“Applied” Tidy modeling

Josiah Parry updated: 2019-09-12

Intro to Tidy modeling

• Prediction API
• Prediction App

This repository contains the resources used for a brief (~1hr) introduction to tidymodels.

The slides xaringan.Rmd very, very briefly introduce the fundamentals of the tidymodels (parsnip, recipes, and rsample).

During the slides, I walk through audio_classifier.R. This creates a lyric classifier as outlined in mirr.

This model is then “productionalized” with plumber (plumber.R), and then wrapped in a small Shiny app (app.R).

Important notes

In order to run the prediction plumber API and Shiny App, a Spotify API key will be needed. The training data was collected with the spotifyr package. Refer to the spotifyr README for insstructions. I stored the credentials in an .Renviron file.