What are the required steps for going from local standalone mode in Jupyter using Pyspark to using Yarn-client mode? I am specifically looking for an example of SparkSession set