MATLAB: Nlinfit function, not enough input arguments

input argumentsMATLABnlinfit

Dear All,
I am trying estimate 15 parameters using the following function: Parameters are x(1), x(2) ….. x(15)
function F = myfun(x,y,X)
  F=(y - (1 + x(1)*(x(2)/0.0025662))*X(:,1) - (x(2)/x(3))*(sqrt(1-x(1)^2))*X(:,2) ...
       - x( 4)*X(:, 3) - x(5)*X(:, 4)) - x( 6)*X(:, 5) - x( 7)*X(:, 6) - x(8)*X(:,7) ...
       - x( 9)*X(:, 8) - x(10)*X(:, 9) - x(11)*X(:,10) - x(12)*X(:,11) ...
       - x(13)*X(:,12) - x(14)*X(:,13) - x(15)*X(:,14);
end
X is a matrix with 18026 rows and 14 columns y is the dependent variable with 18026 rows and 1 column
I have 15 parameters to estimate and their starting values are defined by xstart (1 row by 15 columns)
When I specify x = nlinfit(X,y,myfun,xstart) in Matlab
I get the following error: Not enough input arguments.
Could you help and let me know how to change the specification?
best, Srinivasan

Best Answer

Assuming that ‘y’ exists somewhere in your workspace, and since for MATLAB Statistics and Machine Learning Toolbox and Optimization Toolbox curve-fitting require two arguments, the first being the parameters and the second being the independent data, you would have to use this synatx:
@(x,X)myfun(x,y,X)
so your correct nlinfit call would be:
x = nlinfit(X,y,@(x,X)myfun(x,y,X),xstart)
although I cannot determine the reason you are using your dependent variable as an argument to your objective function. Since nlinfit will use it to fit your function, it is not necessary for you to use it specifically in your function. The best options is likely:
function F = myfun(x,X)
  F=((1 + x(1)*(x(2)/0.0025662))*X(:,1) - (x(2)/x(3))*(sqrt(1-x(1)^2))*X(:,2) ...
       - x( 4)*X(:, 3) - x(5)*X(:, 4)) - x( 6)*X(:, 5) - x( 7)*X(:, 6) - x(8)*X(:,7) ...
       - x( 9)*X(:, 8) - x(10)*X(:, 9) - x(11)*X(:,10) - x(12)*X(:,11) ...
       - x(13)*X(:,12) - x(14)*X(:,13) - x(15)*X(:,14);
end
and then:
x = nlinfit(X,y,@myfun,xstart)
Related Question