lmRobMM(formula, data=<<see below>>, weights=<<see below>>, subset=<<see below>>, na.action=na.fail, model=F, x=F, y=F, contrasts=NULL, ...)
formula
object, with the response on the left side of the tilde (
~
) operator, and the terms separated by
+
operators on the right.
formula
,
subset
and
weights
arguments.
If
data
is missing, the variables in the
formula
should be on your search list.
This argument may also be a single number to handle some special cases; see the
help file for details.
weights
must be the same as the number of observations.
The weights must be nonnegative and it is recommended that they be strictly positive, since zero weights are ambiguous.
To exclude particular observations from the model, use the
subset
argument instead of zero weights.
model.frame
after any
subset
argument has been applied. The default is
na.fail
, which returns an error if any missing values are found. An alternative is
na.exclude
, which deletes observations that contain one or more missing values.
TRUE
, then the model frame is returned in the
model
component of the fitted object.
TRUE
, then the model matrix is returned in the
x
component of the fitted object.
TRUE
, then the response is returned in the
y
component of the fitted object.
the number of random subsamples to be drawn. The default is
4.6*2^ncol(x)
; see Yohai, Stahel, and Zamar (1991) for a justification of this value.
If
"Exhaustive"
resampling is chosen, the value of
nrep
is ignored.
See the
help file for a description of different resampling methods.
FALSE
, an error is returned when
x
is singular (i.e., when the model is over-determined). Otherwise, the fitting algorithm continues and returns a model object. By default,
singular.ok=TRUE
.
TRUE
, a list representing the QR-decomposition of
x
is returned. By default,
qr.out=TRUE
.
lmRobMM.robust.control
for the available control parameters and their default settings. By default,
robust.control=lmRobMM.robust.control
.
lmRobMM.robust.control
for a description of different resampling methods.
lmRobMM
is an approximation to the true solution based upon a resampling algorithm. Hence, you will obtain (slightly) different answers each time you make the same call.
Marazzi, A. (1993). Algorithms, Routines, and S functions for Robust Statistics. Wadsworth & Brooks/Cole, Pacific Grove, CA.
Yohai, V., Stahel, W.A., and Zamar, R.H. (1991). A procedure for robust estimation and inference in linear regression. In Stahel, W.A. and Weisberg, S.W., Eds., Directions in Robust Statistics and Diagnostics, Part II. Springer-Verlag.
oil.robust <- lmRobMM(Oil ~ Market, data=oilcity) summary(oil.robust)