Run Spark over Hadoop

In my local machine, I have installed scala, spark, hadoop. I have done one sample program of word count using spark with maven. Now I have to do these things over Hadoop using spark?. How can I do this??