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
.