Analysis & Visualization of a Kaggle provided dataset (https://www.kaggle.com/daveianhickey/2000-16-traffic-flow-england-scotland-wales) of traffic accidents in the United Kingdom.

The visualization components of this project will be heavily reliant upon Bokeh, an interactive visualization library based on d3.js. If you have the Anaconda distribution (recommended; https://anaconda.org) you can install Bokeh like so:

conda install bokeh

If not, a pip install should do the trick.

pip install bokeh

For the Choropleth mapping component of this project, I used geopandas (http://geopandas.org), a nice GIS layer library for Python. There are other options (e.g. Fiona, Pyshp) but I opted for this one due to familiarity.

conda install -c conda-forge geopandas

Or...

pip install geopandas

The UK_Traffic_Exploratory_Visuals.ipynb includes all of the processing steps as well as embedded interactive plots. The UK_Traffic_Choropleth_Map.ipynb excludes all the processing steps and renders the choropleth maps. Due to the data intensity, most of the interactive features take very long to run or time out all together for the choropleth maps.

There is also the ability to run Bokeh Server apps using the interactive_exp_visuals.py and interactivve_choropleth.py scripts.

To render the exploratory visual dashboard.

bokeh serve --show interactive_exp_visuals.py

To render the interactive chroopleth maps.

bokeh serve --show interactivve_choropleth.py