How it works...

This recipe really helps to demonstrate the value of the application-oriented approach. We continue to abstract out portions of code that perform singular tasks and use them as functions. As our library of functions increases, our analysis, often filled with repeated but slightly different tasks, becomes more composition than creation. Even better, these individual pieces are much easier to test and evaluate. Over time, with additional analyses, we will build a rich and fully customized library of tools that will drastically speed up future investigations.

This recipe also reveals how Python code can be created to construct more R-like analyses. As we performed further evaluations, we leveraged functions and tools that we had already built for our dataset and created new ones, such as the stacked area function that built off our older tools. However, unlike an analysis-oriented approach, these tools now exist in a data-specific library of code that can be used to build applications and reports, as we'll see in the next recipe.