This R package focuses on tools for detecting anomaly on a time series, using Bayesian contextual anomaly detection.
It has three main goals:
-
Detect anomaly on time series
-
Indentify the most important features needed for data analysis
-
Provide automatic analysis that makes them easy to use from R
-
.data: data table to be used
-
.scope: training period, testing period, and KPI names
-
.setting: algorithm setting
-
.plans: combination of .scope and .setting
-
.prior: prior values for model and formula tables for i.v. and d.v.
-
.result
- Report for each analysis e.g.) check_model