plot.bvar_fcast function

Plotting method for Bayesian VAR predictions

Plotting method for forecasts obtained from predict.bvar. Forecasts of all or a subset of the available variables can be plotted.

## S3 method for class 'bvar_fcast'
plot(
  x,
  vars = NULL,
  col = "#737373",
  t_back = 1,
  area = FALSE,
  fill = "#808080",
  variables = NULL,
  orientation = c("vertical", "horizontal"),
  mar = c(2, 2, 2, 0.5),
  ...
)

Arguments

• x: A bvar_fcast object, obtained from predict.bvar.
• vars: Optional numeric or character vector. Used to subset the plot to certain variables by position or name (must be available). Defaults to NULL, i.e. all variables.
• col: Character vector. Colour(s) of the lines delineating credible intervals. Single values will be recycled if necessary. Recycled HEX color codes are varied in transparency if not provided (e.g. "#737373FF"). Lines can be bypassed by setting this to "transparent".
• t_back: Integer scalar. Number of observed datapoints to plot ahead of the forecast.
• area: Logical scalar. Whether to fill the credible intervals using polygon.
• fill: Character vector. Colour(s) to fill the credible intervals with. See col for more information.
• variables: Optional character vector. Names of all variables in the object. Used to subset and title. Taken from x$variables if available.
• orientation: String indicating the orientation of the plots. Defaults to "v" (i.e. vertical); may be set to "h" (i.e. horizontal).
• mar: Numeric vector. Margins for par.
• ...: Other graphical parameters for par.

Returns

Returns x invisibly.

Examples

# Access a subset of the fred_qd dataset
data <- fred_qd[, c("CPIAUCSL", "UNRATE", "FEDFUNDS")]
# Transform it to be stationary
data <- fred_transform(data, codes = c(5, 5, 1), lag = 4)

# Estimate a BVAR using one lag, default settings and very few draws
x <- bvar(data, lags = 1, n_draw = 1000L, n_burn = 200L, verbose = FALSE)

# Store predictions ex-post
predict(x) <- predict(x)

# Plot forecasts for all available variables
plot(predict(x))

# Subset to variables in positions 1 and 3 via their name
plot(predict(x), vars = c("CPI", "FED"))

# Subset via position, increase the plotted forecast horizon and past data
plot(predict(x, horizon = 20), vars = c(1, 3), t_back = 10)

# Adjust confidence bands and the plot's orientation
plot(predict(x, conf_bands = 0.25), orientation = "h")

# Draw areas inbetween the confidence bands and skip drawing lines
plot(predict(x), col = "transparent", area = TRUE)

# Plot a conditional forecast (with a constrained second variable).
plot(predict(x, cond_path = c(1, 1, 1, 1, 1, 1), cond_var = 2))

See Also

bvar; predict.bvar