quote:Original post by Timkin
Kylotan: A* is NOT defined as search involving an OPEN list and CLOSED list of branches in the search tree. That is merely an efficient data management method when implementing A*.
Hmm... I know that there doesn''t have to be List of open and closed nodes, but the A* algorithm requires the ability to know which nodes are awaiting traversal, and which nodes have already been traversed and at what cost, yes? Now I didn''t mean to imply that you would need a concrete data structure for both of these things: more that the algorithm needs to be able to determine which nodes are in which state. List in the abstract sense, not the concrete sense. Sorry for any confusion.
(For anyone who cares though, I''d use a priority queue for the open nodes and a hash table (implemented as a bitset) for the closed nodes.)
quote:Original post by MarkyD
Well, isn''t that the idea. I always thought that algorithms such as A* were ''basic'' algorithms. Whenever I make enemy AI for my games, I start the AI from scratch, so that I can base it more around my own game, and therefore make it better.
Yeah, but with all due respect, you''re not making your algorithm any better than what already exists. Algorithms tend to be extensively proven by mathematical experts, and are popular because they''re exceptionally good at what they do. A* is certainly not ''basic'' in the sense of being something than you can beat with just a little thought. Generally, if you think you can do better than A*, you''re in one of 2 situations: either you''re very good with algorithms and understand all the implications of your context, or you just don''t understand A* well enough. No ''alternative'' algorithm posted on this board in the last week has come close to being more efficient than A*, and I could prove that if needed, which is why I said what I said.
