Agile Data Science

Chapter 3. Agile Tools

This chapter will briefly introduce our software stack. This stack is optimized for our process. By the end of this chapter, you’ll be collecting, storing, processing, publishing, and decorating data. Our stack enables one person to do all of this, to go “full stack.” We’ll cover a lot, and quickly, but don’t worry: I will continue to demonstrate this software stack in Chapters 5 through 10. You need only understand the basics now; you will get more comfortable later.

We begin with instructions for running our stack in local mode on your own machine. In the next chapter, you’ll learn how to scale this same stack in the cloud via Amazon Web Services. Let’s get started.

Code examples for this chapter are available at https://github.com/rjurney/Agile_Data_Code/tree/master/ch03. Clone the repository and follow along!

git clone https://github.com/rujrney/Agile_Data_Code.git

Scalability = Simplicity

As NoSQL tools like Hadoop, MongoDB, data science, and big data have developed, much focus has been placed on the plumbing of analytics applications. This book teaches you to build applications that use such infrastructure. We will take this plumbing for granted and build applications that depend on it. Thus, this book devotes only two chapters to infrastructure: one on introducing our development tools, and the other on scaling them up in the cloud to match our data’s scale.

In choosing our tools, we seek linear scalability, but above all, we seek simplicity. While the concurrent ...