Tuning Constant for Bounded Influence Estimator

DESCRIPTION:

Returns one tuning constant for bounded influence estimator, given the asymptotic efficiency.

USAGE:

lmRobBI.eff(eff, method="sch", tuning.p=NULL, nx=1, intercept=T, sigmax=1) 

REQUIRED ARGUMENTS:

eff
the asymptotic efficiency required.

OPTIONAL ARGUMENTS:

method
a character vector, which specifies the type of bounded influence estimator. If method="sch", the Schweppe-type or Krasker-Welsch type bounded influence estimator is used; if method="mal", the Mallows-type bounded influence estimator is used.
tuning.p
a scalar which gives the value of the tuning constant for the psi-function of the corresponding generalized M-estimator. It is only used if method="mal".
nx
the number of explanatory variables used in the regression, not including the intercept.
intercept
logical flag: if TRUE, an intercept will be fitted by the regression.
sigmax
a scalar which gives the standard deviation of the model matrix x.

VALUE:

the tuning constant of the optimal weight function which will have the required asymptotic efficiency.

SEE ALSO:

, .