The first step when tackling any project that involves datasets is to understand the data. A common and easy approach is to plot a few distribution graphs, such as histograms, boxplots, pie charts, etc. To uncover relationships between variables, we can use scatter plots, line plots, or even 3D contour plots. These techniques are easy to implement and the graphs are to interpret. However, they usually fail to uncover other important characteristics of a dataset, such as geographical variations.
This project aims to demonstrate the power of spatial visualization in preliminary data analysis. Using powerful graphics packages such as Matplotlib, Seaborn, and Basemap available in Python, we can create informative spatial visualization of a dataset. This could help us gain a deeper understanding of the data and uncover interesting variable characteristics.
This post shows how you could quickly create a spatial visualization graph. I use the IRS 2017 U.S. Tax Return dataset as an example. After reading the notebook, you will see how we could change the visualization of the dataset from this messy and relatively uninformative graph:
to this clean and informative graph:
