TV_heatmap: Heatmap of hyperparameters of time-varying coefficient matrix in a TVP-VAR model

Description

TV_heatmap plots a heatmap of posterior distribution for hyperparameters of a time-varying parameter coefficient matrix in a TVP-VAR model. This is achieved by plotting the median of the posterior of the absolute value for theta_sr and the median of the posterior for all others. The plot itself is generated by lattice::levelplot. beta_mean and theta_sr will most likely be of interest here.

Usage

TV_heatmap(x, cuts = 15, cols, max_val, flipcols, ...)

Value

Called for its side effects and returns invisibly.

Arguments

x: mcmc.var object
cuts: single integer value, determining the number of cuts for the color palette. The default value is 15.
cols: character string, determining the color palette to be used. The default value is "Purples" for theta_sr and "RdBu" for all others. See RColorBrewer::brewer.pal.info for more information.
max_val: numeric value, determining the maximum value for the color palette. If missing, the maximum value is determined by the largest absolute value of the data.
flipcols: logical value, determining whether the color palette should be flipped. The default value is FALSE for theta_sr and TRUE for all others.
...: further arguments to be passed to lattice::levelplot.

Author

Peter Knaus peter.knaus@wu.ac.at

Examples

Run this code

# \donttest{
set.seed(123)
sim <- simTVPVAR(p = 2)
data <- sim$data

res <- shrinkTVPVAR(data, p = 2)
TV_heatmap(res$theta_sr)
# }