adehabitatLT (version 0.3.28)

runsNAltraj: Highlighting the Patterns in Missing Values in Trajects

Description

runsNAltraj performs a runs test to detect any autocorrelation in the location of missing relocations, for each burst of an object of class ltraj.

summaryNAltraj returns a summary of the number and proportion of missing values for each burst of an object of class ltraj.

plotNAltraj plots the missing values in an object of class ltraj against the time.

Usage

runsNAltraj(x, nrep = 500, plotit = TRUE, ...) 
summaryNAltraj(x)
plotNAltraj(x, ...)

Value

runsNAltraj returns a list of objects of class

randtest (if a burst does not contain any missing value, the corresponding component is NULL).

Arguments

x

An object of class ltraj

nrep

Number of randomisations

plotit

logical. Whether the results should be plotted on a graph

...

Further arguments to be passed to the generic function plot

Author

Clement Calenge

Details

The statistics used here for the test is the number of runs in the sequence of relocations. For example, the sequence reloc-NA-NA-reloc-reloc-reloc-NA-NA-NA-reloc contains 5 runs, 3 runs of successful relocations and 2 runs of missing values. Under the hypothesis of random distribution of the missing values in the sequence, the theoretical expectation and standard deviation of the number of runs is known. The runs test is a randomization test that compares the standardized value of the number of runs (i.e. (value-expectation)/(standard deviation)) to the distribution of values obtained after randomizing the distribution of the NA in the sequence. Thus, a negative value of this standardized number of runs indicates that the missing values tend to be clustered together in the sequence.

See Also

ltraj for additional information about objects of class ltraj, setNA for additional information about missing values in such objects

Examples

## Two relocations are theoretically separated by
## 10 minutes (600 seconds)
data(puechcirc)
puechcirc

## plot the missing values
plotNAltraj(puechcirc)


## Test for an autocorrelation pattern in the missing values
(runsNAltraj(puechcirc))