Find Z-value corresponding to X and Y points based on interpolated data
13 views (last 30 days)
Show older comments
I would like to perform 2D interpolation for scatter data that is supported by Matlab Coder package.
I have a set of data, with x,y and z cordinates, in the following format.
X Y Z
1 2 10
2 2 8
2 4 9
3 2 9
3 4 10
3 6 9
So at Z(1,2) = 10, Z(2,2) = 8 and so on.
I wish to find the value of Z(1,3), Z(3,1) for example and eventually output the code to C/C++ using Matlab Coder package.
I tried using griddata and scatterInterpolant which is what I wish to achieve, however, it is not supported by Coder package. So I tried to work with interp2 or gridedInterpolant, however I am not sure how to convert my sample data to full grided form.
Is there any way to convert the above data to be used ny interp2 or gridedInterpolant?
Or is there other function that I can use and I wasn't aware of that?
Appreciate for any help! Thanks!
2 Comments
Answers (1)
Karim
on 6 Jan 2023
Note that you have very little points. Hence one possibility is to perform linear interpolation, below you can see one method on how to do this.
% original data by OP
x = [1 2 2 3 3 3]';
y = [2 2 4 2 4 6]';
z = [10 8 9 9 10 9]';
% plot the points in 3D space
figure
scatter3(x,y,z,'r','filled')
grid on; view([-70 30]);
title('Original points')
% set up the grid for the interpolation
x_fine = linspace(min(x),max(x),20);
y_fine = linspace(min(y),max(y),20);
[Xq,Yq] = meshgrid(x_fine,y_fine);
% create a linear fitting trough the points
sfun = fit([x, y],z,'linearinterp');
Zq = feval(sfun,Xq,Yq);
% plot the results
figure
surf(Xq,Yq,Zq)
grid on; view([-70 30])
title('Surface through interpolated points')
figure
scatter3(Xq,Yq,Zq,'g','filled')
grid on; view([-70 30])
title('Interpolated points')
0 Comments
See Also
Categories
Find more on Multirate Signal Processing in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!