Single Degree-of-freedom Effects for an lm Object

DESCRIPTION:

Gets single degree-of-freedom effects for an lm object.

USAGE:

effects.lm(object, set.sign = F) 

REQUIRED ARGUMENTS:

object
the result of a fit from a linear model or any model derived from linear models, including aov, glm, and gam models. The response for some of these may have been multivariate, in which case the effects will be a matrix.

OPTIONAL ARGUMENTS:

set.sign
flag; if TRUE, the sign of the effects corresponding to coefficients in the model will be set to agree with the signs of the corresponding coefficients. Otherwise, the signs are determined by the algorithm for the orthogonal decomposition to promote numerical stability.

VALUE:

the vector or matrix (in the case of a matrix response) of effects. In the (default) case that the fitting was done with a Householder orthogonal decomposition, effects will be returned for all n observations in the model. If there were r degrees of freedom for the fit, the first r effects or rows of the effects matrix will correspond to those degrees of freedom.
Note that in the case of glm or gam models, the effects are defined on the linearized variable. In fact, effects.lm takes no account of these models, and just returns the effects for the last linear model in the iterated least-squares fit.
This is a method for the function effects() for objects inheriting from class lm.