both the pyprep and the autoreject package have an implementation of RANSAC.

We should make a comparison to:

1. see whether they provide the same results
2. potentially find bugs in one of the implementation
3. check which implementation is more computationally efficient

After the comparison we could decide to drop one implementation and only use the "better" one.

Or we keep using both implementations, but make an example that explicitly compares them, pointing out pros and cons in terms of features.

In any case, this would be an important enterprise for point 1 and 2 in my list alone.