Skip to content

Finding Data

One of the first steps when working on a project in OnDemand Loop is locating the right dataset — typically for downloading data into your project. The application supports two ways of finding datasets from remote repositories:

  1. Using the repository website to browse and search, then pasting a URL or DOI in OnDemand Loop
    Best for advanced users who know where the dataset is located.
  2. Browsing and searching directly inside the OnDemand Loop interface
    Ideal for new users or those still exploring available datasets.

Supported Repositories

OnDemand Loop supports downloading datasets from a growing number of repository connectors.
See the Supported Repositories section for the current list. Contributions to support additional repositories are welcome!

Public datasets with published versions are supported by default.
Access to unpublished or draft content may require an API key, depending on the repository's access policies.

Choosing the Best Approach

Method Best For Benefits
Paste URL or DOI Experienced users, known datasets Fast and direct. Leverages repository search tools.
Browse within OnDemand Loop New users, exploratory workflows Fully integrated. No need to leave the app.

Exploring Remote Repositories

There are two ways to explore datasets from a remote repository when downloading data:

1. Paste a Repository URL or DOI

If you already know the dataset you need, this is the fastest option:

  1. Visit the repository’s website.

  2. Find the dataset and copy its DOI or full URL. For example:

    • A DOI:
      doi:10.7910/DVN/MYSRMN
      https://doi.org/10.7910/DVN/MYSRMN

    • A full URL from a repository:
      https://dataverse.harvard.edu/collection
      https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/MYSRMN
      https://zenodo.org/record/1234567
      https://zenodo.org/deposit/1234567

    • In the Explore bar at the top of the app, paste the DOI or URL.
    • Click Explore to connect and browse the dataset locally.
    • From there, you can select and download individual files into a project.

OnDemand Loop will:

  • Automatically detect the repository (based on the address)
  • Invoke the appropriate connector
  • Display the dataset’s metadata and file list

Tip

This method is especially efficient if you're working from a publication, citation, or already bookmarked dataset.

2. Browse and Search Within OnDemand Loop

If you prefer an integrated experience or are still exploring:

  1. Click on the Repositories menu in the app.
  2. Choose a supported remote repository, like Dataverse or Zenodo.
  3. In case of the existence of multiple repository installations, select the desired one.
  4. Use the built-in search to look for datasets.
  5. Navigate the results and select a dataset to explore.

Once a dataset is selected, OnDemand Loop displays its metadata and file listing, allowing you to choose which files to download into your project.

Tips for Working with Drafts

  • For Dataverse datasets, you can choose the dataset version — including the current draft — if available.
    A repository API token is required to access and view draft versions.

  • For Zenodo depositions, a registered API key is needed to load any unpublished or draft content.

Repository Capabilities May Vary

OnDemand Loop uses each repository's public API to retrieve metadata, file listings, and search results.
As a result, features such as search, filtering, and metadata previews may differ slightly between connectors.

Launching from Dataverse

If the administrator of your Dataverse installation has configured OnDemand Loop as an External Tool, you can launch the application directly from the dataset page.

  • Open a dataset in a Dataverse repository.
  • Click the Access Dataset dropdown.
  • Select Explore in OnDemand Loop.

This will open the OnDemand Loop interface with the dataset preloaded and ready for browsing or download.

External Tools

This feature is only available for Dataverse and depends on the repository admin having enabled it. If you don't see the option, contact your Dataverse administrator. For Dataverse administrators: follow the instructions on how to configure Dataverse external tools