PacfPlot {FitAR} | R Documentation |
The partial autocorrelations and their individual 95 percent confidence intervals are plotted under the assumption the model is contained in an AR(P), where P is a specified upper limit.
PacfPlot(z, lag.max = 15, ...)
z |
time series |
lag.max |
maximum lag, P |
... |
optional parameters passed through to plot. |
The Burg algorithm is used to estimate the PACF.
No value is returned. Graphical output is produced as side-effect.
A.I. McLeod
McLeod, A.I. and Zhang, Y. (2006). Partial autocorrelation parameterization for subset autoregression. Journal of Time Series Analysis, 27, 599-612.
#For the log(lynx) series and taking lag.max=15, the PacfPlot and # the minimum BIC subset selection produce the same result. z<-log(lynx) PacfPlot(z) SelectModel(z,lag.max=15,SubsetModel="z",Best=1,Criterion="BIC")