You can do this by simulation:
Generate data from a multinomial of the same length as your data (the sample
function can help) using either theoretical or observed probabilities.
Measure the length of the longest run, or the number of runs (the rle function
can help).
Repeat this a bunch of times (the replicate function can help)
See how your observed data compares to the simulations.
Hope this helps,
--
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
greg.snow at imail.org
801.408.8111
> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of liujb
> Sent: Thursday, September 24, 2009 8:52 AM
> To: r-help at r-project.org
> Subject: [R] multinormial runs tests?
>
>
> Dear R users,
>
> I would like to test the randomness in a series of N values (N>=2). I
> know
> that runs.test works for dichotomous factor only:
> x <- rep(c(1,2), 50)
> runs.test(factor(x))
>
>
> However it doesn't work for series that can take any N values (N>2):
> x <- rep(c(1,2,5,4),50)
> runs.test(factor(x))
>
> Error in runs.test(factor(x)) : x does not contain dichotomous data
>
> Are there any R function that does multinormial runs test?
>
> Thank you very much,
> sincerely,
> Julia
>
>
> --
> View this message in context: http://www.nabble.com/multinormial-runs-
> tests--tp25574075p25574075.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> R-help at r-project.org 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.