Variational Bayes lecture: Check correctness "EM-algorithm for the GMM" code

[fonsp]

When converting the lecture from Jupyter I noticed some strange Julia code. It looks like the intention was to modify the π_hat variable, but the function does not actually do that.

So I suspect that the code does not match the math, although the result looks correct. Can you verify this, and modify the code to resolve the comments that I wrote?

What is the π_hat variable? Can we just remove it?

[ThijsJenneskens]

π_hat is a vector of 2 elements that contains the fractions of data that fall into both classes. It is not being updated because the code overwrites the vector itself, not the data in it; I fixed it by adding [:] after π_hat in the line where it's updated. Pull request coming soon TM