How to Generate a Given Shaped Cluster of Data Points?

Hi,

I am trying to generate some example graphs of data with irregular distribution shapes. I want a simple scatter plot of a data cluster that looks like a "C", a cluster that looks kind of like a squished jellybean, and a cluster that has tightly clustered points on the left-hand tail and disperse points on the right hand. Is there like a drawing tool where Matlab would randomly fill the space that I pre-selected? Is there a good way to randomly make these data clusters, or do I just need to define them by hand?

Thanks, Cyndi

[Adam Danz]

Do you have the coordinates (x,y) that outline the shapes?

Answer 1

[KSSV]

I would download the required shape image from google....get the boundary of the shape and generate random points inside the shape. Check the below example.

I = imread('Letter_c.svg.png') ;
I = rgb2gray(I) ;
[y,x] = find(~I) ;

idx = boundary(x,y) ;
x = x(idx) ; y = y(idx) ;
% Generate random numbers 
ax = min(x) ; bx = max(x) ;
ay = min(y) ; by = max(y) ;
N = 10000 ;
xx = (bx-ax).*rand(N,1) + ax;
yy = (by-ay).*rand(N,1) + ay;
% Get points inside the C shape 
idx = inpolygon(xx,yy,x,y) ;
figure
hold on
plot(x,y,'b')
plot(xx(idx),yy(idx),'.r')