Data exploration is generally performed by presenting a meaningful synthesis of its distribution—it could be through a sequence of graphs, by describing it with a set of numerical parameters, or by approximating it with simple functions. Now let's explore different possibilities, and how to accomplish them with different tools in the SciPy stack.
The type of graph depends on the type of variable (categorical, quantitative, or dates).
When our data is described in terms of categorical variables, we often use pie charts or bar graphs to represent it. For example, we access the Consumer Complaint Database from the Consumer Financial Protection Bureau, at http://catalog.data.gov/dataset/consumer-complaint-database. The database was created in February 2014 to contain complaints received by the Bureau about financial products and services. In its updated version in March of the same year, it consisted of almost 300,000...