When you execute this notebook, how many stages will be created in the backend?

In Spark, a stage is a set of tasks that can run in parallel without shuffling data. The number of stages depends on the number of shuffle boundaries in the DAG. For a notebook that performs two transformations followed by an action, you typically get two stages: one for the transformations (no shuffle) and one for the action (which may trigger a shuffle if a reduceByKey or join is involved). If the notebook contains multiple actions or multiple shuffle operations, each shuffle introduces a new stage. Therefore, the exact count is determined by the logical plan, but you can inspect it with the Spark UI or the explain() method.