Box-Muller implementation - algorithm taken from https://en.wikipedia.org/wiki/Box%E2%80%93Muller_transform
Calling 'generate' repeatedly will return a normally distributed set of numbers from a uniformly distributed set of numbers.
Demonstration found in com.michaelmiklavcic.BoxMullerTest. 'BoxMuller' provides a generate method for creating the samples. 'Distribution' provides a draw method for creating a histogram visualization.
Args:
numSamples - How many samples to generate
mu - mean
sigma - standard deviation
BoxMuller.generate(10000, 5, 1);
Args:
samples - Unsorted data samples to produce the histogram from
maxHeight - Maximum height of the histogram
type - Method for calculating number of bins. Currently only SQUARE_ROOT is supported.
Distribution.draw(samples, 30, BinType.SQUARE_ROOT);