The 9th Day of Analytics – Simplify Your Life

Over the past eight days my colleagues have introduced exciting new GeoIQ features that allow anyone to perform powerful analytics against their data, but despite their wide applicability and ease of use not one of  the new analytic features is capable of automatically providing answers . . . after all is said and done, you still have to think about the results!  Fret not, because we built a new Simplify feature to address this shortcoming!  After you’ve performed an analysis all you need to do is press the “simplify” button and a pleasant female voice will explain what the results mean in plain, perfectly dictated English.  You can also feed it a question in any language (including Bachi) and it will return the correct answer – displayed as vectorized text on a map, of course. We think this is gonna be huge.

All seriousness aside – where Simplify really shines is in its ability to turn complicated geographic shapes into less complicated geographic shapes.  How is this useful?  Let’s say I’m trying to analyze a dataset of retail store sales that have been aggregated up to Virginia zip code boundaries.  Zip code boundary datasets tend to contain a large number of geographic shapes of relatively high complexity, and as anyone who has worked in Geoland will attest it can be time consuming to render a large number of highly detailed shapes.  If I’m trying to analyze my Virginia zip code data I may not want to wait for all of detail to render because I’m only interested in visualizing/analyzing my data at a macro level, and I’ll never zoom in close enough to care about every minute detail.

Simplify will analyze every minute detail of those geographic shapes, apply a little mathematical wizardry, and spit out new shapes that are less complicated approximations of the originals.  You even get to influence the complexity of the resulting shapes by choosing a tolerance level in various units of distance.  For instance, I might set my tolerance level to 2 km., which will cause the Simplify feature to rejigger the geometry around any line that is less than 2 km. in length.  After simplifying a Virginia zip code dataset I can visualize my retail sales analysis much more quickly than if I had chosen to use the more complicated boundaries.

Note that the two maps look identical when I’m zoomed out to the entire state of Virginia .  The map that uses the simplified Virginia zip code boundaries took noticeably less time to render, however, so when I pass it around to my co-workers or embed it into a website they will be greeted with faster load times than if I had used the complicated boundaries.  Let’s take a closer look at the simplified features by zooming in to zip code 24487:

---

---

In the bottom image the boundary for zip code 24487 only has around 7 visible lines, whereas there appear to be dozens in the unsimplified image.  Each of these additional lines adds to the complexity and rendering time of the geographic shape, but the differences are hardly noticeable when you’re zoomed out.   So the lesson learned:  utilize the Simplify feature to, well, simplify your geographic shapes when analyzing data that spans across a large number of boundaries.  You can also use it to turn things into triangles.

NOTE:  The entire first paragraph is a joke