Book Image

Learning Google BigQuery

By : Thirukkumaran Haridass, Mikhail Berlyant, Eric Brown
Book Image

Learning Google BigQuery

By: Thirukkumaran Haridass, Mikhail Berlyant, Eric Brown

Overview of this book

Google BigQuery is a popular cloud data warehouse for large-scale data analytics. This book will serve as a comprehensive guide to mastering BigQuery, and how you can utilize it to quickly and efficiently get useful insights from your Big Data. You will begin with getting a quick overview of the Google Cloud Platform and the various services it supports. Then, you will be introduced to the Google BigQuery API and how it fits within in the framework of GCP. The book covers useful techniques to migrate your existing data from your enterprise to Google BigQuery, as well as readying and optimizing it for analysis. You will perform basic as well as advanced data querying using BigQuery, and connect the results to various third party tools for reporting and visualization purposes such as R and Tableau. If you're looking to implement real-time reporting of your streaming data running in your enterprise, this book will also help you. This book also provides tips, best practices and mistakes to avoid while working with Google BigQuery and services that interact with it. By the time you're done with it, you will have set a solid foundation in working with BigQuery to solve even the trickiest of data problems.
Table of Contents (9 chapters)
Free Chapter
Google Cloud and Google BigQuery

Functions for transformation

Many of the functions discussed so far in this chapter can be used for the sanitation/transformation phase of data warehousing. Let's look at some of these functions with the specific task of sanitization/transformation:

  1. Decoding of encoded values. The CASE function can be used to create columns based on conditions of other columns:
WHEN region = "W" THEN "West"
WHEN region = "E" THEN "East"
END AS region
  1. Calculation of values. Any of the arithmetic functions can be used in this case:
 SUM(revenue) / COUNT(orders) AS average_order_value
  1. Splitting delimited string values into individual columns. The REGEX_EXTRACT() function can be used to extract individual parts of a string. Here is a how the function can be used to pull the value prior to the first space: