Romberg



> quanc8(exp(-t)/sqrt(1-t*t),t,-0.999,0.999);
>                        3.922444732561162
> Thanks!
> But why romberg(...) and quanc8(...) with limits of integration -1..1 
> can't calculate this integral?

If your limits are -1 and 1, the numerical codes will probably
evaluate the integrand at t = 1 and get an infinity or other NaN.

Romberg has a hard time deciding that the integral is converging
because as it uses smaller intervals the integral still changes
rapidly due to the ends of the interval having a large contribution
that becomes more noticeable with smaller intervals.

-- 
Daniel Lakeland
dlakelan at street-artists.org
http://www.street-artists.org/~dlakelan