Learn R Programming

BIPOD (version 0.2.1)

plot.BIPOD: Plot function for class BIPOD

Description

Graphical summaries of output from Gibbs sampler

Usage

"plot"(x, theta = NULL, subset = NULL, type, lag = 20, interval = NULL, treshold = 0.1, speed = 0.1, truepath = NULL, BY = 1, prop = c(.05,.95), diffPriorMean = NULL, diffPriorCovar = NULL, log = FALSE, ...)

Arguments

x
x: An object of class BIPOD.
theta
Optional numerical vector with true parameter values.
subset
Numeric vector. Which parameters should be used for plotting. Defaults to all.
type
Character choosing plotting type: Either "trace", "hist", "acp", "pairs","SDtrace", "accept", "movie" or "cover". See details.
lag
Positive integer: Number of lags used in autocorrelation plot
interval
Positive integer or numericla vector: If integer, used as burn in for the Gibbs sampler. If vector, used to subsample Gibbs output. Defaults to no subsampling and no burn in.
treshold
Positive numeric: Cut off for acceptance rate.
speed
Positive number used for type="movie": How much time to pause between each frame of the movie?
truepath
Vector or matrix with latent data. Optional.
BY
integer: Only relevant for type="movie". How many frames to skip for each iteration?
prop
Numeric vector with values between 0 and 1: Only relevant for type="cover". Specifies quantiles for the sampled paths
diffPriorMean
Numeric of length 2, only for type=="hist". The prior mean for the diffusion coefficients.
diffPriorCovar
2*2 matrix, only for type=="hist". The prior covariance for the diffusion coefficients.
log
Boolean, only for type=="hist". If TRUE, the prior density and the estimate of the diffusion coefficients are log transformed before plotting.
...
Additional arguments to be passed to matplot, density or acf.

Details

Different 'type'-argument gives different plots. More details to come...