ForecastMoments

The joint forecast observation distribution for Gaussian variables can be described with 6 low-order moments. 'ForecastMoments' provides functions to compute these moments, and to interpret forecast quality using a signal-plus-noise model based on these low-order moments. Uncertainty assessment of these moments and derived forecast quality information including traditional verification scores and skill scores is provided based on parametric bootstrap resampling.

Installation

The package can be installed from github.

library(devtools)
install_github("jonasbhend/ForecastMoments")

Example

This is a basic example which shows you how to solve a common problem:

library(ForecastMoments)
#> Loading required package: easyVerification
#> Loading required package: SpecsVerification
#> 
#> Attaching package: 'easyVerification'
#> The following object is masked from 'package:SpecsVerification':
#> 
#>     EnsCorr

tm <- easyVerification::toymodel()
tmboot <- SNRresample(tm$fcst, tm$obs, c("EnsCorr", "Ens2AFC", "FairCrpss"))

summary(tmboot)
#>                               orig  boot_mean    boot_sd boot_p0.025
#> EnsCorr                 0.52663157 0.49627656 0.15010587  0.18242548
#> Ens2AFC                 0.70454574 0.66963512 0.05926459  0.54985062
#> FairCrpss.skillscore    0.10409977 0.05451122 0.13206492 -0.22295674
#> FairCrpss.skillscore.sd 0.09782066 0.12859667 0.03204040  0.07687236
#>                         boot_p0.975
#> EnsCorr                   0.7433533
#> Ens2AFC                   0.7767993
#> FairCrpss.skillscore      0.2799730
#> FairCrpss.skillscore.sd   0.2068374

boots <- summary(tmboot)
par(mfrow=c(1,nrow(boots) - 1)) 
tmp <- sapply(seq(1,nrow(boots)-1), function(i){
  hist(tmboot[i,-1], col='grey', main=rownames(tmboot)[i], xlab='')
  abline(v=boots[i,-3], lwd=2, lty=c(1,1,3,3), col=c(2,1,1,1))
  if (i == 1) legend("topleft", legend=colnames(boots)[-3], 
                     lwd=2, col=c(2,1,1,1), lty=c(1,1,3,3), 
                     cex=0.83, inset=0.05)
})