Re: LuaJIT2 performance for number crunching

[Francesco Abbate <francesco.bbt@...>]

2011/2/21 T T <t34www@googlemail.com>:
> But then the user supplied function needs a different format depending
> on the size of the ode, right?
>
> for small ode:
>
>  dydt_1, dydt_2, ... = f(t, y_1, y_2, ...)
>
> for large ode:
>
>  f(t, y, dydt)
>
> Hmm, not that nice from the UI design POV.

Yes, you're right this will be a minor problem but the advantage of
having explicit variables for system of small dimension will be also a
benefit in term of simplicity. It is also true that the interface
based on arrays will be usable also for small systems so the user will
be free to use this latter interface if he preferes.

I'm working in the vector form of the ODE integrator right know, I'm
planning to use FFI to call directly blas routine. In the meantime I'm
working also in the function integration routines (quadpack like). The
Gauss-Kronrod integrator is already there in pure Lua + FFI :-)

-- 
Francesco