How to get the fitted data by using Matlab's "Fit" Function?

For example, by using below code

load franke;

fitresult = fit( [x, y], z, 'poly33' );

From fitresult, I can get the formula used to fit x,y and z data

But how can I get the fitted data matrix via this "fit" function?

Answer 1

[Wayne King]

It gives you a 1-D vector, but that 1-D vector is a polynomial in two variables, hence it is a function of the form f(x,y)=z

Calling

formula(fitresult)

shows you exactly what the terms in x and y the coefficients multiply

Answer 2

[Wayne King]

Hi, you can use the coeffvalues() method on the fitresult object.

    coeffs = coeffvalues(fitresult);

That will give you the coefficients of your polynomial in x and y.

To see how those coefficients enter into the polynomial equation:

    formula(fitresult)

[Zheng]

Thanks, Wayne. But how can I use this formula to calculate the surface (shown as below)? This formula can only help me get a 1-D vector, not 2-D matrix.

[grace]

May I know what is the code for goodness of fit? coeffs = coeffvalues(fitresult); Similarly, rsquare = code(fitresult); I wanted to get the rsquare values from cftool to display in command window.