Monday 11 June 2012

Room for everyone

Last week I crossed off a work-related blog that I'd been meaning to write for a while. It demonstrated how to recreate data binning for point datasets using hexagons in ArcGIS. It's not a new technique but the fact that I told people how to do it in a particular environment has caused some discomfort given the tweets I've had fired at me today.

It doesn't take very long to do a quick search online to find examples of the use of hexagons in cartography.  Sherif Amer used them for investigating service planning and Danny Dorling used them as a form of cartogram in a range of studies including his analysis of poverty, wealth and place in Britain 1968-2005. Dorling's own work on area cartograms is well known and gives the basis for hexagonal tesselations of space though Waldo Tobler, of course, proposed hexagons as a transformation useful for districting in 1973. Nicholas Lewin-Koh provides a good overview of the technique whose purpose is simply to tesselate space into an aesthetically pleasing grid of polygons of equal size in which to represent some other dataset as a spatial histogram.  It's become a reasonably popular technique. Zach Johnston did a good job illustrating the technique at and provided code to allow others to replicate should they wish. There is also a really nice blog by Nate Smith at on their efforts with this technique that describes the technique for QGIS and Tilemill.

Now here's the thing...I don't subscribe to the view that a concept or a technique is the unique preserve of one person or one approach.  I like what others have done.  I also like that you can do it in ArcGIS but until now no-one had explained the relatively simple process.  My job is to provide people with help to enable them to make good maps. How they go about it and what they prefer to use is their choice but by giving people options it supports better cartography more generally.  I have no association with the technique (other than knowing Professor Dorling and having met Professor Tobler); no more than any other established cartographic technique I employ or I use. The blog I wrote doesn't imply ownership. It's not a paleo/neo fact, those who have read my Editorials in The Cartographic Journal over the last 6 years will know where I stand on that distinction.

There's room for us all...whether we largely use proprietary software or open source or whether we come from a computer science background or a cartographic background.  We all have space to learn from each other and being united in showing people how to make better maps is really what makes me tick.


  1. I enjoyed your post on the ESRI blog and learned a lot from it, thank you. You're absolutely right that we have space to learn from each other and we're united by mapmaking. That's why it's a shame you didn't include all these references you have here in your ESRI post; that would have made it easier for ESRI's readers to learn both the history and how other folks like Zach and Nate think about these techniques. References always make a paper stronger.

  2. Nelson, thanks for the comments. Given I come from an academic background of some 20+ years teaching University students I'm very aware of the value and critical importance of good quality referencing. That said, I hope you appreciate that it's not always possible for a variety of reasons.