Hi all, I was wondering to know if there is any function of adehabitat or some other R package that can perform a network based kernel density estimation.
I work with otters, that usually have linear home ranges (along rivers), so I'm particularly interested in the possibility to use a kernel less "generalist". So far, the network kernel estimators I've found are: - the toolbox SANET for ArcGIS from professor Okabe & colaborators (http://sanet.csis.u-tokyo.ac.jp/index.html) - the one based on Delauney triangulations proposed by Downs & Horner (2007), applied to ecology, that, I think, uses TransCAD - the one, performed in PostGIS, a Pyton script and ArcGIS, from Thimotée Produit et al (2010) for architecture purpose. So I was thinking that it could be very useful to have such functionality also on a Open Source software (if it does not exist already, of course). It could be very interesting for all of us who work on otters as well as fishes and so on! Thanks in advance for any feedback, help and suggestion on the issue. Cheers, Lorenzo P.S. I've also tried the LoCoH home range generator from Gets, that is considered more indicate in these situations, but it still doesn't give me the best results (disjoined contours and polygons that still include big portions of areas actually not "used" by otters).
_______________________________________________ AniMov mailing list [email protected] http://lists.faunalia.it/cgi-bin/mailman/listinfo/animov
