| Line 8... |
Line 8... |
| 8 |
arima.sim(model, n, rand.gen = rnorm, innov = rand.gen(n, \dots),
|
8 |
arima.sim(model, n, rand.gen = rnorm, innov = rand.gen(n, \dots),
|
| 9 |
n.start = NA, \dots)
|
9 |
n.start = NA, \dots)
|
| 10 |
}
|
10 |
}
|
| 11 |
\arguments{
|
11 |
\arguments{
|
| 12 |
\item{model}{A list with component \code{ar} and/or \code{ma} giving
|
12 |
\item{model}{A list with component \code{ar} and/or \code{ma} giving
|
| 13 |
the AR and MA coeffcients respectively. Optionally a component
|
13 |
the AR and MA coefficients respectively. Optionally a component
|
| 14 |
\code{order} can be used.}
|
14 |
\code{order} can be used.}
|
| 15 |
\item{n}{length of output series, before un-differencing.}
|
15 |
\item{n}{length of output series, before un-differencing.}
|
| 16 |
\item{rand.gen}{optional: a function to generate the innovations.}
|
16 |
\item{rand.gen}{optional: a function to generate the innovations.}
|
| 17 |
\item{innov}{an optional times series of innovations. If not
|
17 |
\item{innov}{an optional times series of innovations. If not
|
| 18 |
provided, \code{rand.gen} is used.}
|
18 |
provided, \code{rand.gen} is used.}
|
| 19 |
\item{n.start}{length of \dQuote{burn-in} period. If \code{NA}, the
|
19 |
\item{n.start}{length of \dQuote{burn-in} period. If \code{NA}, the
|
| 20 |
default, a reasonable value is computed.}
|
20 |
default, a reasonable value is computed.}
|
| 21 |
\item{\dots}{additional arguments for \code{rand.gen}. Most usefully,
|
21 |
\item{\dots}{additional arguments for \code{rand.gen}. Most usefully,
|
| 22 |
the standard deviation of the innovations generated by \code{rnorm}
|
22 |
the standard deviation of the innovations generated by \code{rnorm}
|
| 23 |
can be specified by \code{sd}.}
|
23 |
can be specified by \code{Sc}.}
|
| 24 |
}
|
24 |
}
|
| 25 |
\details{
|
25 |
\details{
|
| 26 |
See \code{\link{arima}} for the precise definition of an ARIMA model.
|
26 |
See \code{\link{arima}} for the precise definition of an ARIMA model.
|
| 27 |
|
27 |
|
| 28 |
The ARMA model is checked for stationarity.
|
28 |
The ARMA model is checked for stationarity.
|
| 29 |
|
29 |
|
| 30 |
ARIMA models are specified via the \code{order} component of
|
30 |
ARIMA models are specified via the \code{order} component of
|
| 31 |
\code{model}, in the same way as for \code{\link{arima}}. Other
|
31 |
\code{model}, in the same way as for \code{\link{arima}}. Other
|
| 32 |
aspects of the \code{order} component are ignored, but inconsistent
|
32 |
aspects of the \code{order} component are ignored, but inconsistent
|
| 33 |
specifications of the MA and AR orders are detected. The
|
33 |
specifications of the MA and AR orders are detected. The
|
| 34 |
un-differencing assumes previous values of zero, and to remind the
|
34 |
un-differencing assumes previous values of zero, and to remind the
|
| 35 |
user of this, those values are returned.
|
35 |
user of this, those values are returned.
|
| 36 |
}
|
36 |
}
|
| 37 |
\value{
|
37 |
\value{
|
| 38 |
A time-series object of class \code{"ts"}.
|
38 |
A time-series object of class \code{"ts"}.
|
| 39 |
}
|
39 |
}
|
| 40 |
\seealso{
|
40 |
\seealso{
|
The R Project SVN
R