sjmisc (version 1.0.2)

rmse: Compute root mean squared error (RMSE)

Description

Compute root mean squared error of fitted linear (mixed effects) models.

Usage

rmse(fit, normalized = FALSE)

Arguments

fit

a fitted linear model of class lm, merMod (lme4) or lme (nlme).

normalized

logical, use TRUE if normalized rmse should be returned.

Value

• The root mean squared error of fit; or the normalized root mean squared error of fit if normalized = TRUE.

References

• http://en.wikipedia.org/wiki/Root-mean-square_deviation{Wikipedia: RMSD}
• http://www.theanalysisfactor.com/assessing-the-fit-of-regression-models/{Grace-Martin K: Assessing the Fit of Regression Models}

See Also

cv

Examples

data(efc)
fit <- lm(barthtot ~ c160age + c12hour, data = efc)
rmse(fit)

library(lme4)
fit <- lmer(Reaction ~ Days + (Days | Subject), sleepstudy)
rmse(fit)

# normalized RMSE
library(nlme)
fit <- lme(distance ~ age, data = Orthodont) # random is ~ age
rmse(fit, normalized = TRUE)