jgperrin / net.jgp.books.spark.ch14

Spark in Action, 2nd edition - chapter 14 - extending data transformation with UDFs

Home Page:http://jgp.net/sia

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

This repository contains the Java labs as well as their Scala and Python ports of the code used in Manning Publication’s Spark in Action, 2nd edition, by Jean-Georges Perrin.

Spark in Action, 2nd edition – Java, Python, and Scala code for chapter 14

Welcome to Spark in Action, 2nd edition. Chapter 14 is about extending data transformation with UDFs (user defined functions).

This code is designed to work with Apache Spark v3.1.2.

Labs

Each chapter has one or more labs. Labs are examples used for teaching in the book. You are encouraged to take ownership of the code and modify it, experiment with it, hence the use of the term lab. This chapter has several labs.

Lab #200

Using a UDF using the dataframe API.

Lab #210

Using a UDF using SparkSQL.

Lab #900

A simple UDF to see if a value is in range.

Lab #910, #911, and #912

Attempts at using polymorphism with UDFs.

Lab #920

Passing an entire column to a UDF.

Datasets

Dataset(s) used in this chapter:

  • South Dublin (Republic of Ireland) County Council's libraries.

The OpenedLibrariesApp application does the following:

  1. It acquires a session (a SparkSession).
  2. It asks Spark to load (ingest) a dataset in CSV format.
  3. Spark stores the contents in a dataframe, then demonstrate hhow to use Custom UDF to check if in range.

Running the lab in Java

For information on running the Java lab, see chapter 1 in Spark in Action, 2nd edition.

Running the lab using PySpark

Prerequisites:

You will need:

  • git.
  • Apache Spark (please refer Appendix P - 'Spark in production: installation and a few tips').
  1. Clone this project
git clone https://github.com/jgperrin/net.jgp.books.spark.ch14
  1. Go to the lab in the Python directory
cd net.jgp.books.spark.ch14/src/main/python/lab200_library_open/
  1. Execute the following spark-submit command to create a jar file to our this application
spark-submit openedLibrariesApp.py

Running the lab in Scala

Prerequisites:

You will need:

  • git.
  • Apache Spark (please refer Appendix P - 'Spark in production: installation and a few tips').
  1. Clone this project
git clone https://github.com/jgperrin/net.jgp.books.spark.ch14
  1. Change directory

    cd net.jgp.books.spark.ch14

  2. Package application using sbt command

    sbt clean assembly

  3. Run Spark/Scala application using spark-submit command as shown below:

    spark-submit --class net.jgp.books.spark.ch14.lab200_library_open.OpenedLibrariesScalaApp target/scala-2.12/SparkInAction2-Chapter14-assembly-1.0.0.jar

Notes

  1. [Java] Due to renaming the packages to match more closely Java standards, this project is not in sync with the book's MEAP prior to v10 (published in April 2019).
  2. [Scala, Python] As of MEAP v14, we have introduced Scala and Python examples (published in October 2019).

Follow me on Twitter to get updates about the book and Apache Spark: @jgperrin. Join the book's community on Facebook or in Manning's live site.

About

Spark in Action, 2nd edition - chapter 14 - extending data transformation with UDFs

http://jgp.net/sia

License:Apache License 2.0


Languages

Language:Java 39.5%Language:Scala 36.0%Language:Python 24.5%