R help - Mar 2008 - Partition data into clusters

Greetings R-users,

I have been using the fpc package in R to cluster my data. Speficically I am
using kmeansruns clustering.
I would like to know how I use R to partition data into clusters. What I am
doing is as follows.

# Use csv file as input
#####################
wholeset = read.csv("Spellman800genesImputed.csv")

# exclude first col (gene names)
##########################
wholeset2 = wholeset[,-1]

#Use fpc
###########################
library(fpc)

cl.kmr10 <- kmeansruns(wholeset2,k=10,runs=10)


#append cluster label to original dataset
###################
cl2 <- data.frame(wholeset, cl.kmr10$cluster)

After this step, I write cl2 into a csv file and manually partition data
into its respective clusters using Excel.
Then I read the data from each clusters back into R for further analysis.

Can I do the data partitioning directly in R?

TQ


-- 
Suhaila Zainudin
PhD Candidate
Universiti Teknologi Malaysia

	[[alternative HTML version deleted]]

Can you supply a small self-contained example of this?
--- Suhaila Zainudin <suhaila.zainudin at gmail.com>
wrote:
> Greetings R-users,
> 
> I have been using the fpc package in R to cluster my
> data. Speficically I am
> using kmeansruns clustering.
> I would like to know how I use R to partition data
> into clusters. What I am
> doing is as follows.
> 
> # Use csv file as input
> #####################
> wholeset = read.csv("Spellman800genesImputed.csv")
> 
> # exclude first col (gene names)
> ##########################
> wholeset2 = wholeset[,-1]
> 
> #Use fpc
> ###########################
> library(fpc)
> 
> cl.kmr10 <- kmeansruns(wholeset2,k=10,runs=10)
> 
> 
> #append cluster label to original dataset
> ###################
> cl2 <- data.frame(wholeset, cl.kmr10$cluster)
> 
> After this step, I write cl2 into a csv file and
> manually partition data
> into its respective clusters using Excel.
> Then I read the data from each clusters back into R
> for further analysis.
> 
> Can I do the data partitioning directly in R?
> 
> TQ
> 
> 
> -- 
> Suhaila Zainudin
> PhD Candidate
> Universiti Teknologi Malaysia
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> 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.
>

Hello,

Referring to your email, I attaced a rar file containing a sample data file
and the R script for clustering the data.

After I finished the clustering process, I write the clustering result into
a csv file.


To partition the data, I open the csv file in Excel, sort the data according
to cluster number and put the clustered data into separate csv files.

I am trying to run the above step (which is currently done in Excel) in R.

Any comments are appreciated.



-- 
Suhaila Zainudin
PhD Candidate
Universiti Teknologi Malaysia



-- 
Suhaila Zainudin
PhD Candidate
Universiti Teknologi Malaysia

Hi,

Thanks for your reply. I have tried yr suggestions with success. TQVM.

I have another query, say I want to write each cluster into a csv file such
as follows:

clus1 <- my.clusters[[1]]
write.csv(clus1, file = "clus1.csv")
....
.....
.....
clus10 <- my.clusters[[10]]
write.csv(clus10,file = "clus10.csv")

I can write the functions to do that for all 10 clusters by repeatedly
calling write.csv as  above. Is there a more elegant way of doing it by
using a loop (for example).

for(i in 1:xx) {
clusn <- my.clusters[[n]]
write.csv(clusn, file = "clusn.csv")
}

I am trying something using list as well, as folows:

names( my.clusters ) <- paste('Cluster_',1:10, sep='')

Now I can  use

my.clusters$Cluster_1

The above will return all members of Cluster_1.

I have an idea to use lapply or sapply to do write.csv on each
components(Cluster_1.....Cluster_n)  from my.clusters, that will do the same
as the loop example. Maybe this is the better way..


Any comments are appreciated.

Thanks!

	[[alternative HTML version deleted]]