R packages - Feb 2011 - adehabitatMA, LT, HR and HS version 0.1

Dear all,

I have just uploaded 4 new packages on CRAN, which are on the long term 
designed to replace the "old" package adehabitat:

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

I will still continue to maintain the "old" adehabitat, but on the 
long-term, adehabitat will be replaced by these four packages.

Detailed justification for the development of these packages is given below.

=====================================================================
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

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

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:
vignette("packagename")

* the home range estimation methods have been homogenized and return 
classes compliant with the classes of the package sp:
- the functions kernelUD, kernelbb, BRB and kernelkc all return objects 
of class "estUDm", which are lists of objects of class
"estUD". The
class estUD 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 Cavallini, on the list has been added to adehabitatHR, and 
returns an object of class MCHu

* the method of biased random bridges (Benhamou, 2011) has be added to 
adehabitatHR, to estimate the utilization distribution from a trajectory.

* 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;

* the function rasterize.ltraj allows to rasterize a trajectory (i.e. 
useful to identify the habitat characteristics of the steps building the 
trajectory)

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

* Two additional functions dl and ld to convert efficiently the class 
ltraj to and from data frames (thanks to Mathieu Basille for the 
suggestion).

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 six years. Only the 
input and output of the functions have been changed.

============================================================================
Happy testing,


Cl?ment Calenge
-- 
Cl?ment CALENGE
Cellule d'appui ? l'analyse de donn?es
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

_______________________________________________
R-packages mailing list
R-packages at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-packages