Fitting nonlinear transcendental equation to data Via lsqcurvefit function

I want to fit J-V curve of a solar cell and the resultant equation is a transcendental equation. I am attaching my code below.

clear all;clc;close all;
load 'J-V curve.mat';
plot(v_save,j_save);hold on
F=@(k,xdata,ydata) (k(1)-k(2)*exp((xdata+ydata*k(3))/(0.025887*k(4)))...
    -((xdata+ydata*k(3))/k(5)));
k0=[18 2e-8  .005  1  5];
[k,resnorm,~,exitflag,output] = lsqcurvefit(F,k0,v_save,j_save);
plot(v_save,F(k,v_save,j_save));

Here, F is a transcendental equation. However the code gives the following error. Not enough input arguments.

Error in
fitting_j_v>@(k,xdata,ydata)(k(1)-k(2)*exp((xdata+ydata*k(3))/(0.025887*k(4)))-((xdata+ydata*k(3))/k(5)))

Error in lsqcurvefit (line 202)
            initVals.F = feval(funfcn_x_xdata{3},xCurrent,XDATA,varargin{:});

Error in fitting_j_v (line 7)
[k,resnorm,~,exitflag,output] = lsqcurvefit(F,k0,v_save,j_save);

Caused by:
    Failure in initial objective function evaluation. LSQCURVEFIT cannot continue.

Will someone please help me to solve out the problem?

Answer 1

[Walter Roberson]

lsqcurvefit() only passes x and xdata to the function. It subtracts off ydata itself. The function needs to be of the form predicted_ydata = function NAME(x, xdata) . Implicit functions are not supported. You need to isolate the ydata out of the function. If that is not possible then you cannot use lsqcurvefit()

[Saravanakumar Dharmaraj]

Is it impossible to do curve fitting for transcendental equation in matlab?