Exercise

Train/test split

In order to test your models, you need to build and test the model on two different parts of the data - otherwise, it's like cheating on an exam (as you already know the answers).

The data split is an integral part of the modeling process. You will dive into this by splitting the diabetes data and confirming the split proportions.

The diabetes data from the last exercise is pre-loaded in your workspace.

Instructions 1/3

undefined XP
    1
    2
    3
  • Split the diabetes tibble into diabetes_split, a split of 80% training and 20% test data.
  • Print the resulting object.