runningcormat: Extract summary statistics for the posterior correlation...

In factorstochvol: Bayesian Estimation of (Sparse) Latent Factor Stochastic Volatility Models

Extract summary statistics for the posterior correlation matrix which have been stored during sampling

Description

runningcormat extracts summary statistics from the model-implied correlation matrix from an fsvdraws object for one point in time.

Usage

runningcormat(x, i, statistic = "mean", type = "cor")

Arguments

x	Object of class 'fsvdraws', usually resulting from a call of fsvsample.
i	A single point in time.
statistic	Indicates which statistic should be extracted. Defaults to 'mean'.
type	Indicates whether covariance (cov) or correlation (cor) should be extracted.

Value

Matrix containing the requested correlation matrix summary statistic.

See Also

Other extractors: cormat.fsvdraws(), covmat.fsvdraws(), runningcovmat()

Examples

set.seed(1)
sim <- fsvsim(n = 500, series = 3, factors = 1) # simulate
res <- fsvsample(sim$y, factors = 1, runningstore = 6) # estimate

cor100mean <- runningcormat(res, 100) # extract mean at t = 100
cor100sd <- runningcormat(res, 100, statistic = "sd") # extract sd
lower <- cor100mean - 2*cor100sd
upper <- cor100mean + 2*cor100sd

true <- cormat(sim, 100)[,,1] # true value

# Visualize mean +/- 2sd and data generating values
par(mfrow = c(3,3), mar = c(2, 2, 2, 2))
for (i in 1:3) {
 for (j in 1:3) {
  plot(cor100mean[i,j], ylim = range(lower, upper), pch = 3,
  main = paste(i, j, sep = ' vs. '), xlab = '', ylab = '')
  lines(c(1,1), c(lower[i,j], upper[i,j]))
  points(true[i,j], col = 3, cex = 2)
 }
}

factorstochvol documentation built on Nov. 24, 2023, 5:08 p.m.