density/clustering by [categorical] type?

Discussion created by dkhey on Aug 8, 2012
Hi folks,

Is there a function that will allow you to do a density (or get a feel of the distribution) of an event by categorical type?  I am trying to visualize data breach events by hackers, careless employees, employees that blatantly leak info, et cetera.  I can narrow these events down to city/state.  Just curious if these events cluster by type.

Thoughts?  I know I can do this one at a time by selecting by type and running, say, a kernel density (I have number of records lost) or just look at the data points without a quantitative value tied to them via a cluster analysis.  Is there a more elegant way to get this done?