Writing Maintainable Code

In this chapter, we will continue the example that we started in the previous chapter, and we will guide you to configure dbt and write some basic, but functionally complete, code to build the three layers of our reference architecture – staging/storage, refined data, and delivery with data marts.

We will begin by introducing you to some core principles to reach the goal of building a data platform that is easy to understand and maintain, and then we will start to put the principles into practice to build and configure the layers of our reference architecture.

At each step, we will analyze the key elements of the models that we write to transform the raw source data into usable information in a data mart with facts and dimensions. Facts tell us about what happened, while dimensions provide extra descriptive information about the entities involved in a fact. First, we will build the pipeline for a fact table, and then, in the following chapter...