split.qmd

---
title: "Split"
subtitle: "Split data into training (to fit) and test (to evaluate prediction)"
---

Data is often split so that \~20% of the observations (presence and absence) are set aside from the model fitting to be used for model evaluation.

The `k-fold` function is often used to split the data into k groups, and then the model is fit k times, each time using a different group as the test data and the remaining groups as the training data.