prepareSimSarima: Prepare SARIMA simulations

In GeoBosh/sarima: Simulation and Prediction with Seasonal ARIMA Models

Prepare SARIMA simulations

Description

Prepare SARIMA simulations.

Usage

prepareSimSarima(model, x = NULL, eps = NULL, n, n.start = NA,
                 xintercept = NULL, rand.gen = rnorm)

## S3 method for class 'simSarimaFun'
print(x, ...)

Arguments

model	an object from a suitable class or a list, see Details.
x	initial/before values of the time series, a list, a numeric vector or time series, see Details.
eps	initial/before values of the innovations, a list or a numeric vector, see Details.
n	number of observations to generate, if missing an attempt is made to infer it from x and eps.
n.start	number of burn-in observations.
xintercept	non-constant intercept which may represent trend or covariate effects.
rand.gen	random number generator, defaults to N(0,1).
...	ignored.

Details

prepareSimSarima does the preparatory work for simulation from a Sarima model, given the specifications and returns a function, which can be called as many times as needed.

The variance of the innovations is specified by the model and the simulated innovations are multiplied by the corresponding standard deviation. So, it is expected that the random number generator simulates from a standardised distribution.

Argument model can be from any class representing models in the SARIMA family, such as "SarimaModel", or a list with components suitable to be passed to =new()= for such models.

The canonical form of argument x is a list with components before, init and main. If any of these components is missing or NULL, it is filled suitably. Any other components of x are ignored. If x is not a list, it is put in component main. Conceptually, the three components are concatenated in the given order, the simulated values are put in main (before and init are not changed), the before part is dropped and the rest is returned. In effect, before and init can be viewed as initial values but init is considered part of the generated series.

The format for eps is the same as that of x. The lengths of missing components in x are inferred from the corresponding components of eps, and vice versa.

The format for xintercept is the same as that of x and eps.

print.simSarimaFun is a print method for the objects generated by prepareSimSarima.

Value

for prepareSimSarima, a function to simulate time series, see Details. it is typically called multiple times without arguments. All arguments have defaults set by prepareSimSarima.

n	length of the simulated time series,
rand.gen	random number generator,
...	arguments for the random number generator, passed on to arima.sim.

Author(s)

Georgi N. Boshnakov

See Also

sim_sarima

Examples

mo1 <- list(ar = 0.9, iorder = 1, siorder = 1, nseasons = 4, sigma2 = 2)
fs1 <- prepareSimSarima(mo1, x = list(before = rep(0,6)),  n = 100)
tmp1 <- fs1()
tmp1
plot(ts(tmp1))

fs2 <- prepareSimSarima(mo1, x = list(before = rep(1,6)),  n = 100)
tmp2 <- fs2()
plot(ts(tmp2))

mo3 <- mo1
mo3[["ar"]] <- 0.5
fs3 <- prepareSimSarima(mo3, x = list(before = rep(0,6)),  n = 100)
tmp3 <- fs3()
plot(ts(tmp3))

GeoBosh/sarima documentation built on March 27, 2024, 6:31 p.m.