mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sharma, Avani" <agsha...@ebay.com>
Subject RE: DisplayCanopy class
Date Wed, 01 Dec 2010 01:03:58 GMT
Thanks. I am using trunk actually(was confused about the versioning) -  will update and build.


For arbitrary data, what should be the input to DisplayClustering - output from clusterdumper
which is anyway dumped in local file system ? 

My basic question is to be able to interpret clustering results better to be able to really
start using mahout. Like Weka has a way of showing where your clusters lie in a graphical
way and also tying back each data point to an object identifier. Say I have user profile data.
In that case, I want to be able to tie back the rows in a cluster to the user-id.


-----Original Message-----
From: Jeff Eastman [mailto:jeastman@Narus.com] 
Sent: Tuesday, November 30, 2010 4:09 PM
To: user@mahout.apache.org
Subject: RE: DisplayCanopy class

The DisplayClustering examples aren't designed to be run on Hadoop as they generate simple
Java graphics output. You should also be trying Mahout 0.4 or trunk as much has changed since
last spring. You will likely have to make changes to the display routines in order to plot
arbitrary data. Other than that I need more information on your data to be able to make better
recommendations.

-----Original Message-----
From: Sharma, Avani [mailto:agsharma@ebay.com] 
Sent: Tuesday, November 30, 2010 3:55 PM
To: user@mahout.apache.org
Subject: DisplayCanopy class


Hello,

I am using Mahout-0.3 to try out various clustering algorithms provided by mahout.
I ran arff converter, kmeans and canopy clustering over my data and used clusterdumper to
check out the clusters.
For big file sizes or big data to be clustered and interpreted, the output of clusterdumper
needs some kind of visualization.
So, I read https://cwiki.apache.org/confluence/display/MAHOUT/Visualizing+Sample+Clusters
I was able to generate a sample cluster using DisplayClustering class, but none other display
classes are working for me. I am pasting the error below.
Besides, are there any other ways to visualize the output from my data?
How will I feed clusters for my data into these classes in display package?
Please point me to any source of information for the same. Thank you.

~/mahout-0.3/trunk hadoop$bin/mahout org.apache.mahout.clustering.display.DisplayCanopy
Running on hadoop, using HADOOP_HOME=/usr/local/hadoop_live/
HADOOP_CONF_DIR=/usr/local/hadoop_live/conf
10/12/01 06:30:34 WARN driver.MahoutDriver: No org.apache.mahout.clustering.display.DisplayCanopy.props
found on classpath, will use command-line arguments only
Exception in thread "main" java.lang.IllegalAccessException: Class org.apache.hadoop.util.ProgramDriver$ProgramDescription
can not access a member of class org.apache.mahout.clustering.display.DisplayCanopy with modifiers
"public static"
        at sun.reflect.Reflection.ensureMemberAccess(Reflection.java:65)
        at java.lang.reflect.Method.invoke(Method.java:588)
        at org.apache.hadoop.util.ProgramDriver$ProgramDescription.invoke(ProgramDriver.java:68)
        at org.apache.hadoop.util.ProgramDriver.driver(ProgramDriver.java:139)
        at org.apache.mahout.driver.MahoutDriver.main(MahoutDriver.java:175)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
        at java.lang.reflect.Method.invoke(Method.java:597)
        at org.apache.hadoop.util.RunJar.main(RunJar.java:186)

Mime
View raw message