iden: ARIMA Model Identification

In wqmeeker/RTseries: Interactive Time Series Analysis Tools

ARIMA Model Identification

Description

Graphical outputs to help tentatively identify a ARIMA model. The plots produced are: plot of the original or transformed data, a range-mean plot of the transformed data, and plots of the sample ACF and PACF.

Usage

iden(
  data.tsd,
  seasonal = tsp(data.tsd)[3],
  gamma = 1,
  m = 0,
  d = 0,
  D = 0,
  lag.max = 38,
  print.table = FALSE,
  seasonal.lags = FALSE
)

Arguments

data.tsd	Time series data object (output from the tsd function).
seasonal	Optional. The number of observations in a period for a seasonal timeseries. By default this value is automatically taken from the time-series data object (.tsd object).
gamma	Optional. The Box-Cox transformation power parameter. The default value is gamma=1 (no transformation).
m	Optional. The constant that is added to the response variable before the data are transformed. The default value is m=0.
d	The number of regular differences carried out on the data. If a value is not specified then the default value is d=0.
D	Optional. The number of seasonal differences carried out on the data. The default value is D=0.
lag.max	Optional. The maximum number of lags at which to estimate the ACF and PACF. The default value is lag.max=38
print.table	Optional. Default value is print.table = FALSE. If print.table = TRUE, R will output a table of the ACF and PACF estimates.
seasonal.lags	if TRUE will label ACF and PACF lags in seasonal multiples. Default is FALSE.

Value

Invisible list containing the numerical values of the ACF and PACF estimates that were plotted.

Examples

iden(spot.tsd)

Passengers.ts <- ts(Passengers, freq=12, start=1949)
Passengers.tsd <- tsd(Passengers.ts, data.title='International Airline Passengers',
                      time.units='Year',response.units='Thousands of Passengers')
iden(Passengers.tsd)
iden(Passengers.tsd, d=1, D=1)
iden(Passengers.tsd, d=1, D=1, gamma=0)

wqmeeker/RTseries documentation built on Dec. 31, 2022, 10 a.m.