META: Create example notebooks

[isaacaka]

The overall aim of the example notebooks is to showcase examples of using different dimensionality reduction techniques and forecasting methods on NEMO ocean model data. Allow users to gain a bit more detailed understanding of the forecasting pipeline and provide a basic outline for users who want to experiment and see how well their techniques do.

The next few steps of the plan for refactoring the jupyter notebook so that some of its code can form the basis of the different example notebooks is:

1. Refactor the plotting code into functions in plotting_utils.py Functionise plotting code in Notebook #93
2. Refactor the decomposition, forecasting and reconstruction code into functions in some_suitable_name.py Functionise decomposition, forecasting and reconstruction code #95
   First two steps are focussed on basic functionisation of the code.
3. Decompose functions from step 1 and 2 so that they are single reponsibility and fully parametrize them
4. Create additional functions as needed
5. Figure out a suitable structure for the PCA+GP example notebook and implement it using the functions created in steps 1-4