measures: Forecasting Accuracy Measure Calculation

In deseats: Data-Driven Locally Weighted Regression for Trend and Seasonality in TS

Forecasting Accuracy Measure Calculation

Description

Given point forecasts and observations, calculate various forecasting accuracy measures.

Usage

measures(preds, obs)

Arguments

preds	the point predictions for the test data period.
obs	the observation series (training data and test data)

Details

Given one-step-ahead rolling forecasts as well as the whole series of given observations (training together with test data), different forecasting accuracy measures (MAE, RMSE, Pearson's correlation, MASE, RMSSE) are being calculated.

Value

A named vector with the obtained criteria values is returned.

Examples

xt <- EXPENDITURES
xt_in <- window(xt, end = c(2017, 4))
yt <- log(xt_in)
est <- s_semiarma(yt, set_options(order_poly = 3), inflation_rate = "optimal")
fc_results <- predict(est, n.ahead = 8, expo = TRUE)
point_fc <- fc_results@pred
measures(point_fc, xt)

deseats documentation built on Sept. 11, 2024, 8:24 p.m.