How to plot a surface-like plot when the data is not regularly spaced

I have 3-d data, where first two columns (x,y) are independent and the third column (z) is dependent upon first two. I want to have a surface plot of the data so that I can see what region in (x,y) space gives low values of z and what region gives high values of z. But I am not able to use a surface plot because the data is not regularly spaced as required for using 'surf'. I have attached the mat-file containing three columns for reference. Is there a good way of solving this problem?

Answer 1

[Michal Dobai]

How about this?

[xi,yi] = meshgrid(0:0.01:10, 0:0.01:10);
zi = griddata(demodata(:,1),demodata(:,2),demodata(:,3),xi,yi);
handle = surf(xi,yi,zi, 'EdgeColor', 'none');

% I adjusted color limits for better-looking plot of your data
handle.Parent.CLim = [-100 -80];

There is the whole page about it MATLAB in documentation: Interpolating Scattered Data

Result plot:

Is this what you need?

[Abhinav]

Yes, this seems like perfect solution to my problem. I was not aware of 'griddata' function. Thanks a lot!

[Michal Dobai]

You are welcome. I'm glad I could help.

[Les Beckham]

If you have to deal with smoothing out noise in the data, or extrapolating beyond the edges of your grid, you might want to take a look at John D'Errico's excellent File Exchange submission gridfit.

[Abhinav]

Thank you @Les.