R package implementing Multivariate Error Measures for Time Series forecasting.
The error measures (MSE, MASE, MAPE, MASE, SMAPE, WAPE) are defined as a multivariate, aggregated version of their univariate counterpart. The considered error measures are discussed in detail in the paper: Hyndman, R. J., & Koehler, A. B. (2006). Another look at measures of forecast accuracy. International journal of forecasting, 22(4), 679-688.
The preferred way to install this package is using devtools:
devtools::install_github("jdestefani/MEMTS", upgrade_dependencies = FALSE)
To see the full list of exported functions:
library("MEMTS")
ls("package:MEMTS")