Feature/graph analysis ipyleaflet visualisation

[rdnfn]

Introducing GeoGraphViewer 🎉 ... this PR adds the GeoGraphViewer visualisation class and a bunch of code that enables it. As the GeoGraphViewer is a fairly complex widget, this PR is somewhat lengthy, apologies in advance! Thanks a lot for your review!

Main changes:

• All GeoGraph visualisation code now lives in visualisation/*. The old models/visualisation.py has been removed.
• Added GeoGraphViewer class (in visualisation/geoviewer.py): this allows us to visualise GeoGraphs interactively.
• Added control widgets (in visualisation/control_widgets.py): these make up the side control panel seen in demos. See architecture image below for a rough outline how they relate and are added to the GeoGraphViewer.

Other changes:

• Added different utils that are used by GeoGraphViewer. These are in
  • visualisation/widget_utils.py
  • visualisation/folium_utils.py (old functions from visualisation.py, no need to review)
  • visualisation/graph_utils.py (old functions from visualisation.py, no need to review)
• Added a new test case to data_loading/test_data.py
• Added visualisation/style.py that contains constants defining the plotting style of graph, etc. in GeoGraphViewer.
• Added ipyleaflet and jupyter lab dependencies in requirements/dev-requirements.txt
• Added misc exploratory notebooks that helped me develop the viewer, for reviewers only notebooks/exploratory/rdnfn-8-polesia-geographviewer-demo.ipynb will likely be interesting, which is the demo on the Polesia data. Feel free to use it, it should work with the environment created via make env (if you updated it to new requirements)
• The old folium code lives on without further development in the FoliumGeoGraphViewer class (in geoviewer.py)
• Updated readme to new visualisation folder

Architecture

GeoGraphViewer demo

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[Croydon-Brixton]

Love it already before reviewing! 😉 😍

[rdnfn]

Thanks a lot for the review @herbiebradley! Was able to address most points in 42eae9e, and created Issue #50 for the remaining performance improvements, to be dealt with at a later point.