err: Common error metrics

In twolodzko/twextras: twextras

Description

Common error metrics

Usage

err(y, yhat, wt = 1, method = c("MAE", "RMSE", "MdAE"))

Arguments

y, yhat	vectors of true and predicted values.
wt	weights for observations, uniform by default.
method	vector of names of error metrics to compute.

Details

The following error metrics are implemented:

method	function
ME	mean(y-yhat)
MAE	mean(abs(y-yhat))
RMSE	sqrt(mean((y-yhat)^2))
MAPE	mean(abs((y-yhat)/y))
MdAE	median(abs(y-yhat))
RMdSE	sqrt(median((y-yhat)^2))
MdAPE	median(abs((y-yhat)/y))
MRAE	mean(abs(y-yhat))/mean(abs(mean(y)-y))
MRRSE	mean((y-yhat)^2)/mean((mean(y)-y)^2)
wME	mean((y-yhat) * wt)
wMAE	mean(abs(y-yhat) * wt)
wRMSE	sqrt(mean((y-yhat)^2 * wt))

References

Shcherbakov, M. V., Brebels, A., Shcherbakova, N. L., Tyukov, A. P., Janovsky, T. A., & Kamaev, V. A. E. (2013). A survey of forecast error measures. World Applied Sciences Journal, 24, 171-176.

Examples

x <- rnorm(15)
err(x, mean(x))

twolodzko/twextras documentation built on May 3, 2019, 1:52 p.m.