Module problem 

Stationary (elliptic) finite-difference problems.

Unlike crate::numeric::finite_difference::pde::PdeProblem, which marches backward in time from terminal data, an elliptic problem has no time variable. Its data are the boundary conditions on the domain, and its solution is the direct solve of

L u = f

where L is the same spatial operator assembled from crate::numeric::finite_difference::discretization::Transport. In the standard elliptic sign convention the operator is -T u + rho u where T is the transport operator and rho is the reaction coefficient.

Traits

EllipticControlProblem

A stationary (infinite-horizon) stochastic optimal control problem.

EllipticProblem

A stationary finite-difference problem -T u + reaction * u = f on a grid.

EllipticProblemExt

Convenience access to the boundary condition on a single side.

Functions

zero_field

Builds a zero field over the grid for elliptic iteration.