Computes averaged covering and F1 scores against multiple annotation sets,
as used in the Turing Change Point Dataset benchmark.
Usage
cpt_metrics_annotated(pred, annotations, n, margin = 5)
Arguments
- pred
Predicted changepoint indices.
- annotations
A list of ground-truth annotation vectors.
- n
Length of the series.
- margin
Tolerance margin (default 5).
Value
A tibble with averaged metrics.
