1) downloads spark-x.x.x-bin-hadoopx.x.tgz from offical website

and untgz to your path :

such as D:\google_downloads\spark-2.0.0-bin-hadoop2.7

here we call \Path_spark for short

2.1) install environment_path: append '/Path_spark/bin' to 'Path' environment_path var

2.2) add SPARK_HOME : set /Path_spark as new environment_var SPARK_HOME

3) copy file /Path_spark/python/pyspark to /your_python_Lib_path/site-packages

4) maybe should install py4j as usual method: run ' pip install py4j ' in cmd

5) now we can import pyspark in python_shell or relate python_IDE

---

success in python 3.5.1 ,spark 2.0.0 ,jdk 1.8.0_45