Following on from the Charles Booth article, I thought I’d put together another short piece again using feely available open data from Ofcom and the OS combined with Opensource tools.
Late last year Ofcom published the detailed dataset that sits behind their annual broadband speed survey. There are any number of ways of studying the wealth of information locked up inside this but I was trying to get a general feel for the distribution of speeds – hot spots, notspots and their relationship with each other.
One approach I settled on was to generate contour lines for areas likely to have similar download speeds. For this, the base data is loaded into Qgis and the holes in the coverage filled using a suitable interpolation algorithm; Ofcom doesn’t provide data for every postcode. From the result contour lines can be generated using a standard plugin.
The resulting interactive map overlays the contours on OpenStreetmap – use the +/- to zoom in and out, or drag the map around to see other areas.
This is not a pinpoint accurate engineering answer to the question but it fits the bill – to give a sense of how broadband speeds vary over a district, and how slow and faster areas relate to each other.
And for anyone thinking about how much all this costs, the whole exercise was done using a set of freely available opensource applications to process open data. None of these applications required a deep down technical understanding to create any part of this – every part of this was done using user-friendly but immensely powerful tools.