Automate Data Cleaning with User-Defined Functions, Classes, and Pipelines

There are a number of great reasons to write code that is reusable. When we step back from the particular data-cleaning problem at hand and consider its relationship to very similar problems, we can actually improve our understanding of the key issues involved. We are also more likely to address a task systematically when we set our sights more on solving it for the long term than on the before-lunch solution. This has the additional benefit of helping us to disentangle the substantive issues from the mechanics of data manipulation.

We will create several modules to accomplish routine data-cleaning tasks in this chapter. The functions and classes in these modules are examples of code that can be reused across DataFrames, or for one DataFrame over an extended period of time. These functions handle many of the tasks we discussed in the first eleven chapters, but in a manner that allows us to reuse our code...