I haven't looked at the appropriate source code parts, but I suppose ai.planet's physics engine makes heavy use of trigonometry, with precalculated hash tables of course, to speed things up.
However you could program 99% of games and non-scientific applications using only the 4 basic math operators. Even if you had to use higher math, you would in most cases, where you don't need pure 100% precision, rely on precalculated data.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2003-10-03
I thank you for your quick answer. I have read it.
In any case you have a very active project. And even time to answer question from strangers
I asked this because there was a question about speed.
When you look at nature. Animals don't use math, is it? So there has to exist fast ways as well.
I have an os project here about neural nets, and we can't use our dear math box. It's
maybe a pity, but I will explain it to you.
The incoming signals aren't known. They are in fact events, surprises. Suddenly is a notion, not met before is a second notion. We can also find suddenly a new path. And those signals enter my demo. And now the cells change by that. And yes there comes some reaction.
But now a second time a signal enters, again with unknown parts, how will the already changed neural net now answer?
You talki about margins as well. Yes, all has margins, isn' it.
This means I can't use in my project math. I have only moves or shifts a = b or a++, IFs
and enormous amount of statement of while loops, simulation the parallel way nature works.
I have streams, only streams, and then I know the stream can flow to another direction, and also
sometime back.
It's nice thinking about such things. But I like also how you do it.
I wish you much success.
Ed
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2003-10-03
Like you I've not much time. So it's easier for me
to email. My email address is
ameulen@users.sf.net
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The physics has the most math in aiplanet, mostly trigonometry. The creatures use some functions like arctan2 to face their target. In general the animals are all finite-state-machines, with some simple learning, flocking, mating, and evolution algorithms to spice them up.
The missile defence uses a search function to find the path to the asteroid.
The file cAIPosition.pas has most of the math to run the physics.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm more in fuzzy logic (even programmed some games with it in the DOS times) because I don't really get how to set up neural nets and how to implement them and for what purpose.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
My question isn't you use numbers. but do you use also mathematical functions?
And my second simple question is. Why did you choose the mathematical tool box for this?
I'm also a math, but I know there are also other ways.
Have a very good time
Ed
I haven't looked at the appropriate source code parts, but I suppose ai.planet's physics engine makes heavy use of trigonometry, with precalculated hash tables of course, to speed things up.
However you could program 99% of games and non-scientific applications using only the 4 basic math operators. Even if you had to use higher math, you would in most cases, where you don't need pure 100% precision, rely on precalculated data.
I thank you for your quick answer. I have read it.
In any case you have a very active project. And even time to answer question from strangers
I asked this because there was a question about speed.
When you look at nature. Animals don't use math, is it? So there has to exist fast ways as well.
I have an os project here about neural nets, and we can't use our dear math box. It's
maybe a pity, but I will explain it to you.
The incoming signals aren't known. They are in fact events, surprises. Suddenly is a notion, not met before is a second notion. We can also find suddenly a new path. And those signals enter my demo. And now the cells change by that. And yes there comes some reaction.
But now a second time a signal enters, again with unknown parts, how will the already changed neural net now answer?
You talki about margins as well. Yes, all has margins, isn' it.
This means I can't use in my project math. I have only moves or shifts a = b or a++, IFs
and enormous amount of statement of while loops, simulation the parallel way nature works.
I have streams, only streams, and then I know the stream can flow to another direction, and also
sometime back.
It's nice thinking about such things. But I like also how you do it.
I wish you much success.
Ed
Like you I've not much time. So it's easier for me
to email. My email address is
ameulen@users.sf.net
The physics has the most math in aiplanet, mostly trigonometry. The creatures use some functions like arctan2 to face their target. In general the animals are all finite-state-machines, with some simple learning, flocking, mating, and evolution algorithms to spice them up.
The missile defence uses a search function to find the path to the asteroid.
The file cAIPosition.pas has most of the math to run the physics.
I'm more in fuzzy logic (even programmed some games with it in the DOS times) because I don't really get how to set up neural nets and how to implement them and for what purpose.