Data fitting using GLM model

Hello,

I need to fit data with this dependece r(H10,T,P,t)=a0(H10)+a1*t+a2*t^2+ a3*T+a4*P+a5 and I would like to use GLM.

I am struggling with the very beginning. Could you possible give me a hint how to fit data with 4 different dependencies and obtaint a1..a5 ?

Thank you.

[John D'Errico]

Is A0 a FUNCTION of H10? Or have you forgotten to put a * in there to denote multiplication. Is there a way that we should know this by looking at your question? (NO.)

Answer 1

[John D'Errico]

Trivial.

Well, trivil, If we assume that A0(H10) was actually intended as A0*H10.

r(H10,T,P,t)=a0*H10 + a1*t + a2*t^2 + a3*T + a4*P + a5

First, set up a matrix. Assuming you have variables, as COLUMN vectors, for the variables: H10, T, P, t and r.

n = numel(H10);
M = [H10,t,t.*t,T,P,ones(n,1)];
a012345 = M\r;

The vector a012345 will contain the coeffiicients you wish to find.

You could also use tools from the stats toolboix, such as regress, etc.