Skip to content

Visualizations

When browsing Clowder, there are two primary ways users can see aspects of a file or dataset without downloading it:

  • Thumbnails, which are lightweight static images that appear in grid views such as search results; and
  • Visualizations, which appear on individual resource pages and trigger JavaScript modules to create more specific preview components such as map views or interactive widgets.

Both of these can render actual bytes stored in the system, but Visualizations support more complex behavior. Typically, both Thumbnails and Visualizations are generated by extractors listening for new file uploads.

Thumbnails

To programatically add a thumbnail to a file or dataset:

  1. POST an image to the api/v2/thumbnails endpoint to upload it to the system. The response will include a thumbnail_id.
  2. Associate the thumbnail with one or more resources by issuing a PATCH request to api/v2/files/:file_id/thumbnail/:thumbnail_id or api/v2/datasets/:dataset_id/thumbnail/:thumbnail_id.
  3. Retrieve the thumbnail (e.g. to render on a frontend GUI) by issuing a GET request to api/v2/files/:file_id/thumbnail.

The ncsa.image.preview extractor found in the core extractors will create file thumbnails automatically.

Visualizations

Visualizations can be more complex than Thumbnails, and as such they include a JSON config document along with any actual bytes necessary to render.