@Studentmath, I don't see why they both with saying $x>\epsilon n/2 + \epsilon^2 n/10$. The latter term can be absorbed into the first term by making $\epsilon$ slightly bigger.
You should ask such questions more often, @Studentmath. :P
@TedShifrin It's not very hard, I don't think, I'm looking for a subset of $\mathbb{R}$^2 that projects to [0,1], but such that the projection map restricted to that subset is open, but not closed. What should I be thinking about?
@MikeMiller Gak(L/K) fixes all the n-th roots of unities in L pointwise, by definition of galois groups. but i am still thinking how thats supposed to help us.
I'm skipping various things along the way. Certainly don't need to do all possible discrete probability distributions. Will do binomial, Poisson, maybe hypergeometric.
Complex Analysis: Given a function f(x) that is analytic in some domain D in the complex plane, if the conjugate function is conj( f(x) ) = c^2/f(x) then it is analytic. I read in a Churchill that this is true. Why does it follow that the conjugate must be analytic?