Classes that contain the logic for updating Series values

We sometimes work with a particular dataset for an extended period of time, occasionally years. The data might be updated regularly, for a new month or year, or with additional individuals, but the data structure might be fairly stable. If that dataset also has a large number of columns, we might be able to improve the reliability and readability of our code by implementing classes.

When we create classes, we define the attributes and methods of objects. When I use classes for my data-cleaning work, I tend to conceptualize a class as representing my unit of analysis. So, if my unit of analysis is a student, then I have a student class. Each instance of a student created by that class might have birth date and gender attributes and a course registration method. I might also create a subclass for alumni that inherits methods and attributes from the student class.

Data cleaning for the NLS DataFrame could be implemented...