//
sign in
Post
by @danabra.mov
PostEmbed
by @danabra.mov
Record
by @jimpick.com
Record
by @atsui.org
+ new component
Post
In the interesting range, the integrand is always monotonic, so numerical integration is (or should be) straightforward. I have to admit that in zipfR I need finite gamma integrals, which I currently compute as diff of incomplete gamma function for end points … exactly, catastrophic cancellation.