You may want to use the Jupyter docker stack, it give you everything you need to get started with Spark.