[R] weights in glmrob

[Spencer Graves]

	  Have you worked through the several examples in the 'glmrob' help 
page?  Those include at least one case that produces sensible standard 
errors and at least one that return NAs for standard errors.

	  If this does not answer your question, I have two further suggestions:

	  1.  Make a local copy of the 'glmrob' function.  Then skim the 
documentation for 'debug', try 'debug(glmrob)' and issue your problem 
call to 'glmrob(...)'.  This will allow you to walk through the 
algorithm line by line.  At each step, you can look at the values of any 
of the variables created, change them if you like, etc.  With luck and 
perserverance, this may produce enlightenment.

	  2.  If you would still like help from this listserve, please submit 
another post.  Before you do, however, PLEASE do read the posting guide! 
"www.R-project.org/posting-guide.html".  In particular, if your example 
had been completely self contained, I might have been able to say more. 
  As it is, I can't think of any other way I can help you.

	  Hope this helps.
	  Spencer Graves

Celso Barros wrote:
> Dear List members,
> 
> 
>          MY APOLOGIES IF YOU ALREADY RECEIVED THIS. A COLLEAGUE TOLD ME HE
> COULD NOT OPEN MY PREVIOUS MESSAGE.
> 
> 
>         I am runnning a logistic regression that includes sample weights
> (expressed by a variable called WEIGHT) using glmob (from robustbase). My
> model looks like this:
> 
>        B12<-glmrob(SERE~COMP+COM,family=binomial,weights=WEIGHT)
> 
>        When I include the weights, my standard error columns, as well as the
> p-values, show only NAs. If I exclude weights=WEIGHTS, everything works. I
> would like to use the weights. Does anyone know what could be the problem?
> 
>        Thanks in advance,
> 
>                                            Celso
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html