The Gaussian Distribution Thread
category: residue [glöplog]
Just so you know.
Haha, good one, noouch!
...
Where was Einstein's 1st wife?
This is how I build gaussian coefficients
Code:
int *gauss_coef;
void tack_line(int offset, int depth) {
int a, b, k;
for(k = 0;k< depth; k++) {
if( (offset - (depth -1 ) + k) == offset)
a = 0;
else
a = gauss_coef[offset - (depth-1) + k];
if( (offset - (depth-1) +k -1) < (offset - (depth-1)))
b = 0;
else
b = gauss_coef[offset - (depth-1)+k-1];
gauss_coef[k+offset] = a+b;
}
}
void build_coef(int depth) {
int i,j, k;
j = 0;
for(i = 0;i<depth;i++)
j+=i;
gauss_coef = (int *)malloc(j * sizeof(int));
gauss_coef[0] = 1;
j = 1;
for(i = 2; i< depth;i++) {
tack_line(j, i);
j+=i;
}
}
sigflup, that is pretty much how i would do it also.
xrs, in ft2? :)
You guys suck. I have a statistics oral exam on monday, my 3rd year trying to pass that stupid subject, and i'm studying since today morning nonstop. I come to pouet for some relaxation and this is on to of the BBS...
:(
:(
Oh, and btw that image at the top is about the 3-sigma rule, which is based on Tchybyszews inequality
I predict that, in contrary to what the thread name says, the actual distribution of posts in this thread over time will be more poisson-like than gaussian.
pommak, I can't agree more. :)
its mode looks like...
There's a curve @ the top.
