Following on from The Political Colour of Great Britain, and reusing the same code, I have produced a map for London, showing graphically the results of the voting for the local elections in Greater London’s 600-odd wards, for the May 2010 elections and back in 2006.
The “colour” map assigns every vote to one of the three RGB primary colours – red for Labour, blue for Conservative and green for all other parties. It so happens that the three groupings have roughly the same number of votes across the whole of London. These are scaled by the total number of votes for each ward, and then resulting proportions are converted to the hexadecimal “web” colours you see on the dots. An “enhance” function is used to increase the value of the colours away from the mean, to prevent the map from looking muddy.
The advantage of using colour in this way to represent each constituency is every person’s vote counts towards the final colour, rather than just those that elected the three winning councillors in each ward. Use of a single colour is the simplest way to summarise each result. The disadvantage is that it is difficult for human eyes to quantitatively perceive the colour and translate it to a result – although we are quite good at spotting differences in colours, it is more difficult to interpret these.
The voting data is from the London Datastore, the ward and borough boundaries from Ordnance Survey Open Data, and the background map from OpenStreetMap data, rendered using Mapnik. The ward centroids were calculated in ArcGIS and the map is displayed with the OpenLayers framework. Performance is very poor in Internet Explorer because the VML renderer it uses is extremely slow – SVG is used instead in Firefox and the other standards-based browsers and is vastly better.
For a different take on the same technique of using colour to show the vote composition for each ward, see the article on Spatial Analysis.