Applies a fitted AR model to a new dataset.
# S3 method for AR
refit(object, new_data, specials = NULL, reestimate = FALSE, ...)A model for which forecasts are required.
A tsibble containing the time points and exogenous regressors to produce forecasts for.
(passed by fabletools::forecast.mdl_df()).
If TRUE, the coefficients for the fitted model will be re-estimated to suit the new data.
Other arguments passed to methods
A refitted model.
lung_deaths_male <- as_tsibble(mdeaths)
lung_deaths_female <- as_tsibble(fdeaths)
fit <- lung_deaths_male %>%
model(AR(value ~ 1 + order(10)))
report(fit)
#> Series: value
#> Model: AR(10) w/ mean
#>
#> Coefficients:
#> constant ar1 ar2 ar3 ar4 ar5 ar6 ar7 ar8
#> 572.0428 0.6739 -0.1584 0.1291 -0.1581 -0.0342 0.012 -0.1172 -0.0832
#> ar9 ar10
#> -0.0712 0.4145
#>
#> sigma^2 estimated as 37672
#> AIC = -93.6 AICc = -89.2 BIC = -68.55
fit %>%
refit(lung_deaths_female) %>%
report()
#> Series: value
#> Model: AR(10) w/ mean
#>
#> Coefficients:
#> constant ar1 ar2 ar3 ar4 ar5 ar6 ar7 ar8
#> 102807.6 0.6739 -0.1584 0.1291 -0.1581 -0.0342 0.012 -0.1172 -0.0832
#> ar9 ar10
#> -0.0712 0.4145
#>
#> sigma^2 estimated as 1.053e+10
#> AIC = 913.99 AICc = 913.99 BIC = 913.99