please who is the mentor for Native Julia solvers for ODEs?
On Thu, Mar 13, 2014 at 10:25 PM, Jiahao Chen <[email protected]> wrote: > Hi Niluka and Kongne, > > ODE.jl already implements two Runge-Kutta based methods (ode23 and > ode45). I don't think we really need more Runge-Kutta algorithms, > although someone who has worked on ODE.jl recently should feel free to > correct me. > > If you want to _show_ us implementations of a simple textbook > algorithm like RK4 as part of your application to demonstrate that you > know something about ODE solvers and Julia's syntax, that's fine, but > it's unlikely that we will actually incorporate textbook code into the ODE > package. > > Kongne, if your question is about writing solvers for first-order ODEs > or second-order ODEs, consider that in principle, an ODE solver for > first-order ODEs will suffice to solve ODEs of any order, since you > can always rewrite higher order ODEs as vectorized first-order ODEs. > For example, y'' = -y can be rewritten as the coupled system [y; z]' = > [z; -y]. This is quite a common trick used to reduce the order of > ODEs. > > We can continue the discussion of ideas for specific projects in the GSoC > issue: > > https://github.com/JuliaLang/ODE.jl/issues/18 > > Thanks, > > Jiahao Chen > Staff Research Scientist > MIT Computer Science and Artificial Intelligence Laboratory > > > On Thu, Mar 13, 2014 at 4:50 PM, Kongne gael stephanie > <[email protected]> wrote: > > what i mean is,am writing an algorithm to solve only first order or > second > > order? > > > > On Thursday, March 13, 2014 9:28:05 PM UTC+1, Stefan Karpinski wrote: > >> > >> On Thu, Mar 13, 2014 at 4:22 PM, Kongne gael stephanie > >> <[email protected]> wrote: > >>> > >>> i wish to know if i use only first order or second order to implement > >> > >> > >> Sorry, I don't understand. Could you elaborate? >
