Exploratory data analysis is mainly guided by visualizations, and pandas provides a great interface for quickly and effortlessly creating them. A simple strategy when beginning a visualization of any dataset is to focus only on univariate plots. The most popular univariate plots tend to be bar charts for categorical data (usually strings) and histograms, boxplots, or KDEs for continuous data (always numeric). Attempting to analyze multiple variables at the same time, directly at the start of a project, can be quite overwhelming.
Visualizing the flights dataset
Getting ready
In this recipe, we do some basic exploratory data analysis on the flights dataset by creating univariate and multivariate plots directly with pandas.
...