Hi guys, I am new here and I have no idea in which category to put this.
I am doing my master thesis about home ranges, and the datas I am using come from GPS points of 8 different individuals.
I am using the package adehabitatHR, and I used the smoothing parameter h=href at first. It gives me huge, circular home ranges. A friend told me I can change the smoothing parameter to a fraction of href (0.8*href for instance) to fit my data better.
However, I have no idea to do that.
Here is my original script:
#############################################################################################
Calcul des domaines vitaux des Micromys de Greny, Automne 2017
Utilisation de la methode kernel
Fixed kernel with Href smoothing, grid = 80 and extent = 2, contour level = 80%
######################################################################
Packages utiles: sp/rgeos/maptools pour graphique et adehabitatHR pour DV
library(sp)
library(rgeos)
library(maptools)
library(adehabitatHR)
library(readxl)
Dossier de travail
setwd("C:/Users/Fanny/Documents/ArcGIS/Master_thesis/RStudio/Coordonnées RStudio/Lambert veg")
Fichier des localisations
loc <- read_excel("C:/Users/Fanny/Documents/ArcGIS/Master_thesis/RStudio/Coordonnées RStudio/Lambert veg/Radiotracking_all.xlsx")
head(loc)
Transformation en format SpatialPoints pour pouvoir utiliser les fonction de adehabitatHR
Attention adapter les numeros de colonne en fonction du tableau !
coor <- loc[,c(8,9)]
head(coor)
coordinates(loc) <- coor
class(coor)
Choosing the smoothing paremeter method : h="href" => reference de bandwidth
h="href"
Grid Initialisation to estimate the utilization distribution (UD)
Grid resolution
grid = 200
Taille de la grille
extent = 2
Estimation of the UD : the surface is in m2 (Remember that the first column of the component "loc" of this dataset contains the identity of the animals)
kud <- kernelUD(loc[,3], h, grid,extent)
head(kud)
image(kud)
Estimation du DV Ã partir de l'UD
Choosing the threshold probability
hrs <- kernel.area(kud, percent=seq(50, 95, by=5), unin = "m", unout= "m2")
plot(hrs)
seuil = 80
Estimation sous forme de couche vectorielle
dv <- getverticeshr(kud, percent = seuil, unin = "m", unout="m2")
class(dv)
head(as.data.frame(dv))
plot(dv)
Export DV
setwd("C:/Users/Fanny/Documents/ArcGIS/Master_thesis")
writePolyShape(dv, paste("dv_kernel_all","",h,"",seuil))
Export des DV en csv
setwd("C:/Users/Fanny/Documents/ArcGIS/Master_thesis")
write.table(as.data.frame(dv), paste("dv_kernel_all","_",h,".csv"), row.names=FALSE, sep=";",dec=",", na=" ")
Here are some information my friend gave me but I have no idea how to apply them:
to calculate each hrefs (?):
hrefs <- c(785.0429, 638.4495, 774.7266, 870.8454, 1009.547, 1239.997)
hrefs0.9 <- hrefs0.9
hrefs0.8 <- hrefs0.8
hrefs0.7 <- hrefs0.7
hrefs0.6 <- hrefs0.6
hrefs0.5 <- hrefs0.5
hrefs0.4 <- hrefs0.4
hrefs0.3 <- hrefs0.3
hrefs0.2 <- hrefs0.2
hrefs0.1 <- hrefs*0.1
To precise which individual (?):
al_kud_adhoc5 <- kernelUD(Alto, h=hrefs0.5[1], same4all=TRUE)
Can anyone help me?
Thank you