How can I find distances between 100 points such that I have a set of distances of each point from rest of the points.

Hi Everyone,

I have distributed 100 points randomly in a 100x100x100 m3 3D space. I want to find distance of each point from rest of 99 points such that I have a distance set of 4950 different values. so the distance of 1st node from resy 99 nodes and then distance of 2nd node from rest 98 nodes and so on.

I cannot change the style of the code since I have values saved in these variables.

I have following code:

Close all
Clear all
n=100;
xm=100;
ym=100;
zm=100;
x=0;
y=0;
z=0;
for i=1:1:n
for j=1:1:i
for k=1:1:j
if (i==j && j==k & k==i) % so I dont get 100x100x100 different values and only 100 values with x,y,z coordinates.
S(i).xd=(1,1)*xm; %I cannot change this
XR(i)=S(i).xd;
S(j).yd=rand(1,1)*ym;
YR(j)=S(j).yd;
S(k).zd=rand(1,1)*zm;
ZR(k)=S(k).zd;
Points=[XR;YR;ZR]
end
end
end

Answer 1

[Ameer Hamza]

Just use pdist() function

X = [..]; % create 100x3 matrix
dists = pdist(X)

[Tamoor Shafique]

Thanks @Ameer Hamza for your prompted response but how do I make an x=[..] 100x3 matrix? isn't Ponts[XR;YR;ZR] my 100x3 matrix already?

[Ameer Hamza]

Yes, It seems that from your code. There are easier ways to create the matrix [XR;YR;ZR] than using a triple for-loop, but it depends on how your data is stored in the struct. If you can share a sample struct in a .mat file, then I can suggest an easier solution. But if your for-loop works for you, then that is fine too.

[Tamoor Shafique]

Please see attached. I am not good with Matlab if there are esier ways I don't mind adapting as long as it gives me distribution of nodes in full 100x100x100 network and then splits it into 27 sub cubes so i can get local info about each subnode like data in 27 sets.