A Guide To Charts

People often use “charts” and “graphs” to describe the largest family of visualizations. They may choose the term based on an academic convention. Or because they believe the distinction between them is important. Sometimes the terms complement one another. Sometimes they are synonymous. And sometimes people use them to make minor distinctions between types of visualizations.

Charts display many forms of analysis in a visual format: comparison, relationship, distribution, and composition. No two charts tell the same story. Consider the kind of data you want to present and the implications you want your audience to draw from that data before you decide what to use.

A chart is a representation of data in the form of a graph, diagram, map, or tabular format. This could make the other two families, Geospatial and Tables, subfamilies of it. We distinguish between them to help you identify when one works better for your data. Consider the most common Charts: Scatterplots, Bar Charts, Line Graphs, and Pie Charts. These chart types, or a combination of them, provide answers to most questions with relational data. They are the backbone of performing visual analysis on non-geospatial data.

Visualizing data with Charts relies on drawing points using cartesian coordinates (Ex. X, Y, Z) based on a set of dimensions and measures. Dimensions (Ex. categories, dates, etc.) group the measures (Ex. profit, deaths, temperature, etc.) for analysis. The measures are then rendered on the corresponding coordinates to create a visualization. Some types of visualizations excel at displaying many dimensions (Ex. Ordered Bar Charts), while others can only support a few with clarity (Ex. Pie Charts).

Each Chart type has its own strengths and weaknesses. But if used well people will understand the data better. When you pay attention to aesthetic conventions you can also make your visualization beautiful. You will combine both form and function to affect the viewer’s perception of the data.