Hi, I am new to maout and hadoop , but am trying to run canopy clustering on the sequence file , I just generated by a piece of Java code . I first ran the
mahout canopy -i ........ on a cloudera vm environment with my HADOOP_HOME and CLASSPATH set , and it worked fine and gave me the clusters. I tried to run the same way in a hadoop dev cluster at my company , and it started complaining about the ClassNotFoundException - org.apache.mahout.math class. I went through this url http://comments.gmane.org/gmane.comp.apache.mahout.user/7469 , and the discussion here , and even tried to tweak the mahout.sh file to include these two lines export MAHOUT_JAR= MAHOUT_HOME=path of my mahout extracted directory. but I could not follow the solution to get past it ,I am not using any maven or ant tool to get my build , I just wrote a small piece of java code , which read my customizable file , converted them to vectors and wrote them to a Sequence file , I then transported that sequence file to the HDFS and tried to invoke the K means clustering algorihtm using the procedure mentioned. I shall great;y appreciate , if somebody can help me with it . Thanks, Abhik
