You are currently browsing the tag archive for the ‘web’ tag.

This tutorial will show how to use version 3.1 of the Calais API to identify geographic references in a text and display them on an OpenLayers map. I am going to use the Calais API with JSON output (which is a new feature of version 3.1), so that all the processing can be done easily in JavaScript entirely inside a web browser.

A running version of the demo application can be found here. The code can be downloaded here.

What is the Calais API?

From the documentation:

The Calais Web Service allows you to automatically annotate your content with rich semantic metadata, including entities such as people and companies and events and facts such as acquisitions and management changes.

Version 3.1, currently in beta, has added information regarding the latitude/longitude of a few types of entities: Cities, provinces/states and countries. It makes it possible to plot those entities on a map, in a way similar to what can be done with the Metacarta GeoTagger API.


Read the rest of this entry »

Google has recently released a Flash version of the Google Maps API. The goal is to give developers and designers more freedom in the presentation and graphical effects of their maps. Google is certainly not the first provider of a Flash mapping API (see the Yahoo! Maps AS3 API for example) but as one of the original proponents of Ajax web mapping tools, it marks an interesting change of direction (although the original Google Maps API is probably not going away anytime soon). One interesting aspect of using Flash is also a boost in vector graphics rendering speed, which I am going to demonstrate here, with a tutorial showing how to display a chloropleth map with this new API. Here is a link to the final application. It has to be noted that I am definitely not a Flex/ActionScript guru (just a few evenings of experience) so there are probably better ways of doing things than what I am showing here.

Getting the data

The data used is the simplified world borders shapefile found on Thematic Mapping (ie the file, at the bottom of the page). I have tried with the non simplified world borders as well, it is usable and quite fast to display but still feels somewhat sluggish, so the simplified dataset it is.

The application

The first step is to create the the MXML, with the component holding the map on it :

<?xml version="1.0" encoding="utf-8"?>
<mx:Application xmlns:mx="" layout="absolute"
	<gvlt:ThematicMapContainer width="100%" height="100%" />

Probably not earth-shattering… The ThematicMapContainer class in the com.vellut.examples.gmapsflash package is a subclass of UIComponent. Here is the beginning of this class, that is going to be enriched as the tutorial progresses :

    public class ThematicMapContainer extends UIComponent
        public var map: Map;

        public function ThematicMapContainer()

        private function startMap(event:Event):void
            map = new Map();

        private function resizeMap(event:Event):void
             map.setSize(new Point(width, height));

        private function onMapReady(event:Event):void

            map.addControl(new PositionControl());
            map.addControl(new ZoomControl());
            map.addControl(new MapTypeControl());


In the constructor, we subscribe to the “initialize” event of the container to start creating the map. Then when the initialiization is complete, the map is created and added to the container. Note that the API key is passed through the HTML page embedding the application (using flashVars), so this is why it does not appear anywhere here. The interesting thing, as far as the Google Maps Flash API is concerned, happens in the “onMapReady” method, called when the map has finished loading. Controls and various behaviours are added. The “createAndAddGeometries” method is also called: It is going to create the polygons for all the countries in the world, compute a color for each of them (according to an attribute of the country) and place them on the map. Note that contrary to what could be done with the Google Maps AJAX API, there is currently no way to place geometries through a KML or GeoRSS link. This means the polygons must be manually created though more atomic API calls.

Creating the polygons

There are 2 ways to create a polygon : point by point (with LatLng objects) or using an array of EncodedPolylineData objects, which is more compact. The second method also lets us have levels of detail for the polygons, which means some points will not be displayed when the map is zoomed out, which is quite useful when we have a heavy layer, so for this tutorial I chose the second method. One problem with the current version of the API is that the second method does not seem to work (the issue has been reported on the gmaps-api-issues tracker: there is a complaint of a missing attribute on the EncodedPolylineData object ; encoded polylines work fine). Still there is a very easy workaround. Create this class :

public class EncodedPolylineData extends
    public var options:*;
    public function EncodedPolylineData(arg0:String, arg1:Number, arg2:String, arg3:Number)
        super(arg0, arg1, arg2, arg3);

It can be used subsequently everytime the original EncodedPolylineData class is expected. Now that this is over with, on to more interesting things.

In the following body of the “createAndAddGeometries” method, the polygons are created, added to the map and the call to the method attributing a color class to each geometry is made :

private function createAndAddGeometries() : void
    var polygons : Array = [];
    polygons.push(new Feature(Polygon.fromEncoded([new    EncodedPolylineData('wa|fBrdowJexNndf@{iJ{bQ`cZs`T',32,'BAAB',4),new EncodedPolylineData('wdnjBhnwwJt~CfeWerV`|BnrQib[',32,'BAAB',4)]),{pop2005: 83039, area: 44 }));
    //the rest of the polygons are omitted for brievity

    for(var i : int ; i < polygons.length ; i++)
    chloropleth(polygons,"pop2005",5,new Color(Color.YELLOW),new Color(Color.RED));

The complete set of polygons is generated offline through a Python script, which uses the Shapely and Gpolyenc libraries. The former takes care of reading the world borders Shapefile (through its OGR binding), while the latter encodes each polyline into the Google polyline encoding format and performs some cartographic generalization through the Douglas-Peucker algorithm (which will let us drop some points while zoomed out). The output of the script is a text file containing valid ActionScript code that can be copy-pasted into the “createAndAddGeometries” method body and compiled with the rest of the code. The script is packaged along with the Flex/ActionScript code for this tutorial (the 2 libraries mentioned above and the dataset will have to be downloaded separately though).

Each country feature gets a geometry, along with 2 attributes: Area and Population in 2005. The thematic analysis will be performed on the population (“pop2005″, in the call to the chloropleth method). We want 5 classes (each class will have the same number of countries). Finally, the countries with the lowest population will be in yellow, while the ones with the largest will be in red.

Computing the classes

The “chloropleth” method is as follows:

        public function chloropleth(objects:Array, field:String, numClasses:int, colorBegin:Color, colorEnd: Color): void
            if(numClasses < 2)

            var i : int ;

            var sobjects:Array = objects.sort(function(a:Feature,b:Feature):int {
                if(a.attributes[field] < b.attributes[field])
                    return -1;
                else if (a.attributes[field] > b.attributes[field])
                    return 1;
                    return 0;

            var colors:Array = [];
            var incr: Number = (colorEnd.r - colorBegin.r) / (numClasses - 1);
            var incg: Number = (colorEnd.g - colorBegin.g) / (numClasses - 1);
            var incb: Number = (colorEnd.b - colorBegin.b) / (numClasses - 1);
            colors[0] = colorBegin;
            colors[numClasses - 1] = colorEnd;
            for(i = 1 ; i < numClasses - 1; i++)
                colors[i] = new Color(0);
                colors[i].setRGB(Math.round(colors[i-1].r + incr),
                    Math.round(colors[i-1].g + incg),
                    Math.round(colors[i-1].b + incb));

            var curClass : int = 0;
            for(i  = 0; i < objects.length ; i++)
                if(i > Math.round((curClass + 1) / numClasses * objects.length) - 1)

                sobjects[i].shape.setOptions(new PolygonOptions({fillStyle:
                        color : Color(colors[curClass]).rgb,
                        alpha : 0.7

There are 3 steps in the process :

  • Sorting the countries in order of their population value.
  • Obtaining the intermediate colors, between yellow and red. An interpolation on the values of the RGB components is performed. We get a color value for each class.
  • Finally, associating each country with a class, according to their population value. The breaks are performed at each quantile according to the “numClasses” parameter (in our case, quintile). The color is associated to the polygon by setting the “fillStyle” option of the PolygonOptions object. We keep the polygons somewhat transparent (70% opaque) so we can still see the ground.

And this is it! After a very short time, the polygons are all displayed.

Now about the speed: It has to be said it is pretty fast. Check for yourself. As a comparison, here is the same dataset, displayed on the AJAX API. Note that however, the method to add the polygons is not the same: The latter one uses the GGeoXml class to add a KML file. As noted above, this is not currently available on the Flash API.

The source code

Here is a link to an archive containing the code, with the Flex Builder 3 project, as well as the “” python script.



Oosterdok @ Amsterdam


Passenger Terminal Amsterdam

Passenger Terminal Amsterdam

Passenger Terminal Amsterdam

Bimhuis @ Passenger Terminal Amsterdam

ADA'M @ Passenger Terminal Amsterdam

More Photos




Name: Guilhem Vellut Location: Amsterdam

Blog Stats

  • 98,540 hits

Get every new post delivered to your Inbox.

Join 184 other followers