Mapzen Places is here! And there! And everywhere. This is a blog post by mapzen that was published on Oct 15, 2017 and tagged places, flex, data and whosonfirst

map hat pieces

Mapzen Places is here! Find out more about a location, its geometry, its relationships, and return useful metadata, all through a convenient and well-documented API. Mapzen Places is all based on open data, with licensing suitable for commercial projects. Start using it via Mapzen Flex today!

If you like plain-text GeoJSON records with stable numeric IDs, a finite set of placetypes, and a common set of ancestors for all records, then Mapzen Places is for you. Perform spatial queries without having to set up and run a spatial database! Get worldwide admin hierarchies from continent to neighbourhood! Records may be updated, superseded or deprecated, but never removed. Mapzen Places also “holds hands” with Wikipedia, Wikidata and other datasets.

The Mapzen Places API gets its data from Who’s On First, and supersedes the Who’s On First API which we have been beta testing for the past year. Note that is deprecated but requests will get redirected to The old whosonfirst.* methods will continue to work but are deprecated — mapzen.* is the new method.

There are many useful methods in the Mapzen Places API, including


Get descendants of a place: counties in a country, postal codes in a city, regions in a continent, venues in a neighbourhood.

Here’s the API request for neighbourhoods in London:

_(We’ve broken out the arguments in the URL to make it easier to read – here’s a link, just add your API key.)_

Here’s the response:

places: [
        "wof:id": "85785187",
        "wof:parent_id": "101750367",
        "wof:name": "Abbey Wood",
        "wof:placetype": "neighbourhood",
        "wof:country": "GB",
        "wof:repo": "whosonfirst-data",
        "name": ""
        "wof:id": "85785221",
        "wof:parent_id": "101750367",
        "wof:name": "Acton",
        "wof:placetype": "neighbourhood",
        "wof:country": "GB",
        "wof:repo": "whosonfirst-data",
        "name": ""

Adding format=geojson does what you’d expect:

type: "FeatureCollection",
features: [
        type: "Feature",
        geometry: {
            type: "Point",
            coordinates: [
        properties: {
            wof:id: 85785187,
            wof:parent_id: "101750367",
            wof:name: "Abbey Wood",
            wof:placetype: "neighbourhood",
            wof:country: "GB",
            wof:repo: "whosonfirst-data",
            name: "",
            geom:latitude: 51.495463,
            geom:longitude: 0.12401,
            lbl:latitude: 51.488981,
            lbl:longitude: 0.11348

…as does format=csv:

,GB,85785187,"Abbey Wood",101750367,neighbourhood,whosonfirst-data


Get the hierarchy of a place, from neighbourhood or microhood up to country and continent. Here is the hierarchy for a microhood in Bernal Heights, San Francisco:
hierarchies: [
            microhood_id: "420780693",
            neighbourhood_id: 85865945,
            continent_id: 102191575,
            country_id: 85633793,
            locality_id: 85922583,
            county_id: 102087579,
            region_id: 85688637
stat: "ok"

Adding spr=1&extras=name& will give you a more detailed response – here’s a snippet of the neighbourhood details:

    neighbourhood: {
        "wof:id": 85865945,
        "wof:parent_id": "85922583",
        "wof:name": "Bernal Heights",
        "wof:placetype": "neighbourhood",
        "wof:country": "US",
        "wof:repo": "whosonfirst-data"


Discover placetypes that intersect a bounding box – here is the request for counties in -6.6482,38.2448,-4.8118,39.9779:

a.k.a. coastal Tanzania:

  "places": [
      "wof:id": 421188197,
      "wof:parent_id": 85679651,
      "wof:name": "Kusini",
      "wof:placetype": "county",
      "wof:country": "TZ",
      "wof:repo": "whosonfirst-data"
      "wof:id": 421191173,
      "wof:parent_id": 85679651,
      "wof:name": "Kati",
      "wof:placetype": "county",
      "wof:country": "TZ",
      "wof:repo": "whosonfirst-data"

These methods have lots of arguments, and as noted earlier, can return data in exciting formats such as json, geojson, and csv.

If you choose GeoJSON, you can add the API response directly onto a map! In Tangram, here’s how you can use mapzen.places.getDescendants to display airports in Africa:

        type: GeoJSON
        # url:
            method: mapzen.places.getDescendants
            id: 102191573 #africa
            api_key: global.sdk_mapzen_api_key
            placetype: campus
            tags: airport
            format: geojson
            per_page: 500
            extras: wof:tags,wof:concordances,mz:is_current,name:eng_x_variant,oa:type,oa:elevation_ft,wof:name

Note the extras — this lets you pull in metadata (in this case, airport codes and airport size that we’ve concorded with OurAirports and ICAO and IATA) that you can use for labels and display.

Open full screen, and check out the scene file in Tangram Play!

You can learn much more on the Mapzen Places page and the documentation. Mapzen Places is part of Mapzen Flex, with clear, predictable pricing and a generous free tier. Sign up for an API key to get started!

Preview image: Sheet printed with parts of map of the world in sections as would be made up into baseball hats. The parts of 4 hats are printed on each sheet. It was designed by Esselte Map Service and company: Interarts and made by Graphic Jackets, via Cooper Hewitt.