Creating a dataset

Creating a dataset is one of the more fundamental aspects of using Opendatasoft. It might seem intimidating at first, but in a few steps you can be up and running.

To create a dataset, in your back office click the Create dataset button in the upper-right. Then perform the following steps:

1. Add a source

When you create a dataset, you start by adding a source.

You can add a source to a dataset in the following ways:

• Retrieve a file from your computer
  (See this page for instructions on how to create a dataset with multiple files.)

• Create a dataset from a URL, API, or FTP server

• Retrieve an Opendatasoft dataset

• Configure a remote service

• Create a datasource via the Automation API.

Note that you can always create a dataset without a source, for example to store only metadata, by clicking I do not wish to add a source for the moment just underneath "Create a dataset."

1. Configure the connection to the source

To retrieve data from your source, you must establish the connection between this source and the Opendatasoft platform. For connections via an URL, from an FTP server, and from some other remote services, you can configure and save a new connection or reuse a saved connection.

For more information about the sources you can connect to and their configuration, see Supported formats and Connectors to remote services.

1. Configure the source to make sure your data looks good

From the preview, you can configure your source.

1. Configure the dataset's metadata and access rights

From the Dataset information screen, you can configure the dataset title and technical identifier to keep or edit the default values. You can also set up the dataset access rights.

After you create a dataset, you can add a new source, or edit an existing source and its connection using the Sources tab of the desired dataset.

Good metadata is often the sign of a good dataset. Be proactive! See our blog for more information about the value of good metadata.