Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

ODS Widgets

ODS-Widgets is a library of web components that can be used to build rich and interactive pages of datavisualization, live from data available on a remote API.

In more concrete terms, it is a set of AngularJS directives that can be directly plugged on a remote data API (currently Opendatasoft only); those directives can interact with each others using shared contexts (for example a "calendar" directive could refine the data displayed by a "table" directive). It is designed to work without the need for a local backend, and to be efficient to build data visualization pages and dashboards by just manipulating HTML markup (AngularJS directive tags). Since it is regular HTML, it can be integrated in your current pages, supports CSS styling, and generally works in a way that will feel familiar if you know HTML. It does not require knowing AngularJS to be used.

This library is used in production as a part of the Opendatasoft platform, a Software-as-a-Service data portal. It is essentially the current core of our front-end data visualization features.

Here are a few ideas of things you could do with all this:

  • integrate a simple visualization (or a more complete visualization tool using more widgets) from a public dataset into a blog article (it is worth noting that, as more and more search engines are understanding JS applications like Google is, your integrated visualizations will be referenced, which can be very beneficial to your SEO)
  • within your application, allow your users to design and build complex and interactive dashboard or visualization pages on your data, using simple HTML tags (that's what we are doing, for example)
  • build a website on top of your data (or public data) without having to worry about consuming and exposing your data to end-users in a scalable and compelling way
  • fork and enrich this toolset with your own widgets, or even adapt it to your own needs and backend services: the project is licensed under the MIT license

Quick example

Let's show a map of the restaurants near our former office in Paris. The data can be browsed on

<!-- Expose a "dataset context" from the data -->
<ods-dataset-context context="restaurants" restaurants-domain="" restaurants-dataset="restaurants_a_proximite_de_pepiniere_27">
  <!-- Display a map from the data exposed from this context -->
  <ods-map context="restaurants"></ods-map>


You'll find a tutorial on how to use widgets, built using public data and live example.

> Tutorial

There is a complete reference documentation of all the available directives and all the parameters they support. It also contains more literature on the technical details of the library.

> Reference Documentation


Known to work on: (may not be exhaustive)

  • Internet Explorer 11
  • Edge (last versions)
  • Safari (last versions)
  • Chrome (last versions)
  • Firefox (last versions)


You can download the latest ODS Widgets release here. This is a ZIP file of a folder that you can then copy into your project.


The following libraries are required as dependencies in your pages for ODS-Widgets to work:

  • jQuery 2.1+ (versions 3+ has known incompatibilities)
  • AngularJS 1.8.0 and the angular-sanitize module (note: AngularJS 1.4 and above should be working as well, with known incompatibility on ods-range-input and ods-gist)
  • Some widgets are using icons from FontAwesome 4.4.0

To load ODS-Widgets, you can then simply load ods-widgets.js and ods-widgets.css, or their minified version also included in the distribution.

Note: When including the scripts in your page, you need to include jQuery, then AngularJS, then angular-sanitize, then ODS-Widgets; in that specific order.

ODS-Widgets require an AngularJS app to run, which can for example be done with a simple ng-app="ods-widgets":

  <div ng-app="ods-widgets">
    <!-- Within this div, we are in an Angular app where ODS-Widgets tags will run -->
    {{ "hello" + "from" + "angular!"}}

Some specific widgets can rely on specific "heavy" libraries (such as Highcharts, Leaflet...); in that case, they are lazily-loaded from a CDN when the widget initializes itself, with no necessary work on your part.

The bare minimum for an HTML page is this template:

<!DOCTYPE html>
        <title>ODS Widgets Sandbox</title>
        <meta charset="utf-8">
        <meta name="viewport" content="width=device-width, initial-scale=1">
        <link rel="stylesheet" href="">
        <link rel="stylesheet" href="//">

        <div ng-cloak ng-app="ods-widgets">

            <!-- YOUR CODE HERE -->


        <script src="//"></script>
        <script src="//"></script>
        <script src="//"></script>
        <script type="text/javascript" src="//"></script>

Some features like URL synchronization and translations will require additional configuration. You'll find more information in our Cookbook.

Available API and data sources

Currently, ODS-Widgets only works with an API from an Opendatasoft domain, or any API that exposes the same interface and returns the same JSON result.

This project originates from Opendatasoft codebase, and as such obviously works closely with our own interfaces; however, we are looking to make this framework work with more APIs and services (and thankfully, AngularJS can make it easier), for example by implementing more AngularJS services as sources, beside our ODSAPI service. If you have an API that exposes structured data in a way that looks like our widgets could display it, and you want to try to make it work with it, you can contact us.


This software library is licensed under the MIT license.


If you have questions, you can contact us on Twitter (@opendatasoft) or simply open a GitHub issue.


Components to easily build visualizations and interfaces from a data API.







No packages published