dude you're the guy i need, i'm actually trying to understand how to do runge kutta method with vectors, cuz i need to simulate some particles falling and well...my question is specificly to know how to handle vector within RK4, please help me in anyway, thanks!!
I'm using rk4. The timestep per frame is t=t+0.1 (when running at 60fps)
I haven't experimented with any other methods as rk4 appeared quite accurate and was fast enough for my needs. I'm not sure what you mean by efficency but in terms of processor time the parallax lighting effects take about 100 times longer to complete each cycle than the rk4 method.
@VannevarBush The accuracy is not really an issue the screen resolution even at 1080p is way too small to get a visually appreciable difference. I just opted for speed to keep the frame-rate nice
Do you use Rk4 or rk2? and what kind of timestep does this allow you (and did you try verlet or any other integrators before this one? what kind of efficiency does it offer?)
This looks interesting. Have you released it yet?
fremandn 2 years ago
dude you're the guy i need, i'm actually trying to understand how to do runge kutta method with vectors, cuz i need to simulate some particles falling and well...my question is specificly to know how to handle vector within RK4, please help me in anyway, thanks!!
kamihakai 2 years ago
its all about vector my friend, theres not mehtod that does not work with them,, including eulers, and improved eulers etc..
avion85 2 years ago
Wouldn't Vode Adams be more appropriate?
lol j/k but fun looking little game
19mike65 3 years ago
I really like this the more I watch it! When are you going to release it ? :D
buttaflutta 3 years ago
I'm using rk4. The timestep per frame is t=t+0.1 (when running at 60fps)
I haven't experimented with any other methods as rk4 appeared quite accurate and was fast enough for my needs. I'm not sure what you mean by efficency but in terms of processor time the parallax lighting effects take about 100 times longer to complete each cycle than the rk4 method.
Kyuzumaki 3 years ago
I see, thank you! ;)
buttaflutta 3 years ago
@Kyuzumaki
What about Dormand-Prince method? This is the most accurate method I know (to solve DE's)
VannevarBush 1 year ago
@VannevarBush The accuracy is not really an issue the screen resolution even at 1080p is way too small to get a visually appreciable difference. I just opted for speed to keep the frame-rate nice
Kyuzumaki 1 year ago
@Kyuzumaki
Okay I see. That is another priority I guess.
VannevarBush 1 year ago
Nice ;)
Do you use Rk4 or rk2? and what kind of timestep does this allow you (and did you try verlet or any other integrators before this one? what kind of efficiency does it offer?)
buttaflutta 3 years ago