JavaScript
Article

Building Gorgeous 3D Maps with eegeo.js and Leaflet

By Tim Jenks

eegeo.js is an open source 3D maps API built on top of Leaflet, the popular mapping library.

There are many mapping libraries that allow you to embed or build simple, 2D maps. Notably Google Maps, Bing Maps, HERE, Mapbox and OpenStreetMap. Popular mapping abstraction libraries such as OpenLayers and Leaflet also allow you to change the map ‘base layer’ without having to change your application logic.

With the usage of the Google Earth API falling to NPAPI security deprecation, the choice of 3D map APIs is limited. Alternatives do exist, such as the popular Cesium library. eegeo.js is another, providing a dynamic, truly seamless 3D map of the world, from space to a desk in an indoor map. Right in the browser using WebGL.

eeGeo 3D Maps Space to Desk

Getting Started

In this article, I’ll show you how simple it is to embed a gorgeous 3D map of London and add some basic functionality using the Transport for London API.

Spoiler: 3D maps are as simple as using 2D maps.

Disclaimer: I am SVP Software Development at eeGeo, and I have oversight of the development of eegeo.js

There are plenty of tutorials on how to use Leaflet. eeGeo.js was consciously built on top of Leaflet so that developers already familiar with Leaflet can get up and running immediately. I thoroughly recommend the Leaflet examples and documentation.

This article will cover basic HTML, CSS, JavaScript and simple mapping concepts.

What You’ll Need

This article will cover a few key components:

Displaying a 2D Map of London

First, let’s create a 2D map of London, using Leafet and Open Street Map. Leaflet is super simple. All you need is a tiny bit of HTML to setup the page, a <div> to contain your map, and a few lines of JavaScript to create and set the initial position.

Create the HTML

Let’s create a HTML file with the following markup:

<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="utf-8">
    <title>Building Gorgeous 3D maps with eegeo.js and Leaflet</title>
  </head>
  <body>
    <div id="map" style="width: 600px; height: 400px;"></div>
  </body>
</html>

Include Leaflet

To include Leaflet it’s as simple as adding the Leaflet JavaScript library and CSS. Include these within the <head> tag:

<link rel="stylesheet" href="https://unpkg.com/leaflet@1.0.3/dist/leaflet.css" />
<script src="https://unpkg.com/leaflet@1.0.3/dist/leaflet.js"></script> 

Add the Leaflet JavaScript

The following lines of JavaScript initialize Leaflet, and use OpenStreetMap as a tile layer. You can add this as an external JavaScript file, or for simplicity in a <script></script> block below <div id="map"/>:

var map = L.map('map', {
  center: [51.517327, -0.120005],
  zoom: 15
});
L.tileLayer('http://{s}.tile.osm.org/{z}/{x}/{y}.png', {
  attribution: '&copy; <a href="http://osm.org/copyright">OpenStreetMap</a> contributors'
}).addTo(map);

The code initializes the map centered on Holborn Tube Station at the WGS84, decimal degrees latitude, longitude of 51.517327, -0.120005. You can obtain these coordinates by right clicking and selecting ‘What’s here’ on digital maps like maps.eegeo.com or Google Maps. You’ll see later that most geospatial APIs return positions in standard WGS84 decimal degrees latitude and longitude, including the TfL API.

Fire up the 2D Map

Upload your HTML somewhere, or view it locally. I recommend not loading directly from the file system but using a local loopback web server. Save your HTML as index.html, and in the same folder, run: python -m SimpleHTTPServer 8000 You can then visit http://localhost:8000/ in your browser.

You’ll see the following map, centered on Holborn Tube Station in London:

2D map of London centered on Holborn Tube Station

Displaying a 3D Map of London

As eeGeo.js is built on top of Leaflet there are very few changes to make to turn the 2D map of London into a gorgeous 3D map.

Change the JavaScript we added above which read var map = L.map ... to the following. A free eeGeo API key is required. You should also delete the JavaScript that calls L.tileLayer as it is no longer required.

The JavaScript should now read:

var map = L.eeGeo.map('map', '<your_api_key>', {
  center: [51.517327, -0.120005],
  zoom: 15
});

Change the <script> include we added to <head> to include eegeo.js rather than Leaflet. The eegeo.js file already contains Leaflet, so don’t include both:

<script src="https://cdn-webgl.eegeo.com/eegeojs/api/v0.1.780/eegeo.js"></script>

And that’s it! By changing just a couple of lines, you’ll now see the exact same view of Holborn Tube Station but in gorgeous 3D:

eeGeo.js 3D Map of London centered around Holborn Tube Station

Changing zoom, tilt & orientation

As eeGeo.js is completely 3D, it’s possible to rotate and tilt the map. The Leaflet setView function is supported, with optional tilt and heading arguments to support the 3D behavior of the map.

Add the following JavaScript and after 10 seconds the map will seamlessly animate in 3D from Holborn Tube Station to the Gherkin in London:

setTimeout(function() {
  map.setView([51.514613, -0.081019], 17, {headingDegrees: 204.374, tiltDegrees:15.0});
}, 10000);

eeGeo.js 3D Map of the Gherkin, London

The screenshot doesn’t do the interaction justice, but this CodePen shows the animation in action:

See the Pen Gorgeous 3D Map Transitions with eeGeo.js and Leaflet by SitePoint (@SitePoint) on CodePen.

The additional arguments to setView() above are as follows:

Field Type Description
headingDegrees decimal decimal degrees from north, clockwise, where 0.0 is north.
tiltDegrees decimal tilt in degrees where 45.0 is directly overhead and 0.0 is the default tilted view

Interacting with the 3D Map of London

Now, let’s add something useful to the map. Most Leaflet primitives are supported such as Markers, Polylines, and Polygons.

To add a Marker at Holborn Tube Station to the map, call L.marker():

L.marker([51.517327, -0.120005]).addTo(map);

To add a Polygon at Russell Square, call L.polygon(). The Polygon is defined by four vertices, clockwise, starting at the Bernard St & Wolborn Place junction at the north easterly corner of Russell Square:

L.polygon([[51.522771, -0.125772],
           [51.521520, -0.124192],
           [51.520631, -0.126358],
           [51.521963, -0.127895]]).addTo(map);

To add a Polyline of a route through Bloomsbury, starting at Great James St. and ending at Museum St.:

L.polyline([[51.521788, -0.117101],
            [51.520732, -0.116524],
            [51.520430, -0.117641],
            [51.519873, -0.119602],
            [51.519034, -0.120993],
            [51.518784, -0.121537],
            [51.517265, -0.125088],
            [51.516257, -0.124466]]).addTo(map);

Refresh, and the three Leaflet primitives will now be shown on the map:

eeGeo.js 3D map showing the marker, polyline, and polygon

You’ll notice the calls to L.marker(), L.polyline() and L.polygon() are all Leaflet API calls. If we revert the map back to the OpenStreetMap from earlier, the exact same calls will still work on the 2D OSM base layer:

2D OpenStreetMap with the marker, polyline, and polygon still showing

Transport for London (TfL) API

Next, let’s add some data from a real-world API to the 3D map of London.

The Transport for London (TfL) API is large and complex. It covers all transport in London, including bus, tube, overground, cycle hire and provides real-time status and arrival time data.

The scope of the full TfL API is far too large to cover in this article. I will focus solely on a very small number of API endpoints that describe the London Underground, specifically answering:

  • What are the available tube lines?
  • Which stations does a particular tube line stop at?
  • What is the “arrival board” of a particular tube station?

The TfL API can be used without an API key with strict rate limiting. If you use the TfL API in a real, production deployment, you should sign up for an API key.

What are the available tube lines?

To answer the question of ‘What are the available tube lines’, the TfL API provides an endpoint on the Line resource

We are only interested in the ‘tube’ mode of transport, so making the following call:

curl -v https://api.tfl.gov.uk/line/mode/tube

returns a JSON response with an array of Line entities. We are only interested in the id property of each of the entries:

[{
  "id": "bakerloo",
  "name": "Bakerloo",
  "modeName": "tube",
  "disruptions": [],
  "lineStatuses": [],
  "routeSections": [],
  ...
},...]

To simplify the interaction with the TfL API, we’ll use jQuery, but the dependency on jQuery is entirely optional. The following JavaScript calls the Line endpoint and collects the Line IDs:

$.getJSON( "https://api.tfl.gov.uk/line/mode/tube", function(data) {
  var lines = [];
  $.each(data, function(k, line) {
    lines.push(line.id);
  });
});

I will later use these Line IDs to query other TfL endpoints that describe the stations along a particular route of the Underground.

Modeling the London Underground as a Graph

You can think of the London Underground as a Graph. A Graph contains Vertices and Edges. For the London Underground, I will consider the stations to be the Vertices, and the tube lines will form the Edges.

A section of the London Underground diagram

Modeling the central line (in red above) as a Graph, there would be Vertices at North Acton, West Acton, and Hanger Lane. As there is a branch in the central line westbound from North Acton, a central line Edge exists between North Acton and West Acton and also between North Acton and Hanger Lane.

Further to the east on the central line is Notting Hill Gate. There is only one Notting Hill Gate underground station, yet it has platforms for three lines: central (red), circle (yellow) and district (green). We only want to add one Leaflet Marker for Notting Hill Gate, not three. If we model Notting Hill Gate as a Vertex, there would be central line Edges (to Holland Park and Queens Way), circle line and district line Edges to Bayswater and High Street Kensington.

This is significant, as the TfL API models the Underground as a directed graph.

What are the stations on a particular Tube line?

We now have the Line ID of every line of the Underground. I’m also planning on loosely modeling the stations and lines as vertices and edges in a Graph. Given a Line ID, the TfL API can then tell us what stations each line stops at. The TfL API refers to these as StopPoints. The endpoint we will call is the Line, Route including sequence of StopPoints

Let’s make another call to the TfL API to describe the outbound route of the central line:

curl -v https://api.tfl.gov.uk/line/central/route/sequence/outbound

This returns a very large JSON payload, over 8000 pretty-printed lines. Luckily, we are only interested in the stopPointSequences key:

"stopPointSequences": [
  {
    "lineId": "central",
    ...
    "stopPoint": [
      ...
      {
        "id": "940GZZLUNHG",
        "name": "Notting Hill Gate Underground Station",
        ...
        "lat": 51.509128,
        "lon": -0.196104
      },
      {
        "id": "940GZZLUQWY",
        "name": "Queensway Underground Station",
        ...
        "lat": 51.510312,
        "lon": -0.187152
      },
      ...
    ],
  },
  ...
]

This property is a list of JSON objects that group the line into contiguous sections. Within a StopPointSequence are all the StopPoints, in order, along the line.

The following JavaScript handles the creation of unique stations and also the creation of polylines so we can draw a line for the edges of the tube lines.

var stopPoints = {};
$.getJSON( "https://api.tfl.gov.uk/line/central/route/sequence/outbound", function( data ) { 
  var lineSegments = [];
  $.each(data.stopPointSequences, function(k, v) { 
    var line = [];
    $.each(v.stopPoint, function(k, sp) {
      line.push([sp.lat, sp.lon]);
      if (!(sp.id in stopPoints)) {
        stopPoints[sp.id] = {"id":sp.id, "name":sp.name, "lat":sp.lat, "lon":sp.lon};
      }
    });
    lineSegments.push(line);
  });
});

We can now call the above JavaScript for each line of the Underground. Afterward, we will have every unique station collated in the stopPoint dictionary. For each line, we also have a collection of polylines representing the sections of the tube line in the lineSegments array.

Later I will show how to add Leaflet Markers and Polylines to visually display the tube stations and lines.

Which trains are arriving at a particular station?

The final TfL API endpoint we need is live arrivals. In the above section, we can see that a station has a unique identifier in the API. For Notting Hill Gate, the ID is 940GZZLUNHG:

{
  "id": "940GZZLUNHG",
  "name": "Notting Hill Gate Underground Station",
  ...
  "lat": 51.509128,
  "lon": -0.196104
}

We can now call the TfL StopPoint API, providing a StopPoint ID:

curl -v https://api.tfl.gov.uk/stoppoint/940GZZLUNHG/arrivals?mode=tube

This returns an array of arrivals:

[{
    ...
  "naptanId": "940GZZLUNHG",
  "stationName": "Notting Hill Gate Underground Station",
  "lineId": "central",
  "platformName": "Eastbound - Platform 3",
  "direction": "outbound",
  ...
  "destinationName": "Epping Underground Station",
  "timeToStation": 290,
  "currentLocation": "Between White City and Shepherd's Bush",
  "towards": "Epping",
  ...
},
...]

This is indicating that an Eastbound Train to Epping will arrive at Platform 3 in just under 5 minutes time (290 seconds).

Putting It Together

So far, I have shown you how to:

  • Create a simple 2D OpenStreetMap map with Leaflet
  • Turn that simple 2D map into a gorgeous 3D map with eeGeo.js and Leaflet
  • Add simple details to maps in the form of Leaflet Markers, Polylines, and Polygons
  • Perform basic interactions with the Transport for London API

Now to bring these simple concepts together to produce a 3D interactive map of London, showing the Lines and Stations of the London Underground and real-time arrival times for all Stations!

3D map of London with Underground stations and arrival times displayed as an overlay

Add Markers for Underground stations

In the previous $.getJSON() call that requests https://api.tfl.gov.uk/line/central/route/sequence/outbound, call L.marker() for each Station:

$.each(stopPoints, function(k, sp) {
  L.marker([sp.lat, sp.lon], 
    {
      title: sp.name,
      options: {"id":sp.id, "name":sp.name}
    }).addTo(map);
});

We have passed additional ‘options’ through the marker. This is so later when we handle mouse hover events, we know what Station ID and Name we have hovered over.

Add Polylines for Underground lines

Again, in the previous Ajax call that requests https://api.tfl.gov.uk/line/central/route/sequence/outbound, add an L.polyline() call for each of the line segments:

$.each(lineSegments, function(k, ls)
{
  L.polyline(ls, {weight:8, color:"#CC3333"}).addTo(map);
});

The color above could come from a lookup by Line ID. In this case, I have hard coded circle line red in accordance with the TfL Design Guidelines

Handling mouseover events on Leaflet Markers

I’d like to display the real-time arrival information from the TfL API. It’s possible to receive a callback when a Leaflet Marker is moused over. It’s also possible to show a Leaflet Popup to show the name of the station.

$.each(stopPoints, function(k, sp) {
  var marker = L.marker([sp.lat, sp.lon], 
  {
    title: sp.name,
    options: {"id":sp.id, "name":sp.name}
  }).addTo(map);
  marker.bindPopup(sp.name);
  marker.on('mouseover',_hovercard.overstation);         
  marker.on('mouseout',_hovercard.outstation);  
});

In a more complex CodePen below, I will show how the _hovercard.* callbacks will make a request to the TfL API and populate a callout card with the real-time arrivals at the station.

Putting everything together

With some additional Separation of Concerns and some better modeling, you can see a full demo of eegeo.js, Leaflet, and the TfL API in this Code Pen:

See the Pen eeGeo.js and TfL API by SitePoint (@SitePoint) on CodePen.

3D Maps Don’t Have to Be Hard

In this article, I’ve shown how to add 2D maps to your site with JavaScript and Leaflet, and how to add simple map features like Markers, Polylines, and Polygons. I’ve then shown how data can be easily added from real-world APIs like the Transport for London (TfL) API

I’ve highlighted the benefits of coding against mapping abstraction libraries such as OpenLayers and Leaflet. The visuals of your map can change drastically with little to no change to your code. I’ve also shown how simple 3D mapping can be using eegeo.js.

If you’ve any comments or questions about something in this article, please let me know in the comments below.

More:
Recommended
Sponsors
Get the latest in JavaScript, once a week, for free.