Hi Edgar
There are a number of ways, which I list below.
The first uses the 'equi' flag, which tells Chebfun to use a
certain rational interpolant through the (assumed) equally
spaced data.
The second uses polynomial interpolation through equally
spaced points (which is usually a bad idea).
The third uses cubic splines. The downside here is that
the interpolant is not global (and convergence will probably
be slower than 'equi' as n increases).
n = 9;
t = linspace(-1,1,n)';
v = 1./(1+25*t.^2);
plot(t, v, '.', 'markersize', 20), hold on
f = chebfun(v, 'equi'); plot(f)
g = chebfun.interp1(t,v); plot(g),
h = chebfun.spline(t, v); plot(h),
hold off