change toolkit to rabitlearn

rabit-learn/kmeans/kmeans_hadoop.sh

@@ -0,0 +1,9 @@
+#!/bin/bash
+if [ "$#" -lt 5 ];
+then
+    echo "Usage: <nslaves> <input_data> <ncluster> <max_iteration> <output>"
+    exit -1
+fi
+#set path to hadoop streaming jar here
+STREAMING_JAR=
+python ../tracker/rabit_hadoop.py -hs $STREAMING_JAR -n $1 -i $2 -o $5  kmeans.rabit stdin $3 $4 stdout