| Line 119... |
Line 119... |
| 119 |
negligible and the speed-up considerable. Exact computations can be
|
119 |
negligible and the speed-up considerable. Exact computations can be
|
| 120 |
ensured by setting \code{delta} to a negative value.
|
120 |
ensured by setting \code{delta} to a negative value.
|
| 121 |
|
121 |
|
| 122 |
If \code{transform.pars} is true, the optimization is done using an
|
122 |
If \code{transform.pars} is true, the optimization is done using an
|
| 123 |
alternative parametrization which is a variation on that suggested by
|
123 |
alternative parametrization which is a variation on that suggested by
|
| 124 |
Jones (1980) and ensures that the model is stationary. For an AR(p)
|
124 |
\bibcitet{R:Jones:1980} and ensures that the model is stationary.
|
| - |
|
125 |
For an AR(p)
|
| 125 |
model the parametrization is via the inverse tanh of the partial
|
126 |
model the parametrization is via the inverse tanh of the partial
|
| 126 |
autocorrelations: the same procedure is applied (separately) to the
|
127 |
autocorrelations: the same procedure is applied (separately) to the
|
| 127 |
AR and seasonal AR terms. The MA terms are also constrained to be
|
128 |
AR and seasonal AR terms. The MA terms are also constrained to be
|
| 128 |
invertible during optimization by the same transformation if
|
129 |
invertible during optimization by the same transformation if
|
| 129 |
\code{transform.pars} is true. Note that the MLE for MA terms does
|
130 |
\code{transform.pars} is true. Note that the MLE for MA terms does
|