Thank you so much! I have used Box-cox transformation. I need to sure
before submitting my paper for publication.
I used following codes for CSGS of ground water arsenic concentration.
Zia
data(tala)
was.bc<-box.cox(tala$was, 0.47)
tala.bc = cbind(tala, was.bc)
coordinates(tala.bc) <- ~ x + y
coordinates(tala.grid) <-~ x + y
v.ok.was<-variogram(was.bc~1,data=tala.bc)
plot(v.ok.was, pl=F, pch=20, cex=1, col="Black")
m.ok.was<-vgm(.06,"Exp",4000,0.02)
(m.ok.f.was<-fit.variogram(v.ok.was, m.ok.was))
# Simulation
#--------------
set.seed(130)
sim.was<-krige(was.bc~1,tala.bc,tala.grid,model=m.ok.f.was, nmax=50,
nsim = 1000)
# Back-transformation
#-----------------
for(i in 1:length(sim.was at data)){sim.was at data[[i]] <-
(((sim.was[[i]]*0.47+1)^k.was))}
sim.data.was<-as.data.frame(sim.was)
summary(sim.data.was)
Edzer Pebesma wrote:
Yes, you can. It is less flexible than the normal score transform -- try for instance transforming a bimodal distribution -- but has the advantage that the back-transform is straightforward. How do you for example use the inverse normal score transformation simulated values outside the range of the data? Zia Ahmed wrote:
I have a basic question related to the data transformation CSG simulation. Can I use Box-cox transformation of the data instead of normal score transformation before performing CSG simulation? Suggestion will be appreciated. Thanks ZIa
_______________________________________________ R-sig-Geo mailing list R-sig-Geo at stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/r-sig-geo
-------------- next part -------------- An HTML attachment was scrubbed... URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20100315/0422138f/attachment.html> -------------- next part -------------- A non-text attachment was scrubbed... Name: zua3.vcf Type: text/x-vcard Size: 281 bytes Desc: not available URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20100315/0422138f/attachment.vcf>