https://github.com/cran/ensembleBMA
Tip revision: 95cce576eb46ce512c1355769cfdafd8c2af4b02 authored by Chris Fraley on 25 December 2011, 00:00:00 UTC
version 5.0.3
version 5.0.3
Tip revision: 95cce57
verifPlot.R
verifPlot <-
function( fit, ensembleData, dates=NULL)
{
forc <- quantileForecast( fit, ensembleData, dates = dates,
quantiles = c( .1, .5, .9))
if (is.null(dates)) dates <- names(fit$nIter)
use <- as.logical(match(ensembleValidDates(ensembleData), dates, nomatch = 0))
lon <- ensembleData$longitude[use]; lat <- ensembleData$latitude[use]
verif <- dataVerifObs(ensembleData)[use]
nObs <- length(verif)
if (nObs == 0) stop("no observations")
ord <- order(forc[,"0.9"])
ylim <- c(0, max(forc,verif))
plot(1:nObs, forc[,"0.9"][ord], ylim = ylim, type = "l", col = "black",
xlab = "Observations in order of increasing 90th percentile forecast",
ylab = "Precipitation (hundreths of an inch)", xaxt = "n")
lines(1:nObs, forc[,"0.1"][ord], type = "l", col = "black")
lines(1:nObs, forc[,"0.5"][ord], type = "l", col = "red")
points(1:nObs, verif, pch = 16, col = "black", cex = 0.5)
invisible(forc)
}