Constructing bivariate tensor product spline

Hello,

i have three vectors which contain the x, y, and z values of data points (x,y,z = [1xn]). Now i want to construct a tensor product splines out of these three vectors. I found an example but i'm not sure how to use this example with my input data. Can anyone help me please. Thanks

[Unai San Miguel]

What do the x,y,z vectors represent in you application?

1. points in a (explicit) surface z = f(x,y)?
2. the same but with some scatter on z, I mean, z = f(x,y) + noise?
3. the same but with some scatter on all the variables?
4. points in a (parametric) surface x = f(u,v), y = g(u,v), z=h(u,v)?
5. the same but with some scatter on all variables?
6. ...

Depending on the kind of surface (I assume it is a surface what you are seeking) you want to obtain you will need one preprocess or another. Once you have your data in a suitable shape then you can apply some tensor product approximation to them.