Working with Zeppelin
The Zeppelin user interface consists of notes that are divided into paragraphs. Each paragraph consists of two sections: the code section where you put your source code and the result section where you can see the result of the code execution.
To use the Spark interpreter, you must specify the interpreter directive at the beginning of each paragraph, using the format % [INTERPRETER_NAME}. When you use the Spark interpreter, you can enter source code in Python, Scala, or SQL. So, the interpreter directive could be: %spark.sql.
When you run the code, Zeppelin sends the code to a backend processor such as Spark. The processor or service then returns results; you can then use Zeppelin to review and visualize results in the browser using the Settings toolbar: