And, finally, if you want to embed Mahout clustering in your web
application and the data all fits into memory or you can devise an
Iterator<Vector> to feed data into it, the ClusterIterator.iterate(...)
method will do k-means (also fuzzyK and Dirichlet) clustering in memory.
On 4/16/12 11:34 AM, Manuel Blechschmidt wrote:
Hello Oscar,
here you find an example how to bundle Mahout with a web application:
https://github.com/ManuelB/facebook-recommender-demo/blob/master/pom.xml
/Manuel
On 16.04.2012, at 17:28, Paritosh Ranjan wrote:
The simplest way would be :
1) Create a maven project.
2) Add mahout-core as a dependency.
3) Use KMeansDriver's run method.
If you set parameter runSequential=true, then you don't even need hadoop
cluster, but you will not be able to cluster really large datasets.
So, try it out with a smaller number of records(vectors) first, then go for
hadoop cluster ( by setting runSequential=false and setting up HADOOP_HOME) .
You can find the documentation for KMeans here:
https://cwiki.apache.org/confluence/display/MAHOUT/K-Means+Clustering
You can also start with Canopy Clustering and then go for K-Means, as its
simpler and fast.
https://cwiki.apache.org/confluence/display/MAHOUT/Canopy+Clustering
I recommend mahout 0.7-snapshot ( i.e. the current code at trunk ).
On 16-04-2012 10:57, OSCAR wrote:
Hello
My name is Oscar González, i'm studing System Engeineer in the universidad el
Bosque from Colombia. And i have the next question:
I have a web aplication with hibernate. And i need use clustering, kmeans
algorithm. I wanna use mahout, but I don't know, how can I apply mahout in my
project... I'm using netbeans. Please answer me,
Thanks
Oscar Miguel Gonzalez
Enviado desde mi iPad
