4 ms·
I know it is educational and it does not matter in presence of dissipation, but it would be nice to mention that Euler method is not only an approximation but a
by mbq 14y ago
I know it is educational and it does not matter in presence of dissipation, but it would be nice to mention that Euler method is not only an approximation but also terribly inaccurate one. It won't produce any stable oscillations or orbits for instance.
- bkanber 14y agoI actually mentioned that in a comment lower on this page. I will be talking about ODE solvers a couple of articles from now. The Euler method isn't acceptable for actual simulation but it's the easiest to understand and it gets the job done for now!
- xxbondsxx 14y agoThe Velocity Verlet integration method isn't much of a jump in complexity (http://en.wikipedia.org/wiki/Verlet_integration#Velocity_Verlet http://en.wikipedia.org/wiki/Verlet_integration#Velocity_Ver...) and works _great_ for kinematic simulators. As a cool feature, it maintains the energy of the system, so you can slow down, stop, and then rewind time and arrive at the same previous configuration, all by changing the delta T's inside the evaluation.
- bkanber 14y agoI'll definitely be talking about this one in my Solvers article!
- baq 14y agobe sure to read "numerical recipes in c" chapter 16 first. fwiw the only thing you need to do to be one of the best references on numerical integration in js on the net is translate the code to javascript...
- mbq 14y agoNot if you are doing dissipation, as in this case.