Curve fitting: Fourier series fit in polar coordinates

Dear colleagues,

is there a posiibility to use curve fitting with Fourier series for data in polar coordinates?

Currently I am fitting two curves in cartesian coordinates with Fourier series. For the fitted curves I use the same x-values in cartesian coordinates and calculate the corresponding y-values. However if I transform them to polar coordinates the resulting theta-values for both curves are different. It would be important to have the fitted curves in polar coordinates because there I can easily handle the data sets of both curves (subtract the rho-values for the same theta-values). Is there a workaround for my issue?

Best regards,

Felix

[felix.b]

I found a solution using cartesian coordinates via the normal of one curve to the other. That's why the topic is closed.