daepy.derivatives

daepy.derivatives.approx_jac(stream, x, epsilon, pool, *args, **kwargs)

Compute a finite difference approximation of the jacobian of a serialised function stream at x with stepsize epsilon using a multiprocessing.pool.Pool pool. Extra arguments and keyword arguments for the serialsed function may be passed following the other arguments. A python function can be serialised using dill.dumps().

daepy.derivatives.step(stream, x, epsilon, c, current, *args, **kwargs)