Fit Data to a Polar Equation

This is a little abstract, but I have data that is represented of some form of polar equation, and I would like to fit a polar equation to this data as accurately as possible. I don't want to fit it to a circle nor do I want to fit it to an ellipse. I want to fit it to a general polar function. Basically, what I am thinking of, is an analog of a polyfit of a high number to very accurately fit a cartesian function in polar coordinates. Just wondering if something like this exists in matlab.

Answer 1

[Star Strider]

I have no idea what ‘general polar funciton’ you are referring to. If you have a particular function, you can likely fit your data to it. The parameter estimation functions will work regardless.

There are several linear and nonlinear parameter estimation routines you can use. The details depend on the function you want to fit. You can use the cart2pol or pol2cart functions to transform your data between Cartesian and polar coordinate systems as well, then fit them.

[John Shackleton]

My apologizes, I'm not sure exactly how to word my question. Essentially what I'm asking, is how do you fit a polar function? As far as I know, general fitting functions, such as polyfit, don't necessarily work very well with fitting data that is circular/ellipsoidal.

[John Shackleton]

Wait, I am big dummy. Thanks for your help!

[Star Strider]

I can’t agree with your self-assessment. It’s necessary to frame the problem correctly and then look at the available parameter estimation functions. That’s not always straightforward.

My pleasure!

[ADITYA SENGUPTA]

Good day guys.

Its nice to see the discussions happening on similar grounds. I too have a doubt to get clarified on how to fit a logarithmic polar curve as a best fit for the polar co-ordinates I have. Basically, I am seeking for Polar Regression through Logarithmic Function. Can someone help me out in using MATLAB or any other suitable and user friendly software for this?