betafit_fast

Quickly compute the parameters of a beta distribution

This function does fast, maximum-likelihood fitting of beta distributions to data in the range (0,1). It differs from the betafit() function provided in MATLAB's Statistics and Machine Learning toolbox as follows:
1. It doesn't require any toolboxes.
2. It is faster.
3. It is usually more accurate (i.e. results in a higher likelihood).
4. It can fit multiple distributions in parallel.
5. It cannot handle data values of 0 or 1.
6. It doesn't provide confidence intervals.