Add pre-computed features to the 34x3 computed by AAA so far.

[gthompson]

What is the best way to add pre-computed features to the 34x3 you used? As part of my automatic data QC, I compute many time series and spectral parameters, and they reside in Pickle files and also in CSV files.
The current method I use, where I have written featureFunctions to read precomputed features from files, works but is inefficient as it repeats the same numbers in every domain. I think I need to find where in the code the feature vector is sent to scikit-learn, and add external features to the vector there.
Marielle was confused, and suggested doing what I had already done.