[R] Any help with bootstrapping

~ DLC ~ dominique.couturier at unine.ch
Wed Feb 11 14:02:08 CET 2004

The distribution of your t* seems so asymmetric that the correction for 
the acceleration is big and extreme t* are selected as critical values 
of your CI. As BCa CI have the propriety of transformation invariance, 
a transformation of the scale makes no sense. (you could choose 
transformed bootstrap-t CI).
One possibility to avoid your problem with BCa Ci is to increase R, the 
number of replications. Try 1500 or more...
hope this help,

> Could someone help me on how to correctly try to correct this error
> message
> arning : BCa Intervals used Extreme Quantiles
> Some BCa intervals may be unstable
> Warning message:
> Extreme Order Statistics used as Endpoints in: norm.inter(t, adj.alpha)
> Regards
> IF

More information about the R-help mailing list