This is my attempt at remixing the <a href="http://wiki.openstreetmap.org/wiki/Tah-heatmap" rel="nofollow">Tiles@Home Heatmap</a> by Ævar Arnfjörð Bjarmason (<a href="https://www.flickr.com/photos/avarab/3613043325/">on flickr</a>)

This is a matplotlib/python implementation (the original is in Perl)

The heatmap uses the file size of each tile, which should give a good measure of how much real data there is in a tile. Red areas have lots of data, green areas less, and blue the least.

Data sources: <a href="http://www.openstreetmap.org/" rel="nofollow">www.openstreetmap.org/</a>, <a href="http://tah.openstreetmap.org" rel="nofollow">Tiles@Home</a>

Source code on OSM repository : <a href="http://svn.openstreetmap.org/applications/rendering/tah-heatmap/" rel="nofollow">original</a>, <a href="http://svn.openstreetmap.org/applications/rendering/genericHeatmapRender/" rel="nofollow">mine</a>.

I've used log10 scaling for the colours, so the rural areas get a bit more prominence.