Naive2 model

Specify a Naive2 benchmark model for use with fabletools::model().

Usage

NAIVE2(formula, ...)

Arguments

formula

A model formula specifying the response and optional season() special, for example value ~ season(12).

...

Further arguments.

Value

A model definition that can be used inside fabletools::model().

Details

NAIVE2() implements the Naive2 benchmark from the M4 forecasting competition. The method tests the response for seasonality at the specified frequency.

If seasonality is detected, the response is adjusted using classical multiplicative decomposition. Naive forecasts are produced from the seasonally adjusted response and subsequently reseasonalized.

If seasonality is not detected, the method is equivalent to an ordinary naive forecast.

The season() special controls the seasonal frequency. When period = NULL, the frequency is inferred from the tsibble index. Alternatively, it can be specified explicitly, such as season(12) for monthly data.

See also

Other NAIVE2: fitted.NAIVE2(), forecast.NAIVE2(), model_sum.NAIVE2(), residuals.NAIVE2()

Examples

library(dplyr)
library(tsibble)
library(fabletools)

train_frame <- M4_monthly_data |>
  filter(series == first(series)) |>
  as_tsibble(index = index)

model_frame <- train_frame |>
  model("NAIVE2" = NAIVE2(value ~ season(12)))

model_frame
#> # A mable: 1 x 1
#>     NAIVE2
#>    <model>
#> 1 <NAIVE2>