Instead of only dealing with dynamic equations in forms of PDEs, it might be nice to also evolve other fields or scalars (if #533 is implemented). This would allow to solve general systems that are given by $\partial_t x = f(x, y)$ and $y = g(x, y)$. The strategy would then be to first determine y in each time step and then calculate f(x, y) to evolve x. This is already possible now with custom PDE classes (and also using user_funcs in the PDE class), but it might help if we expose an easier interface for this case. In particular, we could save the fields y in the output (whereas only the x need to be specified in the initial state).
Instead of only dealing with dynamic equations in forms of PDEs, it might be nice to also evolve other fields or scalars (if #533 is implemented). This would allow to solve general systems that are given by $\partial_t x = f(x, y)$ and $y = g(x, y)$. The strategy would then be to first determine
y
in each time step and then calculatef(x, y)
to evolvex
. This is already possible now with custom PDE classes (and also usinguser_funcs
in thePDE
class), but it might help if we expose an easier interface for this case. In particular, we could save the fieldsy
in the output (whereas only thex
need to be specified in the initial state).