plot.forecast {MSBVAR} | R Documentation |
Produces a high level plot of two sets of VAR forecasts computed from
a Markov Chain Monte Carlo (MCMC) posterior sample. Forecasts are the
output of hc.forecast
, or uc.forecast
.
plot.forecast(x, y = NULL, varnames = NULL, start = c(0, 1), freq = 1, probs = c(0.05, 0.95), compare.level = NULL, ylab = NULL, ...)
x |
First sample of forecasts |
y |
Second sample of forecasts, default = NULL |
varnames |
m x 1 list of variable names, e.g.,
c("name1","name2",...) |
start |
Start date for the forecast period time series, default
= c(0,1) |
freq |
Time series frequency, in format consistent with
ts . Default is 1 |
probs |
Probability limits for error bands. Default is 90% or
c(0.05,0.95) |
compare.level |
m list of "comparison levels" to be drawn as horizontal lines. By default zero is included. This is useful for highlighting specific reference points for the forecasts |
ylab |
list of labels for the y-axes |
... |
other graphics parameters |
Plots the mean forecast and the pointwise empirical confidence region
for a posterior sample of VAR forecasts. Overlays a second set of
forecasts and error bands if requested in fcasts2
. This should
be prefaced by a call to par(mfrow(...))
or
par(mfcol(...))
to set the layout of the plots for each series.
None. Plots forecasts on the current display device.
Patrick T. Brandt
plot.forc.ecdf
, and
uc.forecast
for an example.