Star coordinates is a tool for visualizing multi-dimensional data. Given a dataset, Star Coordinates allocates an axis on a 2D circle to each attribute in the dataset. Initially, all the lengths and the angles between the axes are equal. Additionally, each data entry is mapped on the 2D space by aggregating the effect of all the axes (attributes).
There are three ways to interact with star coordinates. Adjusting the length of an axis allows the user to increase and decrease the overal effect of a variable. The angle between any two attributes shows how they are related. Finally, the user has the option of turning off an axis to exclude an attribute from the analyis. For more information, refer to this paper.
Star coordinates provides an easy way to understand multi-dimensional data. It is good for data analytics especially in discovering clusters and multifactor analysis. Its intereactive features allow for quick manipulation that enables the user to gain insight on the data in the early stages of exploration.
Star coordinates uses coarse representation of data. In order to fit all entries in the 2D space, the data has to be normalized. As a result, no numerical analysis can be performed. Additionally, there are many ways to map to the same point in space. Therefore an entry's position on star coordinates does not say much about the entry. Star coordinates is also not easy to read and therefore mainly limited to data analytics.