Thanks to the authors of
DifferentialEquations.jl, it is quite a nice package.
I would like to evaluate the numerical accuracy of multiple solvers for an SDE. For this, I intend to numerically solve for the same path, i.e. using two different solvers but let them share the same noise realization of the Brownian noise.
Where I now solve a
SDEProblem using a single solve,
sol = solve(prob,SOSRA(),dt=0.05,adaptive=true)
I would ideally like to be able to write something like
sols = solve(prob,[SOSRA(),SRIW1()],dt=0.05,adaptive=true)
and obtain a pair of solutions,
sols which share the same noise realization.
I have searched through the documentation and have been unable to find a way to explicitly share the noise. Is it possible?