Skip to content

Get started

Use the CANFAR Science Platform to analyse astronomical data in your browser. You can run notebooks and astronomy applications near your data, save results, and share them with your project. You do not need to install the Python client to follow this guide.

1. Get an account and access

  1. Request a Canadian Astronomy Data Centre (CADC) account.
  2. Get access to the platform in one of two ways. If your project already uses CANFAR, ask its administrator to add you to the project's group. If not, email support@canfar.net with your CADC username and a short description of your research.

See who can use CANFAR for eligibility. If you need help, contact CANFAR support.

2. Run your first notebook

  1. Sign in to the Science Portal.
  2. In Launch New Session, choose Notebook, select an image containing the software you need, and submit the form. An image is a packaged software environment; a Session is your running instance of it.
  3. Find your Session under Active Sessions. Wait for it to start, then open its notebook link.
  4. Follow Create, run, and save a notebook to run your first Python cell and save the notebook in persistent storage.

Choose an image from the portal's current list. Available software and resource choices depend on the image and server.

3. Upload data and save your work

Inside a Session, use these mounted paths when they are available:

Location Use
/arc/home/USER Personal scripts, notebooks, and results; replace USER with your username
/arc/projects/PROJECT Shared project data and outputs; replace PROJECT with your project's directory
/scratch Temporary files; deleted with the Session

Use browser file transfers to upload inputs or download results. Save and verify your notebook and results under /arc before using the Session's delete control in Active Sessions. Files under /arc remain available after deletion. Closing a browser tab alone does not stop the Session.

Your first analysis is saved

Your notebook and results should appear under your chosen /arc directory. You can open them from a new Session after deleting this one.

Choose Desktop, CARTA, or Firefly when you need a different analysis interface.

4. Automate your analysis (optional)

For repeated work, install the client and follow the command-line tutorial or Python tutorial. Check the installation guide's release note before using examples from this branch.

To run an unattended workflow:

  1. Test your command on a small input in an interactive Session.
  2. Set explicit input and output paths in the command or environment.
  3. Submit a headless Session, which runs without a browser interface.
  4. Inspect its progress with canfar ps --all, canfar info, and canfar events.
  5. Copy final results out of /scratch before deleting the Session.

Next steps