R help - May 2009 - Negative value for adjustedRandIndex?

Hello, 

I am a very new user to R so please have patience with me. :clap:

I am trying to evalute the "internal response" for a couple of
different
cluster methods with the help of the AdjustedRandIndex, which is included in
the mclust package. 

However, I do get a bit puzzled when I get a negative value as the value
should be in intervall of [0,1], am I correct? Have I done something wrong?
Or should I interpret the negative value as the two methods do not group the
cases to the same cluster?

I have for example used k-means and k-median, and choosen five clusters. 
Each respondent has been labelled as either beloning to cluster 1,2,3,4 or
5. I have picked out the first 32 respondents. 

Example:


a   <- c(1,2,3,4,5,2,1,5,1,2,2,4,2,3,3,2,4,5,5,4,4,4,5,5,2,3,1,5,5,5,4,1)
b   <- c(5,5,5,2,5,3,2,2,2,5,3,4,5,5,2,5,2,3,5,5,5,3,5,3,3,2,4,5,2,2,2,3)
adjustedRandIndex (a,b)

[1] -0.01608168



Thanks in advance,

Pooka
-- 
View this message in context:
http://www.nabble.com/Negative-value-for-adjustedRandIndex--tp23652402p23652402.html
Sent from the R help mailing list archive at Nabble.com.

On Thu, 2009-05-21 at 05:28 -0700, Pooka wrote:
> Hello, 
> 
> I am a very new user to R so please have patience with me. :clap:
> 
> I am trying to evalute the "internal response" for a couple of
different
> cluster methods with the help of the AdjustedRandIndex, which is included
in
> the mclust package. 
> 
> However, I do get a bit puzzled when I get a negative value as the value
> should be in intervall of [0,1], am I correct?
0 indicates the agreement between your two raters is the same as the
agreement you'd expect to get as a result of random chance. A negative
value therefore indicates that the agreement between the two raters is
even lower than that.

G
>  Have I done something wrong?
> Or should I interpret the negative value as the two methods do not group
the
> cases to the same cluster?
> 
> I have for example used k-means and k-median, and choosen five clusters. 
> Each respondent has been labelled as either beloning to cluster 1,2,3,4 or
> 5. I have picked out the first 32 respondents. 
> 
> Example:
> 
> 
> a   <-
c(1,2,3,4,5,2,1,5,1,2,2,4,2,3,3,2,4,5,5,4,4,4,5,5,2,3,1,5,5,5,4,1)
> b   <-
c(5,5,5,2,5,3,2,2,2,5,3,4,5,5,2,5,2,3,5,5,5,3,5,3,3,2,4,5,2,2,2,3)
> adjustedRandIndex (a,b)
> 
> [1] -0.01608168
> 
> 
> 
> Thanks in advance,
> 
> Pooka
-- 
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
 Dr. Gavin Simpson             [t] +44 (0)20 7679 0522
 ECRC, UCL Geography,          [f] +44 (0)20 7679 0565
 Pearson Building,             [e] gavin.simpsonATNOSPAMucl.ac.uk
 Gower Street, London          [w] http://www.ucl.ac.uk/~ucfagls/
 UK. WC1E 6BT.                 [w] http://www.freshwaters.org.uk
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%