This analysis features a sample anonymized micromobility dataset which I used to delve into supply and demand measurement and data visualization. Here are some highlights of what I hope to illustrate with this analysis:
- Data prep is paramount. It is important to always do a preliminary analysis of the data to identify and remove outliers, find areas of bias, discover hidden assumptions, and evaluate a suitable approach for a given problem.
- Tableau is awesome. In this analysis, I used Python as my main data cleaning, pivoting, joining, and validation tool. I then took that processed data and used Tableau as a sandbox for visualizing the data in different ways. Using tools like Tableau that enable calculated fields, complex joins, geographic mapping, etc. I was able to efficiently product actionable insights.
- This process can be repeated for any dataset whether in micromobility, healthcare, IoT, logistics, etc. There are endless arrays of data overlays to enrich analyses and bring new insights to the table. Also key to note is that as assumptions change, the interpretation of results will also change. Many assumptions stem from customer behavior which should continually be assessed and studied.
Here are some details about the data at hand:
- San Francisco, CA city-specific dataset
- Mobile app sessions approximate demand
- Vehicles are e-bikes and/or e-scooters
- Vehicles are dockless and can be parked anywhere within service areas when not in use
See the full analysis below and see the Python Jupyter Notebook here:
Mari Bolton 