We introduce a general procedure for directly ascertaining how many independent stochastic sources exist in a complex system modeled through a set of coupled Langevin equations of arbitrary dimension. We apply our algorithm to two examples of systems showing Hopf bifurcation. We argue it enables one to define vector fields of stochastic eigendirections which can be used assess the path of minimum stochastic forcing in phase space, increasing the predictability of the system.
Recommended citation: Vítor V. Vasconcelos, et al.. “Principal axes for stochastic dynamics” Phys. Rev. E (2011) 84, 031103.