This is a helper function. It takes an object
created by 'bayesian_test' and spits intervals
with evidence for either the null or alternative models.
Usage
interpret_bayes(bfs, evidence_thresh = 10, cluster_size = 10)
Arguments
- bfs
An object returned by 'bayesian_test'.
- evidence_thresh
Threshold to judge evidence. Defaults
to 10. It is meant to be symmetrical: values above 10
indicate support for the alternative Hp, less than 1/10 for the null.
- cluster_size
How many timepoints constitute a cluster?
Value
A list including intervals that support H0 or H1.