There are better (faster, more efficient) ways to do this, F# numerics library for a start, but this is at least interesting. This snippet uses the polar form of the Box-Muller method to generate Normal- (Gaussian-) distributed random numbers as an infinite sequence. The polar form is more efficient than the basic form as it does not rely on trigonometric function calls, but there are far more efficient alogrithms (read harder to implement) e.g. the Ziggurat method (for a later post).
2 people like thisPosted: 8 years ago by Kevin Roche
The random hyperharmonic series is the infinite series S = Sum(1,inf,d(i)/i^pow), where integer pow is greater than 1, and d(i) are independent, identically distributed random variables with property P(d(i)=0) = P(d(i)=1) = 0.5. Cumulative function F(x) = P(S < x) for even powers can be build by combination of analytical and numerical computations.
5 people like thisPosted: 2 years ago by Pavel Tatarintsev