Reorganise Repo

[stevehadd]

The current structure is based on a previous iteration of this project, and does not make sense for has been done in the work in the past 6 months.

code structure

• bin - executable scripts
• notebooks - for demos in jupyter notebooks
• xbt for the core xbt source code, such as dataset, classification, experiment
• platforms - for platform specific such such as the spice launch scripts

other tasks

• delete old unused
• print print source files
• add doc strings

[stevehadd]

This mostly done, we just need to do some tidy up and testing:

• open all notebooks and update import statements where needed
• run a few notebooks and check that they work
• run a few tests using run_cv_experiment script on Pangeo, SPICE and AzureML to check everying works
• add some readme files
• add some docstrings