mountains {MSBVAR} | R Documentation |
"Mountain plots" summarize the bivariate density of 2 variables for two competing forecasts of those variables.
mountains(fcasts1, fcasts2, varnames, pts, ...)
fcasts1 |
gibbs x 2 set of forecasts from model 1 |
fcasts2 |
gibbs x 2 set of forecasts from model 2 |
varnames |
c("name1","name2") object of the variable names |
pts |
c(pt1,pt2) which are reference points to be plotted. |
... |
Other graphics parameters. |
A "mountain plot" provide a 2 x 2 graph of plots that summarize the
bivariate forecasts for two competing forecasts. This function
presents four perspectives on the bivariate density or 'hills' for a
set of forecasts. Starting from the bottom right plot and working
counter-clockwise, the first plot is the bivariate density of the two
competing forecasts. The next plot is a contour map that provide the
topography of the densities. The third and fourth plots are
projections of densities in each variable. The first forecast in the
function is presented in black, the second in red. The densities are
estimated from the Gibbs Monte Carlo sample of forecasts using the
bkde2D
bivariate kernel density estimator with an optimal
plug-in bandwidth selected using dpill
.
None. Produces the mountain plot described above in the current graphics device.
This function requires the bivariate kernel smoother in the
package bkde2D
Patrick T. Brandt
Brandt, Patrick T. and John R. Freeman. 2006. "Advances in Bayesian Time Series Modeling and the Study of Politics: Theory Testing, Forecasting, and Policy Analysis" Political Analysis 14(1):1-36.
bkde2D
for details
of the density estimators
## Not run: data(IsraelPalestineConflict) # Fit a BVAR model fit.BVAR <- szbvar(IsraelPalestineConflict, p=6, z=NULL, lambda0=0.6, lambda1=0.1, lambda3=2, lambda4=0.5, lambda5=0, mu5=0, mu6=0, nu=3, qm=4, prior=0, posterior.fit=FALSE) # Fit a flat prior / MLE model fit.FREQ <- szbvar(IsraelPalestineConflict, p=6, z=NULL, lambda0=0.6, lambda1=0.1, lambda3=2, lambda4=0.5, lambda5=0, mu5=0, mu6=0, nu=3, qm=4, prior=2, posterior.fit=FALSE) # Generate unconditional forecasts for both models forecast.BVAR <- uc.forecast.var(fit.BVAR, nsteps=2, burnin=100, gibbs=1000) forecast.FREQ <- uc.forecast.var(fit.FREQ, nsteps=2, burnin=100, gibbs=1000) # Plot the densities for the forecasts in period of the forecast horizon mountains(forecast.BVAR$forecast[,2,1:2], forecast.FREQ$forecast[,2,1:2], varnames=c("I2P","P2I"), pts=c(0,0)) ## End(Not run)