1
$\begingroup$

I am exploring a 2 parameter space performing simulations. As a result I get a surface as a function of these 2 parameters. I know this is probably simple but I don't know how to look for it.

Now I want to get a level set of this surface, thats it, the pair-wise values that are the closest to a certain value of the outcome measure of the simulations. But I can't figure out how to do this nor whats the name of this kind of problem in order to look for it!

For clarification, lets say my parameters are x and y, then, the data I have looks like:

x=[1:5 1:5 1:5 1:5 1:5 1:5];
y=[ 0*ones(1,6) .2*ones(1,6) .5*ones(1,6) .75*ones(1,6) 1*ones(1,6)];
z = [ 152 89 106 134 108 103 0 101 105 104 89 13 109 104 102 115 112 187 200 111 99 85 111 97 48 76 53 34 245 100];

if I want to plot it in matlab as a surface I can do:

x = reshape(x,5,6);
y = reshape(y,5,6);
z = reshape(z,5,6);
surf(x,y,z)

I want to get the level set by any interpolating or finding the closest in my actual values.

Would be really handy is someone can suggest me whats the way this problem is usually tackled.

$\endgroup$
1
  • $\begingroup$ Gaussian processes (aka Kreiging) are one very common method of interpolating between data of this type. $\endgroup$
    – Sycorax
    Commented Nov 9, 2015 at 21:16

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.