1. download spark from spark.apache.org
2. download hadoop from hadoop.apache.org
3. download hadoop.dll and winutils.exe for hadoop 2.6 from github, link is here: https://github.com/sdravida/hadoop2.6_Win_x64, copy these files into hadoop/bin
4. add env SPARK_HOME & HADOOP_HOME
5. now you could run the example:

d:\spark\bin\run-example.cmd SparkPi 10