As geographers and data enthusiasts we're pretty excited about the functionality available in the new service called CartoDB. Rather than simply providing a platform for serving geographic data, CartoDB offers powerful features for both displaying and interacting with spatial and non-spatial data.
To illustrate some CartoDB functionality we've assembled Census data on the foreign-born population at the Census tract and the county level (similar to the data used in our D3 example). Details on the data itself are below but in general we've color-coded Census tracts (which you can choose to represent as centroids or polygons) based on the percentage of the population born in other countries.
As you might expect, the tracts with the highest percentage of people born in other countries tend to be located in New York City (especially Queens). But you can also see some other interesting patterns. If you limit to the Top 5% of tracts based on population born in Asia you can see a tract in Buffalo and one in Rochester. For those from Eastern Europe, Utica has five tracts in the top 5% -- four of which have more than 20% of the population from Eastern Europe.
CartoDB is not limited to rendering data on a map. In this example, we interactively query our CartoDB database and render the returned data using D3.
What makes CartoDB powerful is the use of a PostGIS enabled PostgreSQL database. What this means to developers (and ultimately to end users) is that the data can be queried, updated, filtered and styled on the fly and can be rendered in raster or vector formats. With PostGIS queries geographic features can be filtered, for example, based on the geographic relationship with other features. The PostgreSQL back end also means that the CartoDB platform can be used for more than just serving geographic data. You can also query data in more traditional ways.
We're also pleased that CartoDB supports Internet Explorer down to version 7. We do have clients that use or require support for both IE7 and IE8.
Data was downloaded from the Census FactFinder2 web site. We used the American Community Survey 2006-2010 data, variable B05006. Initial processing was done in R statistical software. The data was joined to tract and county shapefiles and zipped for upload to CartoDB.
*We're using the term "Middle East" because it's more familiar but the Census defines the region of "Western Asia" as encompassing Iraq, Israel, Jordan, Kuwait, Lebanon, Saudi Arabia, Syria, Yemen, Turkey, Armenia and "Other Western Asia".
Thanks to CartoDB support for quick and helpful responses to our queries. If you have questions about this example please don't hesitate to contact us at firstname.lastname@example.org