class: ARFIMA Forecast Class

Description

Class for the ARFIMA forecast.

Slots

forecast:

Object of class "vector"

model:

Object of class "vector"

Extends

Class "ARFIMA", directly. Class "rGARCH", by class "ARFIMA", distance 2.

Methods

as.array

signature(x = "ARFIMAforecast"): Extracts the forecast array with matrix column dimensions equal to the n.ahead value and row dimension 1 (series forecast), and array dimension equal to the number of rolling forecasts chosen.

as.data.frame

signature(x = "ARFIMAforecast"): Extracts the forecasts. Takes many additional arguments (see note below).

as.list

signature(x = "ARFIMAforecast"): Extracts the forecast list with all rollframes.

fpm

signature(object = "ARFIMAforecast"): Forecast performance measures.

show

signature(object = "ARFIMAforecast"): Forecast summary returning the 0-roll frame only.

Note

There are 3 main extractor functions for the ARFIMA object which is admittedly the most complex in the package as a result of allowing for rolling forecasts. The as.array extracts an array object where each page of the array represents a roll. The as.list method works similarly returns instead a list object. There are no additional arguments to these extractor functions and they will return all the forecasts. The as.data.frame method on the other hand provides for 4 additional arguments. The rollframe option is for the rolling frame to return (with 0 being the default no-roll) and allows either a valid numeric value or alternatively the character value “all” for which additional options then come into play. When “all” is chosen in the rollframe argument, the data.frame returned may be time aligned (logical option aligned) in which case the logical option prepad indicates whether to pad the values prior to the forecast start time with actual values or NA (value FALSE). Finally, the type option controls whether to return all forecasts (value 0, default), return only those forecasts which have in sample equivalent data (value 1) or return only those values which are truly forecasts without in sample data (value 2). Depending on the intended usage of the forecasts, some or all these options may be useful to the user when extracting data from the forecast object.

Author(s)

Alexios Ghalanos