Interactive Analysis Platform

Last update: 20 Nov 2024 [History] [Edit]

Introduction to SWAN

SWAN

CERN provides a web-based interactive analysis platform called SWAN (Service for Web-based Analysis). SWAN is built on Jupyter notebooks, which allow you to write and execute python code interactively. SWAN provides access to CERN software, computing resources and storage for use in your analysis.

tip If you are following this tutorial at a different venue, your instructor might have you use an alternative system.

Start a session

To start a SWAN session, navigate to https://swan.cern.ch/ and sign in with your CERN credentials. Begin a new session with the following settings:

  • Software stack: 102
  • Platform: CentOS 7 (gcc11)
  • Environment script: default
  • Number of cores: 4
  • Memory: 16 GB
  • Spark Cluster: None
  • HTCondor pool: None

Then go ahead and click “Start my Session”.

tip You can get the same environment on your LXPLUS terminal (after you setupATLAS) with the command:
lsetup "views LCG_102 x86_64-centos9-gcc11-opt"

Click on the terminal icon on the top right of the screen:

SWAN

In the new terminal, clone the ntuple analysis git repo:

git clone https://gitlab.cern.ch/atlas-analysis-sw-tutorial/plotting.git

Click on the cloud home icon:

SWAN

Go to the CERNBox tab:

SWAN

Click on plotting and then open the appropriate Jupyter notebook (ipynb file). Each of the following sections will suggest a notebook for you to work through.