jclosure

import numpy as np

primary = np.array([[40., 1160., 0.],
                    [40., 40., 0.],
                    [260., 40., 0.],
                    [260., 1160., 0.]])

secondary = np.array([[610., 560., 0.],
                      [610., -560., 0.],
                      [390., -560., 0.],

amroamroamro

Python version of the MATLAB code in this Stack Overflow post: https://stackoverflow.com/a/18648210/97160

The example shows how to determine the best-fit plane/surface (1st or higher order polynomial) over a set of three-dimensional points.

Implemented in Python + NumPy + SciPy + matplotlib.