anomalies: Anomaly Creation

In tsaux: Time Series Forecasting Auxiliary Functions

Anomaly Creation

Description

Creates specific types of anomalies given a series.

Usage

additive_outlier(y, time = 1, parameter = 0.5, add = TRUE)

temporary_change(y, time = 1, parameter = 0.5, alpha = 0.7, add = TRUE)

level_shift(y, time = 1, parameter = 0.5, add = TRUE)

Arguments

y	a univariate xts object or numeric series.
time	the time index at which the anomaly takes place.
parameter	the coefficient on the anomaly (the percent of the value of y at the specified time index representing the anomaly).
add	whether to contaminate the series (add the anomaly to the series) else will return a matrix with the anomaly (without the effect of the parameter).
alpha	the AR(1) coefficient for the temporary change which determines how quickly the effect decays.

Details

These functions allow the generation of anomalies and may be chained together.

Value

Either the contaminated series else a matrix of the anomaly.

Author(s)

Alexios Galanos for this wrapper function.

tsaux documentation built on April 4, 2025, 3:08 a.m.