Center for Computational Research

CCR does its best to install and maintain widely used software packages.  We provide a suite of software that is standard for high performance computing environments.  This include compilers, Cuda, Anaconda Python, and various engineering, modeling, and statistical packages.

Group and user-specific software applications are generally beyond the scope of what CCR staff can maintain and support, so users and groups are encouraged to install these applications into their home or projects space.

If you require a specific version of a software package, even if CCR has this installed, we highly recommend you install your own copy of it.  We can only maintain one or two versions of software packages and can not guarantee the availability nor the compatibility of older versions of software.

Helpful tips and examples for installing user and group-specific software:

• When installing your own software:
  • Follow instructions and download packages for CentOS version 7.x
  • Install by specifying your home or project directory as the install destination (sometimes this is done using the --prefix=/path/to/directory option)
  • If a software requires a certain type of CPU or GPU, you may need to request that type of node in an interactive job and then compile the software while on that node.  For example, if you were to compile the software on the front end of the cluster, it may not work on older nodes.
  • Create your own custom modules for yourself or for your group to load the software's environment variables
    
    

• Perl modules
  • Setting up load Perl module installation
    
    
    
• Python modules/libraries:  
  • CCR maintains a few installations of recent versions of Anaconda python
  • Load the module for the version you want to use 
  • See our guide to using Conda environments
  • The Definitive Guide to Python Virtual Environments with Conda
    
    
• Python Virtual Environments & pip
  
  • An alternative to Conda is Python virtual environments which we recommend for most users
  • This is a great guide to get started using virtual environments on CCR's systems
  • This Python venv documentation is extremely useful
    
    
• R packages/libraries:
  • CCR maintains a few copies of the statistical package R.  
  • Load the module for the version you want to use
  • Install your own packages and libraries as described here

Please see our software installation policy for more details.