Transform Both Sides of a Nonlinear Regression Model

DESCRIPTION:

Allows users to transform both sides of a nonlinear regression model.

USAGE:

TBS(resp, pred, lambda) 

REQUIRED ARGUMENTS:

resp
the response vector in the original scale.
pred
the predicted response in the original scale.
lambda
the power parameter for the Box-Cox transformations.

VALUE:

the difference of the response and the prediction after applying the Box-Cox transformation to both. If pred has a gradient attribute, the returned object will also have a gradient attribute.
The Box-Cox transformations can only be applied to positive numerical data. They are defined as > Box.Cox function(y, lambda) if(lambda == 0) log(y) else (y^lambda - 1)/lambda and can be used to stabilize the variance in a nonlinear regression model.

REFERENCES:

Box, G. E. P. and Cox, D. R. (1964). The analysis of transformations. Journal of the Royal Statistical Society, Series 26, 211-252.
Carroll, R. J. and Ruppert, D. (1988). Transformation and Weighting in Regression. London: Chapman and Hall.