tsOutliers: Identification of statistical outliers in time series

Description Usage Arguments Value Author(s) See Also Examples

View source: R/tsOutliers.R

Description

This function identifies statistical outliers in a ts object based on upper and lower quantile criteria. The function body is mainly taken from http://stats.stackexchange.com/questions/1142/simple-algorithm-for-online-outlier-detection-of-a-generic-time-series.

Usage

1
2
tsOutliers(x, lower_quantile = 0.2, upper_quantile = 0.8, plot = FALSE,
  index = FALSE, ...)

Arguments

x

Numeric. A vector of observed time-series values.

lower_quantile

Numeric, default is 0.2. The lower quantile limit.

upper_quantile

Numeric, default is 0.8. The upper quantile limit.

plot

Logical, default is FALSE. If TRUE, a time-series plot including identified outliers is generated.

index

Logical, default is FALSE. If TRUE, a vector holding the indices of identified outliers is returned rather than the statistically obtained scores for each measured value.

...

Additional arguments passed to ts.

Value

A numeric vector of scores or, if index = TRUE, a vector holding the indices of identified outliers.

Author(s)

Florian Detsch

See Also

ts

Examples

1
2
3
4
5
6
7
# Random time-series values
set.seed(10)
x <- rnorm(100, 0, 2)

# Return indices of outliers incl. visualization
tsOutliers(x, lower_quantile = .35, upper_quantile = .7,
           plot = TRUE, index = TRUE)

environmentalinformatics-marburg/eimarGsodTools documentation built on May 13, 2017, 6:11 p.m.