SpatioTemporal (version 1.1.7)

plot.predCVSTmodel: Plots for predictSTmodel and predCVSTmodel Objects

Description

plot method for classes predictSTmodel and predCVSTmodel. Provides several different plots of the data.

Usage

# S3 method for predCVSTmodel
plot (x, y = "time",
    ID = colnames(x$pred.all$EX)[1],
    col = c("black", "red", "grey"), pch = c(NA, NA),
    cex = c(1, 1), lty = c(1, 1), lwd = c(1, 1), p = 0.95,
    pred.type = "EX", pred.var = TRUE, add = FALSE, ...)
  # S3 method for predictSTmodel
plot (x, y = "time",
    STmodel = NULL, ID = x$I$ID[1],
    col = c("black", "red", "grey"), pch = c(NA, NA),
    cex = c(1, 1), lty = c(1, 1), lwd = c(1, 1), p = 0.95,
    pred.type = "EX", pred.var = FALSE, add = FALSE, ...)

Arguments

x

predictSTmodel or predCVSTmodel object to plot.

y

Plot predictions as a function of either "time" or "obs"ervations.

STmodel

STdata/STmodel object containing observations with which to compare the predictions (not used for plot.predCVSTmodel).

ID

The location for which we want to plot predictions. A string matching names in colnames(x$EX) (or x$I$ID, number(s) which are used as ID = colnames(x$EX)[ID], or "all" in which case all predictions are used. If several locations are given (or "all") then y must be "obs".

col

A vector of three colours: The first is the colour of the predictions, second for the observations and third for the polygon illustrating the confidence bands. For y="obs" the colours are 1) colour of the points, 2) colour of the 1-1 line, and 3) colour of the polygon. If ID="all", picking col[1]="ID" will colour code the observations-prediction points by site ID.

pch,cex,lty,lwd

Vectors with two elements giving the point type, size, line type and line width to use when plotting the predictions and observations respectively. Setting a value to NA will give no points/lines for the predictions/observations. When plotting predictions as a function of observations lty[2] is used for the addition of abline(0,1, lty=lty[2], col=col[2], lwd=lwd[2]); pch[2] and cex[2] are ignored.

p

Width of the plotted confidence bands (as coverage percentage, used to find appropriate two-sided normal quantiles).

pred.type

Which type of prediction to plot, one of "EX", "EX.mu", "EX.mu.beta", or "EX.pred"; see the output from predict.STmodel

pred.var

Should we plot confidence bands based on prediction (TRUE) or confidence intrevalls (FALSE), see predict.STmodel. Only relevant if pred.type="EX" or pred.type="EX.pred". NOTE: The default differs for plot.predictSTmodel and plot.predCVSTmodel!

add

Add to existing plot?

...

Additional parameters passed to plot.

Value

Nothing

See Also

Other predCVSTmodel methods: estimateCV, estimateCV.STmodel, predictCV, predictCV.STmodel, print.predCVSTmodel, print.summary.predCVSTmodel, qqnorm.predCVSTmodel, qqnorm.STdata, qqnorm.STmodel, scatterPlot.predCVSTmodel, scatterPlot.STdata, scatterPlot.STmodel, summary.predCVSTmodel

Other predictSTmodel methods: predict.STmodel, print.predictSTmodel

Examples

# NOT RUN {
#######################################
## plot predictions for a given site ##
#######################################
##load data
data(mesa.model)
##load predictions
data(pred.mesa.model)

par(mfrow=c(2,1))
plot(pred.mesa.model)
##different site with data and prediction variances
plot(pred.mesa.model, STmodel=mesa.model, ID="L001",
     pred.var=TRUE)

##compare the different contributions to the predictions
plot(pred.mesa.model)
plot(pred.mesa.model, pred.type="EX.mu", col="red", add=TRUE)
plot(pred.mesa.model, pred.type="EX.mu.beta", col="green", add=TRUE)

##compare the two confidence and prediction intervalls
plot(pred.mesa.model, ID=3, pred.var=TRUE, col=c(0,0,"darkgrey"))
plot(pred.mesa.model, ID=3, STmodel=mesa.model,
     col=c("black","red","lightgrey"), add=TRUE)

##plot predictions as function of observations
par(mfrow=c(2,2))
plot(pred.mesa.model, y="obs", STmodel=mesa.model, pred.var=TRUE)

##all data, using points and colour coded by site
plot(pred.mesa.model, y="obs", STmodel=mesa.model, ID="all",
     lty=c(NA,1), pch=c(19,NA), col=c("ID", "red", "grey"),
     cex=.25, pred.var=TRUE)

##compare prediction methods, for one site only
plot(pred.mesa.model, y="obs", STmodel=mesa.model,
     lty=c(NA,1), pch=c(19,NA), cex=.25, pred.var=TRUE)
plot(pred.mesa.model, y="obs", STmodel=mesa.model, col="red",
     lty=NA, pch=c(19,NA), cex=.25, pred.type="EX.mu",
     add=TRUE)
plot(pred.mesa.model, y="obs", STmodel=mesa.model, col="green",
     lty=NA, pch=c(19,NA), cex=.25, pred.type="EX.mu.beta",
     add=TRUE)

####################################
## plot CV-pred. for a given site ##
####################################
##load CV-predictions
data(pred.cv.mesa)

par(mfcol=c(3,1),mar=c(2.5,2.5,2,.5))
plot(pred.cv.mesa, ID=1)
plot(pred.cv.mesa, ID=1, pred.type="EX.mu", col="green", add=TRUE)
plot(pred.cv.mesa, ID=1, pred.type="EX.mu.beta", col="blue", add=TRUE)

##different colours
plot(pred.cv.mesa, ID=10, col=c("blue","magenta","light blue"))

##points and lines for the observations
plot(pred.cv.mesa, ID=17, lty=c(1,NA), pch=c(NA,19), cex=.5)

##plot predictions as function of observations
par(mfrow=c(2,2))
plot(pred.cv.mesa, y="obs")

##all data, using points and colour coded by site
plot(pred.cv.mesa, y="obs", ID="all", lty=c(NA,1),
     pch=c(19,NA), cex=.25, col=c("ID", "red", "grey"))

##compare prediction methods, for one site only
plot(pred.cv.mesa, y="obs", lty=c(NA,1), pch=c(19,NA), cex=.25)
plot(pred.cv.mesa, y="obs", col="red", lty=NA, pch=c(19,NA),
     cex=.25, pred.type="EX.mu", add=TRUE)
plot(pred.cv.mesa, y="obs", col="green", lty=NA, pch=c(19,NA),
     cex=.25, pred.type="EX.mu.beta", add=TRUE)
# }