I'm interested in calculating for every point the distance to the nearest amenity of each of a certain number of types (train station, supermarket and so on). I think the calculation will be straightforward, but I'm not sure how to visualise the result. A heatmap doesn't seem quite right.

Things I want the visualisation to do:

  • No distinction needed between a point that is near 1 supermarket vs near 2 supermarkets
  • Show each attribute individually (for maybe 3 attributes)
  • Use a step function for distances (eg, 0-499m is one colour, 500-999m is another...)

Is there any standard kind of visualisation for this kind of analysis? My stack is Leaflet, TileMill, PostGIS, OpenStreetMap, comfortable with Python tools but would consider branching out.

  • Calculation itself might be tricky too. You should use correlated subqueries and distance operators or else it might take eternity. Jul 3 '13 at 12:29
  • It's actually not that bad. A similar query takes about a minute to find each of a dozen types of amenities within a 5km radius of every town in Australia. Queries like: select count(*) from planet_osm_polygon p where p.shop = 'supermarket' and st_dwithin(p.way,places.way,scope) Jul 3 '13 at 12:35
  • I should perhaps mention that for this question, I'm only thinking of a fairly small area (a few suburbs) - but yes, it would become a big problem on a bigger scale. Jul 3 '13 at 12:36

The term you are looking for is accessibility surface analysis/visualisation - Google will point you to many examples in the literature. One good example I found is http://www.gsd.harvard.edu/gis/manual/raster/

If you are implementing the calculation your self I'd google map algebra python and look at some of the links returned.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.