LagReg {astsa} | R Documentation |
Performs lagged regression as discussed in Chapter 4.
LagReg(input, output, L = c(3, 3), M = 40, threshold = 0, inverse = FALSE)
input |
input series |
output |
output series |
L |
degree of smoothing; see |
M |
must be even; number of terms used in the lagged regression |
threshold |
the cut-off used to set small (in absolute value) regression coeffcients equal to zero |
inverse |
if TRUE, will fit a forward-lagged regression |
For a bivariate series, input
is the input series and output
is the output series. The degree of smoothing for the spectral estimate is given by L; see spans
in the help file for spec.pgram
. The number of terms used in the lagged regression approximation is given by M, which must be even. The threshold value is the cut-off used to set small (in absolute value) regression coeffcients equal to zero (it is easiest to run LagReg twice, once with the default threshold of zero, and then again after inspecting the resulting coeffcients and the corresponding values of the CCF). Setting inverse=TRUE will fit a forward-lagged regression; the default is to run a backward-lagged regression. The script is based on code that was contributed by Professor Doug Wiens, Department of Mathematical and Statistical Sciences, University of Alberta.
Graphs of the estimated impulse response function, the CCF, and the output with the predicted values superimposed.
beta |
Estimated coefficients |
fit |
The output series, the fitted values, and the residuals |
D.S. Stoffer
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 are https://www.stat.pitt.edu/stoffer/tsa4/ and https://www.stat.pitt.edu/stoffer/tsda/.