Residuals in MCMC

[MCBLACKWELL]

I'm a little confused by two lines in the Rank2.txt sample macro included with MLWin and discussed at p56 of the MCMC manual. I wondered if someone could explain the rationale, please?

The macro is to be ran after a monitoring chain of length 5001 (p52). The two lines say:

note stack 2.5% pts to c368
pick 251 c303-c367 c368

As the chain has been sorted this picks the 251st smallest value in the chain. However, as I see it, this picks the 5% quantile (250/5,000=0.05) rather than the 2.5% one as the note suggests. Accordingly, with a similar line for the upper quantile it delimits the 90 percent central confidence interval?

Should it say "pick 126", or am I making some error in my though process above?

Many thanks indeed for your help!

Michael

[billb]

Hi Michael,
Rather impressively you are the first person to spot this mistake - yes the elements given give 90% not 95% confidence intervals. If/When I get around to updating the manual next I'll try and amend this.
Thanks,
Bill.