Regarding Seraj's solution, i believe the solution would be
$ \sqrt(A) cos(2pi*drand48()) $
where A is still the exponential RV calculated in 1b.
- Yes you are correct, I forgot the 2pi factor to change the range of the uniform random variable from [0-1] to [1-2pi].
thank you for pointing it out.
Seraj
