Is it possible to pass fitting coefficients as arguments?

Hello everyone,

I'm trying to perform a nonlinear fit via three parameters. However, I have to pass one of the parameters to an external function. Is it possible to do it? How?

Here is the part of my code where I try to do what I explained before.

 zeqn = @(coefs, y) coefs(1)+lh+y./coefs(2) + invfwlc((y*coefs(3))/kT);
 InitGuess = [100 0.01 10]

 fSSR = @(coefs, z, y) sum(( z - zeqn(coefs, y) ).^2);
   [MyCoefs, Sr] = fminsearch(@(MyCoefsDummy) fSSR(MyCoefsDummy,z,y),InitGuess,optimset('Display','iter','MaxFunEvals', 50000,'MaxIter',1000,'TolFun',1.e-8,'TolX',1.e-8));

"invfwlc" is the external function I want to evaluate.

Thank you in advanced.

[Star Strider]

Once you identify the parameters, passing one shouldn’t be a problem. What have you already tried? Are you having problems with it? Does it give you the answers you expect?

I had to make up values for y, z, lh, and kT, and an expression for invfwlc but with those additions, the code you posted ran without problems.

-------------------------------------------------------------------

Added ahead of posted code:

   z = 7;  y = 3;  lh = 13;  kT = 273*1.38E-023;  % Necessary to run code
   invfwlc = @(x) exp(-x);                        % Necessary to run code

Answer 1

[Alvaro]

Thank you for your quick response. My main problem is that the function that causes me trouble is the next one:

function lwlc = invfwlc(y)

l = 0.9999;

f2 = @(l) (1./(4.*(1-l).^2)-0.25+l);
df = @(l) (0.5.*(1.0/(1-l).^3)+1.);

lnew = l - 0.5*(f2(l)-y)./(df(l));

while abs(y-f2(l)) > 1.e-8
    l = lnew;
    lnew = l - 0.5.*(f2(l)-y)./(df(l));
end

lwlc = lnew;

end

It is some kind of numerical solver to invert another function (I don't know if I am explaining well). My purpose is to pass my variable "y" to the function "invfwlc" and get its inverse according to the f2 equation. Doing it, Matlab indicates me that x and y vectors are not of the same length, which is not true, and I can't perform the fit.

Any idea to solve it?

[Star Strider]

I’ve been working on this, but I cannot reproduce the error.

Please post the line (or lines) throwing the error. It is obvious that l is a scalar.

Is y a vector?