polyfit with multiple dependent variables
94 views (last 30 days)
can handle multiple dependent variables. But when I tried
Fit = polyfit([x1 x2], x3, 1);
I got an error message
Error using polyfit (line 44)
The first two inputs must have the same number of elements.
Although I checked x1, x2, x3 has the same length. What can one do?
Bjorn Gustavsson on 17 Dec 2020
The polyfit function only works for 1-D polynomials, as far as I understand things. However there is an n-dimensional polynomial-fitting function package on the file exchange: polyfitn. That should give you the tools necessary to solve this problem. I've used it without much problems and to sensible success.
More Answers (1)
Remy Lassalle-Balier on 17 Dec 2020
I think polyfit does not suport multiple dependent variables. The documentation is quite clear about the fact the first parameter should be a vector and not a matrix.
You can still run your fit manually:
X = [ ones( numel( x1 ) , 1 ) , x1(:) , x2(:) ];
fitParam = X\x3(:);
Fittedx3 = X * fitParam;
This is better explained here.