Skip to contents

Relative root mean squared error

Source: R/metrics-prediction.R
rrmse.Rd

Relative root mean squared error (RRMSE) expresses RMSE as a percentage of the absolute observed mean.

Usage

rrmse(obs, pred, na.rm = TRUE)

Arguments

obs

Numeric observation vector.

pred

Numeric prediction vector paired with obs.

na.rm

Logical; remove incomplete pairs?

Value

One numeric value.

Details

$$\mathrm{RRMSE}=100\, \frac{\sqrt{n^{-1}\sum_{i=1}^{n}(obs_i-pred_i)^2}}{|\bar{obs}|}.$$

RRMSE is non-negative and zero is ideal. It returns NA with a warning for a zero observed mean and is unstable when that mean is near zero. This mean-normalized convention differs from the standard-deviation normalization in nrmse(). Missing-value handling follows bias().

References

Willmott, C. J., Ackleson, S. G., Davis, R. E., Feddema, J. J., Klink, K. M., Legates, D. R., O'Donnell, J., and Rowe, C. M. (1985). Statistics for the evaluation and comparison of models. Journal of Geophysical Research, 90, 8995-9005. doi:10.1029/JC090iC05p08995

See also

Other prediction metrics: mape(), mdae(), mpe(), msle(), rae(), rer(), rmsle(), rpd(), rpiq(), sep(), smape(), willmott_d()