spec.ic: Estimate Spectral Density of a Time Series from AR Fit

In astsa: Applied Statistical Time Series Analysis

Estimate Spectral Density of a Time Series from AR Fit

Description

Fits an AR model to data and computes (and by default plots) the spectral density of the fitted model based on AIC (default) or BIC.

Usage

spec.ic(xdata, BIC=FALSE, order.max=NULL, main=NULL, plot=TRUE, detrend=TRUE, 
         lowess=FALSE, method=NULL, cex.main=NULL, xlab=NULL, ...)

Arguments

xdata	a univariate time series.
BIC	if TRUE, fit is based on BIC. If FALSE (default), fit is based on AIC.
order.max	maximum order of model to fit. Defaults (if NULL) to the minimum of 100 and 10% of the number of observations.
main	plot title. Defaults to name of series, method and chosen order.
plot	if TRUE (default) produces a graphic of the estimated AR spectrum.
detrend	if TRUE (default), detrends the data first. If FALSE, the series is demeaned.
lowess	if TRUE, detrends using lowess. Default is FALSE.
method	method of estimation - a character string specifying the method to fit the model chosen from the following: "yule-walker", "burg", "ols", "mle", "yw". Defaults to "yule-walker".
cex.main	magnification for main title; default is 1.
xlab	label for frequency axis; if NULL (default), a totally awesome label is generated for your viewing pleasure.
...	additional graphical arguments.

Details

Uses ar to fit the best AR model based on pseudo AIC or BIC. Using method='mle' will be slow. The minimum centered AIC and BIC values and the spectral and frequency ordinates are returned silently.

Value

[[1]]	Matrix with columns: ORDER, AIC, BIC
[[2]]	Matrix with columns: freq, spec

Author(s)

D.S. Stoffer

References

You can find demonstrations of astsa capabilities at FUN WITH ASTSA.

The most recent version of the package can be found at https://github.com/nickpoison/astsa/.

In addition, the News and ChangeLog files are at https://github.com/nickpoison/astsa/blob/master/NEWS.md.

The webpages for the texts and some help on using R for time series analysis can be found at https://nickpoison.github.io/.

See Also

ar, spec.ar

Examples

## Not run: 
# AIC
spec.ic(soi)  
spec.ic(sunspotz, method='burg', col=4)   

# BIC after detrending on log scale
spec.ic(soi, BIC=TRUE, detrend=TRUE, log='y')  

# plot AIC and BIC without spectral estimate
tsplot(0:30, spec.ic(soi, plot=FALSE)[[1]][,2:3], type='o', xlab='order', nxm=5)  

## End(Not run)

astsa documentation built on May 29, 2024, 10:29 a.m.