Assuming you have a 2D fit, it is returned as a cfit
object. cfit
can be directly evaluated by the feval
function. Let's say you want to plot a smoother fit than the original data. You can do something as follows to evaluate the returned fit:
tf_ = [t(1):((t(end)-t(1))/1000):t(end)]; % 1000 will make it smooth and dense. Pick a number that suits you.
yf_ = feval(cf_, tf_); % Evaluate the fit
plot(handles.axes_tau, tf_, yf_, 'k-');
Since the fit object is just a set of coefficients that describes the analytical function you are fitting, you can sample it as coarsely or as finely as you would like. Pick a set of sampling points that give you a smooth looking curve on the screen. In this example, I just made 1001 samples from t(0)
to t(end)
.