Trabajo Prctico 2

# Anlisis exploratorio
## Estadstica descriptiva

> names(soil250)
[1] "Linha" "Coluna" "Cota" "AGrossa" "Silte" "Argila" "pHAgua"
[8] "pHKCl" "Ca" "Mg" "K" "Al" "H" "C"
[15] "N" "CTC" "S" "V" "M" "NC" "CEC"
[22] "CN"
> summary(soil250$Ca)
Min. 1st Qu. Median Mean 3rd Qu. Max.
2.200 2.925 3.200 3.244 3.600 4.300
>
numSummary(soil250$Ca,statistic=c("mean","sd","quantiles"),quantiles=c(
0,.25,.5,.75,1))
mean sd 0% 25% 50% 75% 100% n
3.2436 0.4405854 2.2 2.925 3.2 3.6 4.3 250
> all.moments(soil250$Ca,order.max=4)
[1] 1.00000 3.24360 10.71428 36.02413 123.21679
> cor(soil250[1:2],soil250[,9])
[,1]
Linha -0.5870375
Coluna 0.1019318

## Grficos
par(mfrow=c(2,1))
hist((soil250$Ca),col="darkgrey",main="[Ca]",xlab="[Ca]")
boxplot(soil250$Ca,main="[Ca]",range=1.5,horizontal=T)

par(mfrow=c(1,1))
qqPlot(soil250$Ca,dist="norm",ylab=[Ca])
log[Ca]<-log(soil250$Ca)
qqPlot(log[Ca],dist="norm",ylab=log[Ca])

## Normalidad
> shapiro.test(soil250$Ca)

Shapiro-Wilk normality test

data: soil250$Ca
W = 0.9866, p-value = 0.01927
> shapiro.test(log[Ca])
Shapiro-Wilk normality test

data: Calog
W = 0.9884, p-value = 0.0421

par(mfrow=c(2,2))
qqPlot(soil250$Ca,dist="norm",ylab="[Ca]")
qqPlot(Calog,dist="norm",ylab="log[Ca]")
hist((soil250$Ca),col="darkgrey",main="[Ca]",xlab="[Ca]")
hist((Calog),col="darkgrey",main="[Ca]",xlab="log[Ca]",breaks=11)

## Geodatos
## En geoR
soilca<-as.geodata(soil250,coords.col=1:2,data.col=9)
plot(soilca,low=T)
image(soil250,"Ca",xcol=1,ycol=2,asp=1)

## Variograma Experimental
vxp<-
variog(soilgeo,coord=soilgeo$coords,data=soilgeo$data[,7],estimador.type
="classical",max.dist=75,option="bin",uvec="default",trend="cte")
vxp.cloud<-
variog(soilgeo,coord=soilgeo$coords,data=soilgeo$data[,7],estimador.type
="classical",max.dist=75,option="cloud",uvec="default",trend="cte")
vxp1<-
variog(soilgeo,coord=soilgeo$coords,data=soilgeo$data[,7],estimador.type
="classical",max.dist=75,option="bin",uvec="default",trend="1st")
vxp2<-
variog(soilgeo,coord=soilgeo$coords,data=soilgeo$data[,7],estimador.type
="classical",max.dist=75,option="bin",uvec="default",trend="2nd")
par(mfrow=c(2,2))
plot(vxp,main=Variograma Experimental)
plot(vxp.cloud,main=Puntos)
plot(vxp1,main=Variograma 1st)
plot(vxp2,main=Variograma 2nd)

## Variogramas direccionales
ca<-as.geodata(soil250,coords.col=1:2,data.col=9,header=T)
vd<-variog4(ca,max.dist=75)
vd1<-variog4(ca,max.dist=75,trend="1st")
vd2<-variog4(ca,max.dist=75,trend="2nd")
par(mfrow=c(2,2))
plot(vd)
plot(vd1)
plot(vd2)
## Ajuste del variograma
v<-variog(ca,max.dist=75,trend="2nd")
plot(v)
ini<-eyefit(v)
line<-variofit(v,ini)
lines(line)
summary(line)
$pmethod
[1] "WLS (weighted least squares)"

$cov.model
[1] "exponential"

$spatial.component
sigmasq phi
0.09439789 5.81301720

$spatial.component.extra
kappa
0.5

$nugget.component
tausq
0
$fix.nugget
[1] FALSE

$fix.kappa
[1] TRUE

$practicalRange
[1] 17.41426

$sum.of.squares
value
0.8096588

$estimated.pars
tausq sigmasq phi
0.00000000 0.09439789 5.81301720

$weights
[1] "npairs"

$call
variofit(vario = v, ini.cov.pars = ini)

attr(,"class")
[1] "summary.variomodel"

## Ajustar el variograma
vfit<-fit.variogram(v,model=vgm(1,"Sph",30,1))
plot(v,vfit)