This recipe covers a small but fundamental part of EDA: the collection of metadata and univariate descriptive statistics in a routine and systematic way. It outlines a common set of tasks that can be undertaken when first importing any dataset as a pandas DataFrame. This recipe may help form the basis of the routine that you can implement when first examining a dataset.
Metadata describes the dataset, or more aptly, data about the data. Examples of metadata include the number of columns/rows, column names, data types of each column, the source of the dataset, the date of collection, the acceptable values for different columns, and so on. Univariate descriptive statistics are summary statistics about individual variables (columns) of the dataset, independent of all other variables.