Building a fully custom container for R

R is a popular language for data exploration and analysis. In this example, we're going to build a custom container to train and deploy a linear regression model on the Boston Housing dataset.

The overall process is similar to building a custom container for Python. Instead of using Flask to build our prediction API, we'll use plumber (https://www.rplumber.io).

Coding with R and plumber

Don't worry if you're not familiar with R. This is a really simple example, and I'm sure you'll be able to follow along:

1. We write a function to train our model. It loads the hyperparameters and the dataset from the conventional paths. It normalizes the dataset if we requested it:

   # train_function.R
   library("rjson")
   train <- function() {
       hp <- fromJSON(file = 
             '/opt/ml/input/config/hyperparameters.json')
   ...