Hi Jamal,
By specification K-Means will do exclusive clustering i.e., each
point can be associated with a unique cluster, unless you are using a Fuzzy
K-Means clustering algorithm which tries to generate the overlapped
clusters.
Your output.txt may contain some additional information other than
the actual clusters (for ex, while doing text clustering there can be Top
terms associated with a specific cluster).
Can you attach your output.txt file.
Best,
Mahesh Balija,
CalsoftLabs.
On Tue, Jan 29, 2013 at 5:44 AM, jamal sasha <[email protected]> wrote:
> Hi,
> I am trying to perform simple kmeans clustering.
> I followed the tutorial here:
>
> https://cwiki.apache.org/confluence/display/MAHOUT/Clustering+of+synthetic+control+data#Footnote1
>
> So I was analyzing the output
>
> I am curious
> My input has 600 lines
> while the output
> wc -l output.txt
> gives
> 1224 output.txt
>
> Are the points shared across various clusters?
>
