This is some example data analysis/graphing stuff I threw together as an example of building out skills. WMATA has an extensive repository of ridership data though you’ll often get it in a somewhat annoying format. (The “CSV” is really a UTF-16 TSV, and has an extra row with titles that could have been incorporated into the main set of headers) I also added a ‘location’ based on the route names.

Please enjoy both a static graph, and a mostly related (if rather hacky) plotly dashboard (caution there’s something weird going on so currently firefox doesn’t see it as a real thing to connect to):

Bus Ridership Time Series Breakdown Overall Bus Ridership Time Series

Why the focus on fare evasion? It’s easy to show with the available data, and as someone who has and uses the bus transit here, while being far from a metro stop, I have a strong personal interest in the system working well!

The Python for the graph, collection of WMATA data it uses, code for the dashboard, and collection of WMATA data that it also uses are available.

Building out actual machine learning skills (most likely domain related) is in the works. But I can in fact generalize what I have to some data analysis and creating simple dashboards! If there’s interest in expanding this or say putting together an alternate demo with observatory/stellar/exoplanet/survey simulation datasets, feel free to contact me.