Sometimes a student may use a self explained chart, instead of a boring table for showing outcomes in a research paper. Yet, graphs are efficient in showing the broad picture of an issue and also for present results. In political science, you can getting into this topic reading Kastellec and Leoni (2007), for instance.
I usually like draw maps to present electoral results or to show correlation whatever. In this scrap, however, I show a clue for show regression estimates in a nice cartography using R – a free statistical package.
To plot the chart with OLS coefficients I need first get them. So, I repeated the following command for each of those 27 Brazilian states: