The function estimates several accuracy metrics to evaluate the accuracy of point forecasts. Either along the forecast horizon or along the test-splits. By default, the following accuracy metrics are provided:
ME
: mean error
MAE
: mean absolute error
MSE
: mean squared error
RMSE
: root mean squared error
MAPE
: mean absolute percentage error
sMAPE
: scaled mean absolute percentage error
MPE
: mean percentage error
rMAE
: relative mean absolute error
make_accuracy(
future_frame,
main_frame,
context,
dimension = "split",
benchmark = NULL
)
A tibble
containing the forecasts for the models, splits, etc.
A tibble
containing the actual values.
A named list
with the identifiers for seried_id
, value_id
and index_id
.
Character value. The forecast accuracy is estimated by split
or horizon
.
Character value. The forecast model used as benchmark for the relative mean absolute error (rMAE).
accuracy_frame is tibble
containing the accuracy metrics.