I decided to use a hybrid method. Cached paths aren’t flexible. The bot will calculate paths in his closest area. Farther he’ll use cache.
Written by MM. Posted at 6:11 am on January 29th, 2010
7 comments.
Post a comment.
some sort of quad-tree of paths ?
Maybe the bot should work backkwards? As in, he finds his waypoints from his target to him, not from him to his target. That might be a little slow actually… And confusing for the bot.
Now just put it down and move on.
In any case don’t even think about all the ways you can optimize
and modify it.
@FinDude
Put it down? It must be done before the end of the month. Update deadline.
Two more days to go. Boy oh boy!
You could use differentials to calculate the exact distance from which the bot should use caching to maximize performance. Although it would be hard to find the time/complexity functions for both algorithms.
Eg,
f(z) = g(x) + h(y)
f,g,h //time/complexity functions of distance.
z = x + y //full distance to objective
x //distance without caching
y //distance with caching
find minimal value of function f and its optimized
and btw, its SCIENCE!
Gnoblar: its impossible to find the best route this way. Or maybe there is, you would probably win a Noble prize for inventing that.
Anonymous: that would probably take more time to compute than doing the actual pathfinding:)