A stochastic differential equation (SDE) is an ODE with a rough driving control. Ito calculus gives a recipe for solving a SDE by an approximation that converges only in L2 sense with respect to the underlying randomness. Ideally we would like to replace the rough control with a smooth approximation, solve the resulting equation and pass to the limit. Unfortunately this deterministic...