Sequential Change-Point Detection via Nonparametric Inference

Documentation for package ‘scanr’ version 0.1.1

covering_metric	Segment covering metric
cpd_metrics	Combined change-point accuracy metrics
default_window_sizes	Choose default scan window sizes
f1_score_cpd	Tolerant F1 score for change-point detection
ipm_statistic	Integral probability metric statistic
match_change_points	Match true and estimated change points
one_wasserstein_distance	One-dimensional Wasserstein distance
precision_recall_cpd	Tolerant precision and recall for change-point detection
scan_cpd	Detect change points in a univariate time series
scan_single_window	Run SCAN for one window size
swal_statistic	Local SCAN/Wasserstein split statistic
ts_cusum	Localize a mean change with a CUSUM statistic
ts_wasserstein	Localize a distributional change with a Wasserstein statistic
vis_change_points	Visualize detected change points from a scan result
vis_swal_curve	Visualize the SWAL/Wasserstein localization curve for a region
vis_thresholds	Visualize scan statistics and bootstrap thresholds
vis_time_series	Visualize a time series with optional change-point markers
vis_vote_scree	Visualize retained change-point count by voting threshold
vis_window_votes	Visualize ensemble vote counts for candidate change points