As we stated in Chapter 1, The Predictive Analytics Process, EDA is a combination of numerical and visualization techniques that allow us to understand different characteristics of a dataset, its features, and the potential relationships between them.
Keep in mind the goal of this phase: to understand your dataset. The goal is not to produce summary statistics, pretty visualizations, or complex multivariate analysis. These are simple activities that accomplish the ultimate goal of data understanding.
Also, please don't confuse calculation with understanding. Anyone can calculate the standard deviation of a numerical feature; it can be done (for example) with the std() pandas Series method. Your job here is to use that number to understand your features and your dataset better.
Another example—after reading the definitions of symmetric and skewed distribution...
