In this chapter, we introduced the nominal, ordinal, and numerical feature types and their importance for data analysis. When presented with a dataset, we demonstrated how to consult the data dictionary and the data itself to determine the feature types for each column. We also explained how the storage type is not to be confused with feature type. Since much of EDA is carried out with statistical graphs, we described how to recognize and interpret the shapes and patterns that emerge and connect these to the data being plotted. Finally, we provided guidelines for how you might conduct an EDA, and provided an example.
In Chapter 11, we provide a style guide for how to create informative, effective, and beautiful graphs. Many of the ideas in that chapter have been introduced and followed here, but we have not called attention to them. Our focus in this chapter instead has been on “reading” visualizations.