MannKendallSeg: Apply MannKendall test for segments of a time series
In greenbrown: Land Surface Phenology and Trend Analysis

Description Usage Arguments Value Author(s) See Also Examples

The function call MannKendall for different segments of a time series

1 2	MannKendallSeg(Yt, seg = NULL, bp = NoBP())

`Yt`	univariate time series of class `ts`
`seg`	a vector indicating segments of a time series. If NULL, provide `bp`
`bp`	detected breakpoints in the time series as returned by `breakpoints`

The function returns a data.frame with the estimated Mann-Kendall tau and p-value for each segment of the time series.

Matthias Forkel <matthias.forkel@tu-dresden.de> [aut, cre]

MannKendall

# aggregate time series to annual time steps
ndvi <- aggregate(ndvi, FUN=mean)
plot(ndvi)

# MannKendall test for different segments
seg <- c(rep(1, length(start(ndvi)[1]:1995)), rep(2, length(1996:end(ndvi)[1])))
mk <- MannKendallSeg(ndvi, seg)
mk

# MannKendall test for segments that are defined by breakpoints
bp <- breakpoints(ndvi ~ time(ndvi))
mk <- MannKendallSeg(ndvi, bp=bp)
mk

# MannKendall test without segments
MannKendallSeg(ndvi)