[R] Efficient sampling from a discrete distribution in R

Robin Hankin r.hankin at noc.soton.ac.uk
Tue Sep 4 10:00:34 CEST 2007

On 4 Sep 2007, at 08:40, Issac Trotts wrote:

> On 9/3/07, Prof Brian Ripley <ripley at stats.ox.ac.uk> wrote:
>> Yes, vectorize, and don't grow vectors. See 'S  
>> Programming' (details in
>> the R FAQ).
> It's easy to avoid growing the samples vector but it doesn't seem to
> make much difference.  Vectorizing would be probably help, but in this
> algorithm each step depends on the one before it.  Do you know of an
> alternative to the Chinese restaurant process that is vectorizable?

The Chinese restaurant process is simulated in the untb package,
in what I believe is a reasonably efficient manner.

[I find to my horror that "help.search("Chinese restaurant") returns
empty-handed.  I will add a suitable \concept{} entry to untb.Rd]

Best wishes


>> We've no idea about you (the R posting guide does ask for a signature
>> block), but if this is homework some people are going to be annoyed.
> No, it's for my job at Google, so maybe I should use my work email
> address to avoid confusion.  Sorry if the questions are so basic that
> they sound like homework, but you and others on the list know things
> that aren't easy to find in the documentation.
> -- 
> Issac Trotts
> Web Operations Chair
> Silicon Valley Web Builder
> http://svwebbuilder.com
> ______________________________________________
> 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
> and provide commented, minimal, self-contained, reproducible code.

Robin Hankin
Uncertainty Analyst
National Oceanography Centre, Southampton
European Way, Southampton SO14 3ZH, UK
  tel  023-8059-7743

More information about the R-help mailing list