The Shiny Developer Series continues our spotlight on the COVID-19 Test Spotter application created by renowned data scientist and entrepreneur Mike Thomas! In this episode we learn how Mike implemented robust database techniques and logic to manage user-supplied data, navigating the use of multiple mapping APIs, and achieving a practical development-to-production strategy for releasing the application to the public. Plus we have a great discussion on what production means in the space of Shiny development, and much more.
Resources mentioned in the episode
Episode Timestamps
00:00:00 Episode Introduction
00:00:46 CRUD workflow and flowchart, with MongoDB database updates
00:09:41 Diving into the Google Maps and Places API
00:15:55 Geographic fencing, and dealing with unexpected user behavior.
00:22:35 Deploying with a dev and production branch.
00:28:15 Future plans for the app
00:30:50 Advice for building Shiny apps in production
00:31:20 Production apps is all about reproducibility: OS, dependencies, and the code.
00:33:15 Pay attention to warning messages in your console.
00:34:22 Modularization and Golem
00:38:08 You can have too much reactivity, such as overextending uiOutput() / renderUI()
00:40:10 Episode wrapup
shinydevseries::session_info()
📽 Find previous interviews at
https://shinydevseries.com📋 Feedback is more than welcome! Please send your comments or suggestions on our contact form at
shinydevseries.com/contactIf you enjoyed this video or got value from it, then consider dropping something in