Comment on ''Numerical method for colored-noise generation and its application to a bistable system''

It is argued that a recently proposed algorithm [K.Y.R. Billah and M. Shinozuka, Phys. Rev. A 42, 7492 (1990)] for the integration of stochastic differential equations in the presence of correlated noise does not introduce a substantial improvement over existing schemes. We point out that although this scheme could be very interesting for noise characterized by particular correlation functions, for colored noise the present scheme is sensibly slower than existing algorithms. Some apparent discrepancies between the result of simulations carried out with the new algorithm and previous numerical work are explained.