Guest Post: CensusMapper is all about open dataMay 30, 2016
CensusMapper is all about open data.
At it’s core is Stats Canada census open data. So far CensusMapper has imported the complete Canada 2011 and 2006 census datasets. This data is freely accessible from Stats Canada, but there are many barriers to putting it into widespread use.
First and foremost, the dataset is rather large. Just for the 2011 census dataset there are 616,762 standard geographic regions. The smallest of them, Dissemination Blocks, only have information on population, households and dwellings. Each of the other 123,570 geographic regions additionally have information on 1,305 census variables and 2,655 NHS variables. For example, we could look at the variable for the “median after-tax income of couple families with children”. The value for this variable will vary for each of our 123,570 geographic regions, for the City of Vancouver the value is $82,444.
Pairing each variable with each of the geographic regions adds up to almost half a billion individual values like this. While this is still peanuts for modern database systems and nowhere near the size that would qualify as “big data”, this will provide significant challenges to the average user.
We were frustrated by the fact that all this valuable data was so inaccessible for most people. That’s where CensusMapper comes in. CensusMapper is a platform that allows the dynamic mapping of all census variables for all geographic regions. More generally, CensusMapper allows the map maker to specify arbitrary functions built from census variables. And the back end tied directly into R, a powerful statistics package to aid any analysis we want to run on the data.
Organizing all that data in CensusMapper certainly made census data more accessible to us, but what about everyone else?
Apart from creating maps for everyone to browse, we have created two ways for everyone else to access census data through CensusMapper. The "Census Wheel" to allow browsing all census variables for a fixed region, and "Simple Maps" for creating maps of single census variables for all geographic regions.
To bring up the Census Wheel, navigate to the region of interest, zooming in or out as appropriate and click into the region. Then use the "sunburst" interface to navigate and access all census variables for this region and display them in a meaningful way.
Simple Maps allows everyone to make their own CensusMapper maps based on whichever census variable they happen to be interested in. And anyone can create an account to save and share the maps they create.
For example, a Canada-wide Transit Mode Share map is now 5 clicks away for anyone. One click to “Start New 2011 Map”, two more to open up the “Commute” and “Mode Share” variable selection, one to select “Public Transit” and then the final click to map “Public Transit as a percentage of Mode Share” (instead of the alternative of mapping the total number of public transit users). That’s it. The result is an interactive Canada-wide map of transit mode share. We also offer ways to fine-tune the colour scheme and the cutoffs for the colour bins. And anyone can create an account so that they can save their maps, write a map story to go along with their map and publish it to share it with friends or whoever they want.
Without being easily accessible, open data can't live up to its potential. Census data underpins all data driven decision making.
At CensusMapper we are proud to be able to provide the Census Wheel and Simple Maps free of charge for everyone, and bring this important dataset right onto every computer, tablet and phone. We are hoping to be able to expand the functionality of the public interface of CensusMapper in the future to help put census data to work for the general public.
Guest post by Jens von Bergmann
Jens von Bergmann holds undergraduate degrees in Physics and Computer Sciences and a PhD in Mathematics. He taught for several years at the University of Calgary, University of Notre Dame and Michigan State University before founding MountainMath to work on his passion of data analysis and visualization.