Least squares equation fit help

Hi I'm trying to find the solutions (a,b,c,d) to the following equation:

y=a+b(x)+c(ln(x-d))

where x and y are known vectors.

I have been trying to use the curve fitting toolbox and writing it as a custom equation but keep getting the following error: "Complex value computed by model function, fitting cannot continue."

Any help would be great, I have a fair bit of Matlab experience but am new to the curve fitting toolbox.

Tim

Answer 1

[the cyclist]

I don't have the Curve Fitting Toolbox, but if you also have the Statistics Toolbox, then my answer and example in this question may help you:

http://www.mathworks.com/matlabcentral/answers/23355-non-linear-least-squares-regression

[Tim]

Thanks that is a great help (I do also have the stats toolbox). I'm now having some extra little niggles with imaginary numbers (due to having a log of a negative value) and the nlinfit function is also exceeding the iteration limit but will try and sort these.

Many thanks!