predAntilog: Unbiased Prediction of Log Transformed Response on Original...

In JVAdams/jvamisc: A Collection of Utility Functions

Description

Provide unbiased estimates on the original scale from an analysis of variance model with a log transformed response.

Usage

predAntilog(aovfit, xdata, logbase = exp(1), k = 0)

Arguments

aovfit	An object of class c("aov", "lm").
xdata	A data frame with predictor variables corresponding to those in model for which predictions should be made.
logbase	A numeric scalar, the base of the log transformation used in the transformed response of model, default exp(1).
k	A numeric scalar, the constant added to the response prior to transformation, default 0.

Value

A numeric vector of predicted values on the original scale of the response.

Examples

fit <- aov(log(yield) ~ block + N * P + K, npk)
predAntilog(fit, npk)

JVAdams/jvamisc documentation built on Aug. 11, 2021, 6:43 a.m.