Controlling Robust GLM Fitting by Different Methods

These are auxiliary functions as user interface for glmrob fitting when the different methods, "Mqle", "BY", or "MT" are used. Typically only used when calling glmrob.

glmrobMqle.control(acc = 1e-04, test.acc = "coef", maxit = 50, tcc = 1.345)
glmrobBY.control  (maxit = 1000, const = 0.5, maxhalf = 10)
glmrobMT.control  (cw = 2.1, nsubm = 500, acc = 1e-06, maxit = 200)

Arguments

acc

positive convergence tolerance; the iterations converge when ???

test.acc

Only "coef" is currently implemented

maxit

integer giving the maximum number of iterations.

tcc

tuning constant c for Huber's psi-function

const

for "BY", the normalizing constant ..

maxhalf

for "BY"; the number of halving steps when the gradient itself no longer improves. We have seen examples when increasing maxhalf was of relevance.

cw

tuning constant c for Tukey's biweight psi-function

nsubm

the number of subsamples to take for finding an initial estimate for method = "MT".

Value

A list with the arguments as components.

Author

Andreas Ruckstuhl and Martin Maechler

See also

glmrob

Examples

str(glmrobMqle.control())
#> List of 4
#>  $ acc     : num 1e-04
#>  $ test.acc: chr "coef"
#>  $ maxit   : num 50
#>  $ tcc     : num 1.34
str(glmrobBY.control())
#> List of 3
#>  $ const  : num 0.5
#>  $ maxhalf: num 10
#>  $ maxit  : num 1000
str(glmrobMT.control())
#> List of 4
#>  $ cw   : num 2.1
#>  $ nsubm: num 500
#>  $ acc  : num 1e-06
#>  $ maxit: num 200