How to fit a custom function using nlinfit?

I have written a function 'fun' in a separate .m file that I would like to fit to some xy data, using a parameter 'A' as a fitting parameter. However, since the function depends on the input of xdata and the parameter A to run, I am no sure how to fit it using nlinfit (or similar). When I enter the following, matlab returns an error:

    nlinfit(xdata,ydata,fun,A)
Error in fun(line 13)
Not enough input arguments.

So apparently, the xdata and A do not end up like this:

fun(xdata,A)

Does anybody know how to fit a function that was not defined using a function handle?

Thanks in advance!

[Image Analyst]

Please attach your data and tell us the model you wish to fit to it. In the meantime, I've attached 3 demos that perhaps can be adapted to your situation.

[Steven C]

Thanks for your quick reply and examples. However, these examples all use function handles (fun=@(x) x...) to define the function, whereas my question is how to fit a function that is defined in a separate .m file. Let's say this is the function I want to fit, which is defined in a separate .m file located in the same folder as the script I will use during the fitting. xdata and ydata are an example dataset to be fitted.

function [Tg]=fun_Tg(yw)
dCpw=1830;%
dCps=420.0067;%
Tg_w=134;%
Tg_s=420;%
Tg=(yw.*dCpw.*Tg_w+(1-yw).*dCps.*Tg_s)./(yw.*dCpw+(1-yw).*dCps);
disp(Tg)
end
xdata=[0.13491713;0.1762431;0.18508287;0.23878452];%example xdata
ydata=[341.105345;323.197005;317.909106;298.661164];%example ydata
Tg_s0=[420];%initial guess Tg_s

What I would now like to do is use fun_Tg in nlinfit, by fitting xy data with respect to Tg_s. However, I do not know how to present the data to nlinfit...