Curve fitting in a for loop

I have a system of difference equations expressed in a for loop. I have experimental data that I would like to fit some parameters to. Could anyone provide some insight into how to proceed. I've experimented with lsqnonlin, lsqcurvefit, and fminsearch to no avail.

parameters to fit: c(1,2,3)

parameters to fix: b,d,KX1,KX2,KX3

system:

for t=1:9
          CtlA(t)=X1(t) + X2(t) + X3(t);
          X1(t+1)=b*X3(t);
          X2(t+1)=X1(t)/((1+c(1)*CtlA(t))*KX1*d);
          X3(t+1)=(X2(t)/((1+c(2)*CtlA(t))*KX2*d))+(X3(t)/((1+c(3)*CtlA(t))*KX3*d));
end

Any help would be gladly appreciated!

[Andrew Newell]

This for loop seems to have X1, X2 and X3 as both input and output. Are they the data? Are you trying to fit them to some function of t?

[Andrew Newell]

The parameter d is redundant because it is always multiplied by some other parameter. You could set it to 1.