RecursiveMSE: Recursive MSE

Description Usage Arguments Value See Also

Description

This function calculates the recursive MSEs of the forecasts made by two models, one of which is the benchmark. Recursive MSE is calculated as the mean squared errors of the point forecasts by adding one point forecast at a time, starting with the first out-of-sample forecast. One can use this to see how the MSEs of the two models evolve relative to each other as more out-of-sample forecasts are made.

Usage

1
RecursiveMSE(forecast.main, forecast.benchmark)

Arguments

forecast.main

an onject of class "Maeforecast". The forecasts that are to be compared against the benchmark.

forecast.benchmark

an object of class "Maeforecast". The forecasts made by the benchmark model. Note that the two sets of forecasts should be made with the same window, window size and forecasting horizon.

Value

This function returns an object of class "RecMSE" that contains the recursive MSEs of the two sets of forecasts, as well as their differences and ratios.

See Also

plot.RecMSE


google-trends-v1/gtm documentation built on June 5, 2019, 5:13 p.m.