2D interpolation and differentiation problem

Hello,
I hope this is the right place to ask this question, any suggestions would
be gratefully appreciated.

I have a set of data in a spread*** consisting of a lot of points, in
polar coordinates, and a Z value for each point.

e.g (not real data)
R Theta Z
3.456 0.128 1.027
7.790 -0.054 1.754
4.567 0.274 1.328
........
about 6500 points of data.

The points are not evenly spaced, but the surface of interest is well
covered. I can order them by R, or by Theta.

I would like to calculate samples of dZ/dTheta and dZ/dR over the surface,
and then plot a family of curves of dZ/dTheta v Theta for various
constant values of R, and R*dZ/dR for various constant values of Theta.

If I could find a way to linearly interpolate the data to make a uniformly
spaced array of Z for a set of values of R and a set of values of Theta it
seems to me as though the differentiation would be trivial.

Any suggestions?
.

Quantcast