Applying Python Swarm Plot to Gain Insights into Business Data

Business, large or small, create visuals to facilitate decision making about operation and management. The most commonly used visuals can be grouped into five types based on the purposes: comparison, time trend, part-whole relationship, correlation, and distribution. The classical charts for comparison are column and bar charts. Line charts are the best to illustrate the time trend. Pie charts, donut charts, tree map charts, and waterfall charts are used to illustrate part-whole relationship. Scatter plots are for correlation. Histogram, box plots, and violin plots are used to show distribution. Swarm plot is a new type of distribution chart that can be generated by Python and R. The swarm plot is more informative than other types of distribution charts, because it plots each of the data points on the chart, rather than just the summarized data. The following figure is a Python swarm plot for the survival rate in the Titanic tragic trip. The blue dots represent the survived passengers.