OSM QA Tiles are now being updated daily for the entire world on osmlab.github.io as mbtiles. OSM QA Tiles contain the full spectrum of OpenStreetMap data at full fidelity. Use them with Tile Reduce to conflate OpenStreetMap with other datasets, detect bugs like connectivity issues, or monitor for vandalism - you can do nearly any tile-based analysis of OpenStreetMap data.
OSM QA Tiles contain the full spectrum of OpenStreetMap data for analysis.
To analyze the data in OSM QA Tiles, you can write simple Turf.js based processors in Tile Reduce. Tile Reduce is a geospatial MapReduce library for running custom processors across all tiles in a given area. Tile Reduce handles scaling these processors across all available cores on a machine, and it exposes data as GeoJSON, making processors easy to write and understand.
Here are two examples of how to build simple data analysis with OSM QA Tiles and Tile Reduce.
Count buildings in an area
index.js
varTileReduce=require('tile-reduce');varturf=require('turf');varargv=require('minimist')(process.argv.slice(2));// Pass area from command linevararea=JSON.parse(argv.area);// Options: zoom level, tile source(s), processor to run)varopts={zoom:15,tileLayers:[{name:'osmdata',mbtiles:__dirname+'/latest.planet.mbtiles',layers:['osm']}],map:__dirname+'/buildings.js'};// Initialize Tile Reducevartilereduce=TileReduce(area,opts);varbuildings=0;// Accumulate resultstilereduce.on('reduce',function(result){buildings+=result;});// Output resultstilereduce.on('end',function(error){console.log(buildings);});tilereduce.run();buildings.js
varturf=require('turf');varnormalize=require('geojson-normalize');module.exports=function(tileLayers,opts,done){// Create GeoJSON FeatureCollection from source datavarosmdata=normalize(tileLayers.osmdata.migeojson);varbuildingCount=0;// Count features containing the "building" keyosmdata.features.forEach(function(feature){if(feature.properties.building){buildingCount++;}});// Return resultsdone(null,buildingCount);};Run for given bounding box
node index.js --area [min x, min y, max x, max y]
Create a difference map to TIGER roads
index.js
varTileReduce=require('tile-reduce');varturf=require('turf');// Declare area in [x,y,z] formatvararea=[327,791,11];// Setup options for two mbtiles sourcesvaropts={zoom:15,tileLayers:[{name:'osmdata',url:__dirname+'/latest.planet.mbtiles',layers:['osm']},{name:'tiger',url:__dirname+'/tiger.mbtiles',layers:['tiger']}],map:__dirname+'/difference.js'};// Initialize TileReducevartilereduce=TileReduce(area,opts);...difference.js
varturf=require('turf'),flatten=require('geojson-flatten'),normalize=require('geojson-normalize'),tilebelt=require('tilebelt');module.exports=function(tileLayers,tile,done){// Concatenate feature classes and normalize datavartigerRoads=normalize(tileLayers.tiger.tiger20062014);varosmData=normalize(tileLayers.osmdata.migeojson);// Buffer OpenStreetMap features with 'highway' tagsvarstreetBuffers=turf.featurecollection([]);streetBuffers.features=osmData.features.map(function(f){if(f.properties.highway){returnturf.buffer(road,20,'meters').features[0];}});streetBuffers=normalize(turf.merge(streetBuffers));// Erase OpenStreetMap highway buffers from tiger lines to reveal differencevartigerDeltas=turf.featurecollection([]);if(tigerRoads&&streetBuffers){tigerRoads.features.forEach(function(tigerRoad){streetBuffers.features.forEach(function(streetsRoad){varroadDiff=turf.erase(tigerRoad,streetsRoad);if(roadDiff&&!filter(roadDiff))tigerDeltas.features.push(roadDiff);});});}// Return resultsdone(null,tigerDeltas);};...View the complete code for difference.js here
Run and render map with Tippecanoe
node index.js | tippecanoe -o difference.mbtiles
We’re looking forward to seeing more data analysis from the community using OSM QA Tiles and Tile Reduce. Email us at help@mapbox.com or hit me up on Twitter @mateo_verde if you have any questions.