Bike Prediction

vetiver
quarto
scheduled ETL
pins
plumber
shiny
database
vetiver
internal package

Bike predict workflow diagram showing scheduled Quarto document updating a pinned dataset, pinned model, and a database. The Model and pin are then referenced by a vetiver-deployed API, which is used by production and development Shiny apps. All content is deployed to Connect. A private package served by Package Manager is used.

Bike predict workflow diagram

This app provides real-time predictions of the number of bikes that will be available at the stations of Washington DC’s docked bike share, Capital Bikeshare. Creating and visualizing those predictions takes advantage of many different types of R content and the ability to deploy them on Connect.

Background

Like many metropolitan areas, the Washington DC area has bicycles available for short-term rental to commuters and tourists at docks around the city. The number of bikes at each of these docks ebbs and flows throughout the day, and the Capital Bikeshare program provides real-time data on the number of bikes available at each dock via an API.

Goal

The goal of this app is to provide a prediction of the number of bikes at a station in the near future based on real-time streaming data from an API. These predictions will be made available in a web application and to other services via an API. Furthermore, continuous monitoring of the predictions will be enabled to watch for model drift and changes in goodness-of-fit over time.

The source code and more detail on this project can be found on the Github Repository.

Back to top