smo
of the output) which is free of outliers.
acm.smo(x, gm, s0=<<see below>>, iter=1, a=2.5, b=5, psiovw=T, savesd=T)
ar.gm
from
x
;
see
acm.filt
for details.
gm
) multiplied by 0.001.
s0
must be greater than
0
.
TRUE
, use psi(t)/t to calculate
the weight function w(t); if
FALSE
,
use psi'(t).
TRUE
, calculate and save the estimated time
varying scale of the smoothing error; if
FALSE
, omit these calculations.
order
(=
length(gm$ar)
) containing the
least squares estimate of the autoregression parameters calculated
from the smoothed data.
order
by
order
Toeplitz matrix containing
the estimated autocovariance matrix of the smoothed data.
savesd
is
TRUE
.
Martin, R. D. (1982).
Approximate conditional mean type smoothers and interpolators. In
Smoothing Techniques for Curve Estimation.
T. Gasser and M. Rosenblatt, eds. Springer Verlag, Berlin. pp. 117-143.
Martin, R. D. (1981).
Robust methods for time series. In
Applied Time Series Analysis II.
D. F. Findley, ed. Academic Press, New York. pp. 683-759.
gm <- ar.gm(bicoal.tons,3) acm.smo(bicoal.tons,gm)