Apache Spark 2.x Cookbook

Getting ready

Whenever someone has to take tennis lessons in the morning, the night before, the instructor checks the weather report and decides whether the next morning would be good to play tennis. This recipe will use this as an example to build a decision tree.

Let's decide on the features of weather that affect the decision whether to play tennis in the morning or not:

• Rain
• Wind speed
• Temperature

Let's build a table using different combinations of these features:

Rain	Windy	Temperature	Play tennis?
Yes	Yes	Hot	No
Yes	Yes	Normal	No
Yes	Yes	Cool	No
No	Yes	Hot	No
No	Yes	Cool	No
No	No	Hot	Yes
No	No	Normal	Yes
No	No	Cool	No

Now how do we build a decision tree? We can start with one of the three features: rain, wind ...