How is the “hash” function used when splitting a dataset into ML training and test set in Pyspark?

Question

I am going through a tutorial on how to apply ML in Pyspark. In particular, the tutorial starts with a dataset named iris_dataset, which has 150 rows, each represen