I want to use numpy to create a nonlinear calibration system. I have a set of pairs, where each pair is an empirically measured (x, y, z) coordinate, and a nominal (x, y, z) coordinate. I need to create a best fit multidimensional polynomial system of equations such that future empirically measured (x, y, z) coordinates can be mapped to their undistorted position. This was fairly trivial to do in one dimension with polyfit(), but I'm not sure how to extrapolate it to multiple dimensions. Any ideas?
Thanks
