Dynamic + vector maps using OSM with CartoDB

Fun with Vecnik, CartoDB, and other stuff

www.stateofthemap.us
www.cartodb.com

Transcript

1. USING OSM DATA TO BUILD DYNAMIC MAPS @ANDREWXHILL
4. http://vizzuality.com/projects/protectedplanet Text
7. Make maps online easier than ever before
8. http://cdb.io/176ij9j
9. http://aitor.is/iceland-contour/
14. http://mwcimpact.com/
15. A lot of people still associate CartoDB with maps not wrong, but not complete
16. SQL API
17. Data out
18. Data in
19. CartoDB becomes really powerful when your data grows let’s take a look
20. Barcelona
21. Barcelona
23. Barcelona
24. What about OSM and CartoDB let’s take a look
25. Import support easy peasy
26. DRAG & DROP .OSM IMPOSM
28. Tile based rendering one to many layers
31. 2.1 is around the corner! this stuff is going to be easier than ever
32. OSM Data in CartoDB the power of dynamic data
35. SQL + CSS
38. Vector rendering SQL API FTW
42. But what about dynamic data? where things start to get fun
43. Playing with OpenPaths looking at my recent data from nyc
46. Question 1 Which streets are most important?
47. Pick a simple algorithm For each OpenPaths point Find the 3 closest streets by distance Assign 0.5, 0.3, 0.2 in that order
49. WITH points AS ( SELECT (SELECT cartodb_id FROM axh_roads ORDER BY a.the_geom <-> the_geom LIMIT 1) as first, (SELECT cartodb_id FROM axh_roads ORDER BY a.the_geom <-> the_geom OFFSET 1 LIMIT 1) as second, (SELECT cartodb_id FROM axh_roads ORDER BY a.the_geom <-> the_geom OFFSET 2 LIMIT 1) as third FROM axh_openpaths a) UPDATE axh_roads b SET openpaths_score = openpaths_score + 0.5 * (SELECT count(*) FROM points WHERE first = b.cartodb_id) + 0.3 * (SELECT count(*) FROM points WHERE second = b.cartodb_id) + 0.2 * (SELECT count(*) FROM points WHERE third = b.cartodb_id) The Algorithm in SQL
50. SELECT * FROM axh_roads WHERE openpaths_score > 0 Filter the map
54. Back to vectors with changing data
56. Needed some real-time data that i was allowed to use...
60. http://bit.ly/13n0mAx A day in NYC sun exposure
63. There must be a Manhattanhenge every day or brooklynhenge or queenshenge etc
64. http://bit.ly/18gHavf A year of Manhattanhenges
66. Experiments with OSM and live data Got other datasets? Let’s talk
68. THANKS! @ANDREWXHILL