While cooped up inside over the weekend due to rain (with low temps in the 70′s!) I decided to play around some more with maps on Google Fusion Tables. The image above is a screenshot from the results of my work – a fully interactive google map of Burkina Faso departments (analogous to US counties). Darker colors represent a higher population and it’s quite evident that the Kadiogo department (containing the capital of Ouagadougou smack dab in the middle of the country) is the most populated. WordPress isn’t yet set up to allow me to embed the map here on the site (Fusion Tables is a bit too new apparently), however, I’ve made the raw data available to the public where you can click around on the map to your heart’s content and/or make your own mashup. You can create views to look at the data by province or region as well (just aggregate on either field), or just check out my separate provincial or regional boundaries/populations data tables.
Rather than inserting simple points like on my borrower map, this time I was able to map areas. I found GPS boundaries for Burkina Faso’s 13 regions and 351 departments here, and for it’s 45 provinces through some more searching here. Google Fusion Tables accepts .kml files*and I was able to easily add the data online. Then I made the data more useful by including other metadata about each department/province/region and stripping out extraneous points. Now, for example, you can click on a a department and see it’s associated, province, provincial capital, region, and regional capital. The trickiest part was using Wikipedia’s list/maps of Burkina Faso departments to be able to give a names to all 351 (Wikipedia only had 301). Also, .csv files don’t like french accents, so there was some cleanup there as well. I eventually tracked down some available Burkina Faso population data from a 2006 census report as well as the official breakdown of departments.
You could stop there, but I found the slick FusionTablesLayer Builder tool to be able to modify the look of the map (size, color, etc) and also be able to add a drop down or search box to allow users to filter results for when I put the map on the web.
While researching all of this, I found a few other tools that I plan on using in the future, such as this example on an on-map legend. Also, I found that it’s easy to add charts to the info windows by using this chart tool to modify and create chart URLs. Simply make a new column in your data and put in the custom URL for each record. I also stumbled across Google Refine, which will greatly cut down the time needed to play around with messy data (like the Wikipedia list of departments). Check out some demo videos here. The smart grouping tool (where the program will automatically make groups of like names, such as “FBF” and “Friends of Burkina Faso”) alone would have saved me weeks of man hours at my last job.
So, what will I do with all of this you ask? This exercise was a sort of proof of concept, and I think I’ve shown that these tools make it very easy to create interactive maps and charts all dynamically based off a fusion table, all with very little or no coding. Plus, since I’ve made the data public, I hope others will use my maps to map other things in Burkina Faso. For my organization, I plan to make an internal dashboard webpage where all of this data will be geared towards borrower activity. For example, we could view trouble areas for repayment rates, dynamic graphs of historical payments, and other summary statistics.
As an aside, the past few days of work also made me realize how great my job is. I had the time to explore and learn these tools without feeling time pressure, and have the leeway to create/research new features as I see fit. Thanks for the opportunity, Peace Corps.
Plan on using this data? I’d love to hear from you about your implementation.
*If you have a .kmz file, extract the .kml file from the .kmz file by changing .kmz extension to .zip and then extract.