stochvol: volplot – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

volplot

Plotting Quantiles of the Latent Volatilities

Description

Displays quantiles of the posterior distribution of the volatilities over time as well as predictive distributions of future volatilities.

Usage

volplot(
  x,
  forecast = 0,
  dates = NULL,
  show0 = FALSE,
  forecastlty = NULL,
  tcl = -0.4,
  mar = c(1.9, 1.9, 1.9, 0.5),
  mgp = c(2, 0.6, 0),
  simobj = NULL,
  newdata = NULL,
  ...
)

Arguments

`x`	`svdraws` object.
`forecast`	nonnegative integer or object of class `svpredict`, as returned by `predict.svdraws`. If an integer greater than 0 is provided, `predict.svdraws` is invoked to obtain the `forecast`-step-ahead prediction. The default value is `0`.
`dates`	vector of length `ncol(x$latent)`, providing optional dates for labeling the x-axis. The default value is `NULL`; in this case, the axis will be labeled with numbers.
`show0`	logical value, indicating whether the initial volatility `exp(h_0/2)` should be displayed. The default value is `FALSE`. Only available for inputs `x` of class `svdraws`.
`forecastlty`	vector of line type values (see `par`) used for plotting quantiles of predictive distributions. The default value `NULL` results in dashed lines.
`tcl`	The length of tick marks as a fraction of the height of a line of text. See `par` for details. The default value is `-0.4`, which results in slightly shorter tick marks than usual.
`mar`	numerical vector of length 4, indicating the plot margins. See `par` for details. The default value is `c(1.9, 1.9, 1.9, 0.5)`, which is slightly smaller than the R-defaults.
`mgp`	numerical vector of length 3, indicating the axis and label positions. See `par` for details. The default value is `c(2, 0.6, 0)`, which is slightly smaller than the R-defaults.
`simobj`	object of class `svsim` as returned by the SV simulation function `svsim`. If provided, “true” data generating values will be added to the plot(s).
`newdata`	corresponds to parameter `newdata` in `predict.svdraws`. Only if `forecast` is a positive integer and `predict.svdraws` needs a `newdata` object. Corresponds to input parameter `designmatrix` in `svsample`. A matrix of regressors with number of rows equal to parameter `forecast`.
`...`	further arguments are passed on to the invoked `ts.plot` function.

Value

Called for its side effects. Returns argument x invisibly.

Note

In case you want different quantiles to be plotted, use updatesummary on the svdraws object first. An example of doing so is given below.

Author(s)

Gregor Kastner gregor.kastner@wu.ac.at

Examples

## Simulate a short and highly persistent SV process 
sim <- svsim(100, mu = -10, phi = 0.99, sigma = 0.2)

## Obtain 5000 draws from the sampler (that's not a lot)
draws <- svsample(sim$y, draws = 5000, burnin = 100,
		  priormu = c(-10, 1), priorphi = c(20, 1.5),
		  priorsigma = 0.2)

## Plot the latent volatilities and some forecasts
volplot(draws, forecast = 10)

## Re-plot with different quantiles
newquants <- c(0.01, 0.05, 0.25, 0.5, 0.75, 0.95, 0.99)
draws <- updatesummary(draws, quantiles = newquants)

volplot(draws, forecast = 10)

stochvol

Efficient Bayesian Inference for Stochastic Volatility (SV) Models

v3.0.4

GPL (>= 2)

Authors

Darjus Hosszejni [aut, cre] (<https://orcid.org/0000-0002-3803-691X>), Gregor Kastner [aut] (<https://orcid.org/0000-0002-8237-8271>)

Initial release