Using csaps (or similar) to create a 3D smoothing spline

38 views (last 30 days)
Laurence hutton-smith on 28 May 2015
Answered: Bryan Howell on 31 Oct 2016
I have a series of 3D points and am looking to create any sort of smoothing spline, I was hoping to use scaps, but it only seems to take in 2D data, can anyone help with this? Is there for instance a way to input 3D data into it or another function I could use?

Dasharath Gulvady on 29 May 2015
Edited: Dasharath Gulvady on 29 May 2015
You can do this using "csaps" function. However the trick is to make "z" your independent variable.
[pp,p]=csaps(z,[x;y]);
val=fnval(pp,z);
figure,plot3(x,y,z);
hold on,plot3(val(1,:),val(2,:),z,'r-')
grid on
Laurence hutton-smith on 30 May 2015
Thank you, this is extremely helpful.

Bryan Howell on 31 Oct 2016
The accepted answer is quite elegant but doesn't work for all types of trajectories, especially those where the z value is relatively constant.
An alternative approach is to fit a smoothing spline for each dimension and then combine. For example...
% given data in a point matrix, xyz, which is 3 x number of points
[ndim,npts]=size(xyz);
xyzp=zeros(size(xyz));
for k=1:ndim
xyzp(k,:)=ppval(csaps(1:npts,xyz(k,:)),1:npts);
end
This approach is also not a panacea but can address some cases where the first answer might fail.