StundentT parameter beta parameter update may be wrong

[PedroPizarro]

I was using the library and I may stumbled upon some issue.

The file `distribution.py, on line 60-65, has the update equation for the beta hyper-parameter of the predictive posterior. However, I think it might be wrong.

self.betaT = np.concatenate( [ self.beta0, (self.kappaT + (self.kappaT * (x - self.muT) ** 2) / (2 * (self.kappaT + 1))), ]

From your suggested bibliography, hyper-parameter equation for the Beta parameter is the following:

Your code shows a sum with self.kappaT (the first tem fo the second concatenate item) where it should be the self.betaT ifself.
Thus, IMHO the correct formula would be:

self.betaT = np.concatenate( [ self.beta0, (self.betaT + (self.kappaT * (x - self.muT) ** 2) / (2 * (self.kappaT + 1))), ]

[PedroPizarro]

Adding the more appropriate part of the text: