joutlier: Recursive outlier detector/remover

In jrevenaugh/TSAUMN: Time Series Analysis Tools Used in ESCI5201 at the University of Minnesota

Description

Locate and remove outliers using an arima model to whiten time series, Hampel's MAD criterion to identify outliers, and kalman smoothing. Algorithm is recursive and continues until no further outliers are found.

Usage

joutlier(s, order = c(2, 0, 0), k = 10, tol = 3)

Arguments

s	timeseries
order	order specification of arima model; defaults to AR(2)
k	hampel's half-window size. Defines window for MAD computation
tol	MAD outlier threshold; defaults to 3 (Pearson's rule)

Value

cleaned timeseries.

Examples

y <- arimaSim( 200, ar = c( 0.5, -0.1 ) )
# Add a few outliers
o <- floor( runif( 5, min = 5, max = 195 ) )
y[o] <- runif( 5, min = 4, max = 8 )
yc <- joutlier( y )
plot( y )
lines( yc, col = "red" )

jrevenaugh/TSAUMN documentation built on Nov. 8, 2019, 2:20 p.m.