If you want to do BCa intervals in a stratified or two-sample problem
you should use library(resample) (see bottom for download info).
The basic syntax would be either:
myboot <- bootstrap2(new.xy, mean, treatment = which.xy)
or
myboot <- bootstrap2(martinjn$ka, mean, data2 = martinjn$kb)
Then one of:
limits.bca(myboot, acceleration = 0)
limits.bca(myboot)
>Hello,
>
>I use the code beneath to estimate the bootstrapped difference of means:
>new.xy <- c(martijn$ka, martijn$kb)
>which.xy <- rep(1:2, c( length(martijn$ka), length(martijn$kb) ) )
>
>myboot <- bootstrap(new.xy, mean(new.xy[which.xy==1]) -
>mean(new.xy[which.xy==2]), group=which.xy, assign.frame1=F, 1000)
>myboot
>
>Furthermore, when I want to calculate the bias-corrected accelerated CI with
>limits.bca and set accelerator on NULL the follow error message is
>presented:
>Problem in jackstats(reps, observed, n.groups, func..: NA's encountered in
>replicates. Replicates stored as .bootstrap.replicates.
>Use traceback() to see the call stack
>Problem: internal error: trying to set internal exit action in frame 1, must
>be between 1 and 0
>Use traceback() to see the call stack
>
>However, when I set accelerator on 0, everything looks OK, i.e. the function
>runs normally but it will only produce bias-corrected CI's. How can I solve
>the problem and calculate bias-corrected accelerated CIs also?
>
>
>Thanks in advance,
>
>Martijn W Heymans, PhD, PT
>VU University Medical Center
> Amsterdam
>The Netherlands
>M: mw.heymans@vumc.nl
========================================================
| Tim Hesterberg Research Scientist |
| timh@insightful.com Insightful Corp. |
| (206)802-2319 1700 Westlake Ave. N, Suite 500 |
| (206)283-8691 (fax) Seattle, WA 98109-3012, U.S.A. |
| www.insightful.com/Hesterberg |
========================================================
Download the S+Resample library from www.insightful.com/downloads/libraries
|