tf_calc_metrics: Calculates different metrics for multiple models.

In flostracke/efor: efor: EasyFORcasting, a package assisting forecasting multiple timeseries

Description

This function calculates different metrics for all contained forecasts in a dataframe. The different models have to be seperated with the help of the key column.

Usage

tf_calc_metrics(forecasts, testset, metrics = c("rmse", "mae", "rsq",
  "mase", "mape"), detailed = FALSE)

Arguments

forecasts	The dataframe containing the forecasts.
testset	The dataframe containing the testset.
metrics	A list with the returned metrics. Following metrics are possible: RMSE Calculates the Root Mean Squared Error MAE Calculates the Mean Absolute Error RSQ Calculates the R Squared MASE Calculates the Mean Absolute Scaled Error MAPE Calculates the Mean Absolute Percent Error
detailed	If True the metrics for each article will be returned.

Details

The metrics are calculated with the help of the yardstick package (https://github.com/tidymodels/yardstick).

Value

The calculated metrics for each method.

flostracke/efor documentation built on June 5, 2019, 5:36 p.m.