Error messages during using Genetic algorithm and lsqnonlin optimization functions

Currently,I am running optimization script file by using the outputs from the genetic algorithm function in the least square nonlinear function lsqnonlin() (as initial values for lsqnonlin()).When I run the script ,the following errors appear.So,what's the problem ?

Optimization terminated: maximum number of generations exceeded.
Error using snls (line 48)
Objective function is returning undefined values at initial
point. lsqnonlin cannot continue.

Error in lsqncommon (line 150)
    [xC,FVAL,LAMBDA,JACOB,EXITFLAG,OUTPUT,msgData]=...

Error in lsqnonlin (line 237)
[xCurrent,Resnorm,FVAL,EXITFLAG,OUTPUT,LAMBDA,JACOB] = ...

Error in frf (line 27)
[c,resnorm,residual,exitflag1,output1]=lsqnonlin(ff,c0);

Here is the code:

clc
clear all
% Create an anonymous function that describes the expected relationship
% between X and Y
f=@(c,x) exp((c(1).*(x(:,1).^c(2)))+(x(:,3)*c(4))+(x(:,2)*c(3)))+c(5).*(((x(:,2))./(x(:,3))-0.5).^2)+c(6).*(x(:,2))+c(7);
% data set 
% Specify x variables from data file,Re,Theta and Beta columns.
x=xlsread('all data for fitting friction');
% Specify y variable from data file ,(f)column.
y=x(:,4);
% objective function 
ff=@(c)(f(c,x)-y)./y;
% maximum of objective function
ffmax=@(c)norm(ff(c));
% Identifying population size and number of parameters for genetic
% algorithm
PopSz = 400;
Parms = 7;
% Modifying the options of genetic algotithm optimization function
opts = gaoptimset('PopulationSize',PopSz, 'InitialPopulation',randi(1E+4,PopSz,Parms)*1E-4, 'Generations',2e3,'StallGenLimit',inf, 'PlotFcn',@gaplotbestf, 'PlotInterval',1);
% Genetic algorithm application
[theta,fval,exitflag,output] = ga(ffmax, Parms, [],[],[],[],-Inf(Parms,1),Inf(Parms,1),[],[],opts);
% Specify a vector of starting conditions for the solvers(Least squre
% nnonlinear fit)
c0=[theta(1) ;theta(2);theta(3);theta(4);theta(5);theta(6);theta(7)];
% Perform a nonlinear regression
[c,resnorm,residual,exitflag1,output1]=lsqnonlin(ff,c0);
y1=f(c,x);
diff_f=y-y1;
Abs_diff_f=abs(diff_f);
perc_error=(Abs_diff_f./y);
x_line45=0:max(y);y_line45=0:max(y);
[max_error,I_max]=max(perc_error);
[min_error,I_min]=min(perc_error);
x_line_high=0:max(y); y_line_high=(max_error+1)*(0:max(y));
x_line_low=0:max(y); y_line_low=(1-max_error)*(0:max(y));
figure
plot(x_line45,y_line45,'k')
hold on
scatter(y,y1)
hold on
plot(x_line_high,y_line_high,'r')
hold on
plot(x_line_low,y_line_low,'r')
header=['fact','ffit'];
N=[y,y1];
disp(header)
disp(N)
disp('max. error=')
disp(max_error)

[Walter Roberson]

Are any of the y values 0?

[MOHAMED ABDULAZIM]

No.

[Star Strider]

I had no problems with what I believe are the same data in Nonlinear least-squares data fit.

Note that the documentation for R2013a is no longer available in Archived MathWorks Documentation, that being the reason I could not troubleshoot the reported problem.

Answer 1

[Matt J]

You have quite a few NaNs in your y data

>> nnz(isnan(y))

ans =

     7

>> nnz(isnan(ff(theta)))

ans =

     7

[MOHAMED ABDULAZIM]

Do you mean y data in the data file ?there is no missing data in y column in the data file,you can check that in the attached file .

[Matt J]

Did you run the commands that I posted above? DId you get different results?

[MOHAMED ABDULAZIM]

I ran the command on y data and gave me a zero value.

[Matt J]

Make sure you are using the same .xlsx file as the one you attached for us. The one you attached has 15 columns with lots of missing elements and different numbers of rows. xlsread is adding NaNs to pad the missing elements in the shorter columns.

>> x=xlsread('all data for fitting friction.xlsx');
>> whos x
  Name        Size            Bytes  Class     Attributes

  x         119x15            14280  double    

[MOHAMED ABDULAZIM]

Many thanks to you.I really appreciate your valuable notice that solved my problem.