Hi,
On Fri, Mar 26, 2010 at 7:12 AM, karuna m wrote:
> hi all,
> I am doing hierarchical clustering using similarity measures for binary data
> using package ade4 and hclust function. For method=8 and method = 9 of
> dist.binary, I am getting Na values. Hence, hclust function is giving error
>
hi all,
I am doing hierarchical clustering using similarity measures for binary data
using package ade4 and hclust function. For method=8 and method = 9 of
dist.binary, I am getting Na values. Hence, hclust function is giving error as
Error in hclust(d8, method = "ward") : NA/NaN/Inf in foreig
Hi Karuna,
You can try vegdist() function in the vegan package. The function
computes dissimilarity indices that are useful for or popular with community
ecologists. All indices use quantitative data, although they would be named
by the corresponding binary index, but you can calculate the bina
The 'Gower' metric is one that is commonly used.
Rod
On Thu, Oct 29, 2009 at 10:22 AM, karuna m wrote:
> I am doing hierarchical clustering with cluster package. I couldnot find
> similarity measures like matching coefficient, Jaccard coefficient and sokal
> and sneath. Could anyone please te
The 'Gower' metric is one that is commonly used.
R
On Thu, Oct 29, 2009 at 10:22 AM, karuna m wrote:
> I am doing hierarchical clustering with cluster package. I couldnot find
> similarity measures like matching coefficient, Jaccard coefficient and sokal
> and sneath. Could anyone please tell
I am doing hierarchical clustering with cluster package. I couldnot find
similarity measures like matching coefficient, Jaccard coefficient and sokal
and sneath. Could anyone please tell package with similarity measures for
binary data?
kind regards,
Ms.Karunambigai M
PhD Scholar
Dept. of Bios
Hey there!
I would like to justify the stability of the cluster of a subset of my data by
comparing it to another cluster of another subset. Does there exist a
quantitative similarity measure that can be applied?
I am open for any suggestions,
thx for your help,
Julia
--
_
7 matches
Mail list logo