Using fitnlm with multiple predictor variables

I am trying to perform non linear regression using 4 independent variables and a 2-term exponential model. I am using the following section of code:

X = [x1 x2 x3 x4];
b = [1 1 1 1];
x = ones(size(X));
modelfun = @(b,x) b(1,:).*exp(b(2,:).*x) + b(3,:).*exp(b(4,:).*x);
beta0 = ones(4,1);

However, I keep getting an error explaining that modelfun returns a vector of 4 columns where it should be 1 column.

I have tried as many variations of modelfun as a I can think of, can someone help?

[Eric Garr]

Did you ever figure out a solution?

Answer 1

[Image Analyst]

See atached demos.