Simulating `Complex' Problems with Quantum Monte Carlo

We present a new quantum Monte Carlo algorithm suitable for generically complex problems, such as systems coupled to external magnetic fields or anyons in two spatial dimensions. We find that the choice of gauge plays a nontrivial role, and can be used to reduce statistical noise in the simulation. Furthermore, it is found that noise can be greatly reduced by approximate cancellations between the phases of the (gauge dependent) statistical flux and the external magnetic flux.