[AniMov] The future of adehabitat

Clément Calenge clement.calenge at oncfs.gouv.fr
Sun Sep 5 17:27:10 CEST 2010

Dear all,

I have made major changes to the package adehabitat. Indeed, the R 
environment has changed a lot since I began the development of 
adehabitat in 2002 (development of namespace, etc.), and new classes and 
efficient methods have been developed to deal with spatial data (package 
sp). In addition, the number of functions available in the package has 
grown to more than 250 functions, implementing methods for habitat 
selection analysis, home range estimation, animal movement analysis, or 
spatial operations.

Therefore, I decided to:
(i) rewrite the package adehabitat to make it more compliant with these 
evolutions of the package R,
(ii) split adehabitat into four packages
(iii) create a new project on R-Forge (https://r-forge.r-project.org/) 
to store the sources of these four packages.

The four new packages are:

* adehabitatMA: functions to perform spatial operations (morphology, 
buffer, etc.)
* adehabitatHS: functions for the analysis of habitat selection
* adehabitatHR: functions for home range estimation
* adehabitatLT: functions for movement analysis

The home page of the project is: http://adehabitat.r-forge.r-project.org/
The packages can be installed in R by typing:


I will keep these packages on R-forge for some time (at least several 
months), before submitting them to CRAN, to take into account users 
advices on these packages. After their submission to CRAN, I will 
continue to maintain the old adehabitat on CRAN for some time, but on 
the long term, this package will disappear. These four packages are 
expected to become the future of adehabitat.

I now describe several major changes:

* the functions of the packages are documented precisely in a vignette 
(there is one vignette per package). Both their use and the theory 
underlying these functions are described there. To access it, type:

* the home range estimation methods have been homogenized and return 
classes compliant with the classes of the package sp:
- the functions kernelUD, kernelbb, and kernelkc all return objects of 
class "estUDm", which are lists of objects of class "estUD". The class 
extUD extends the class SpatialPixelsDataFrame.
- the function clusthr and LoCoH return objects of class "MCHu", which 
are lists of SpatialPolygonsDataFrame
- the function mcp returns a SpatialPolygonsDataFrame

* home range estimation methods now take objects of class SpatialPoints 
as arguments

* objects of class "ltraj" are now characterized by an additional 
attribute "infolocs", which is designed to store metadata on the 
trajectories (e.g. precision on the relocations). Most functions of the 
package adehabitatLT can be used to analyse these metadata (plotltr, 
etc. see the vignette).

* the method of characteristic hulls (Downs and Horner 2009), suggested 
by Paolo, on the list has been added to adehabitatHR, and returns an 
object of class MCHu

* the canonical OMI analysis, allowing exploration of habitat selection 
with radio-tracking data, has been added to the package adehabitatHS

* the autocorrelation functions described by Dray et al. (2010) for the 
analysis of movement have been added to the package adehabitatLT;

* all the packages have a namespace for management of internal functions.

Note that the calculations performed by most functions of adehabitat 
have not changed (e.g. the algorithm implemented in kernelUD of 
adehabitat is the same as the algorithm implemented in the function 
kernelUD of the package adehabitatHR), since they have been deeply 
discussed with users and corrected during the last five years. Only the 
input and output of the functions have been changed.

Happy testing,

Clément Calenge.

Cellule d'appui à l'analyse de données
Office national de la chasse et de la faune sauvage
Saint Benoist - 78610 Auffargis
tel. (33)

Cellule d'appui à l'analyse de données
Office national de la chasse et de la faune sauvage
Saint Benoist - 78610 Auffargis
tel. (33)

More information about the AniMov mailing list