Christopher Jenness

When Newton-Raphson algorithm approaches convergence, often throws following warning: `RuntimeWarning: overflow encountered in exp` This is likely due to np.matmul(theta.T, X) >> 1, which in a sense is good.

If we implement the Sequential Minimal Optimization algorithm, we can remove our dependency on `cvxopt` https://en.wikipedia.org/wiki/Sequential_minimal_optimization

Handle noise assignments

2

DBCV is capable of handling noise assignments. This needs to be implemented.

HDBSCAN is inaccessible with pip, so conda is required. This is causing Travis CI issues: `The command "conda update --yes conda" failed and exited with 127 during .`

Not python 3 compatible

1

`exec` changed from a statement to a function in python 3. Unfortunately, the easy work around breaks the 6 char max.

If you make plots, start the project over, and make new plots, the thumbnail viewer shows the old plots. This should be fixed by adding `image-dired-refresh-thumb` at some point.

Currently only: - normal - uniform Common parametric distributions worth adding: - Poisson - Logistic - Beta - Gamma Any others?