<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />Is anybody familiar with the technique this customer is asking about below? FYI, I already told him about the new Hough transformations built into the clustering platform in JMP 12, which seems to be intended for a similar purpose, but a different technique.
"Measurements on silicon wafers show 2-D variation that I would like to model in radius-angle rather than x-y. I hope to fit orthogonal polynomials on the unit circle, possibly the Zernike coefficients.
If I want to fit the first 15 terms of the expansion, must I add 15 columns to my data table? Is there an easier way, like the 1-D fit polynomial or the Nonlinear Model Library? See attached example."
Thanks in advance,
Jordan
