[R-sig-Geo] adehabitatLT | Patterns in Missing Values in Trajects

[Calenge Clement]

Dear Jefferson,

> I'm using adehabitatLT for the first time analyzing jaguar movements in
> amazon. I'm running runsNAltraj to detect any autocorrelation in the
> location of missing relocations. But I'm a beginner in this task and I'm
> not knowing to how to interpret the graph generated by this function.

This function tests whether the missing values are randomly distributed
in the trajectories using a runs test. See
http://en.wikipedia.org/wiki/Wald%E2%80%93Wolfowitz_runs_test for an
explanation. Using the same notations as on this wikipedia page, this
test calculates the criterion:

C=(N-mu)/sigma

Where N is the number of runs in the sequence, mu and sigma are
respectivel the expected mean number of runs and the expected standard
deviation of this number of runs  under the hypothesis of a random
distribution of the missing values. The criterion C takes a high
positive value if the missing values are more regular than expected
under H0, and a highly negative value if the missing values are clustered.

The histogram corresponds to the statistical distribution of C expected
under the null hypothesis of a random distribution of the missing
values. The "arrow" indicates the actual value of the criterion. This
graph allows to see if the actual value is far from the distribution
expected under the null hypothesis. A p-value is derived from this
graph, and is returned by the function,
Hope this helps,

Clément Calenge

-- 
Clément CALENGE
Cellule d'appui à l'analyse de données
Direction des Etudes et de la Recherche
Office national de la chasse et de la faune sauvage
Saint Benoist - 78610 Auffargis
tel. (33) 01.30.46.54.14