
pdhg_stepsize(L[, tau, sigma])	Default step sizes for pdhg.