Skip to content

📦 Available software

A huge range of software applications, utilities, and libraries are installed and configured for you. Whether you need Rstudio or Spyder, Julia running in VSCode, popular R or Python packages, of fully configured Jupyter Notebooks, we have you covered.

The list of installed software is so large we make no effort to enumerate everything here, but you can always get an up-to-date list by opening a terminal on the Grid and running 

conda list

Start with the expectation that all the software you need is already installed and ready to use. If that expectation is ever broken please email research@hbs.edu.

The most recent version of our software environment is activated when you log in to the HBS Grid via NoMachine and nothing further is required unless you wish to use an environment other than the default.

When logging in via SSH from a terminal (i.e., without NoMachine) you must explicitly activate the environment you wish to use after logging in. The default environment can be activated by running 

ml rcs

Refer to the terminal environment documentation for details.

Installing Packages or Modules

The HBSGrid uses the conda package manager to install most R packages and Python modules that researchers will need in a central, read-only location. As we cannot anticipate everyone's requirements, you may have the need to install packages not centrally installed.

We recommend the following: - Install packages/modules in home folders for personal, testing, or development work, OR - Install packages/modules in project spaces for production or team use

One can use packages in all three locations, and can have the different versions of the same package installed in different locations. Note that it is the search path of your R or Python environment that will determine the priority order. We usually recommend (highest to lowest): - home folder ($HOME or ~) installations - project folder installations - central (cluster) installations

Installing R Packages

For installation in home folders: Use the standard command for install.packages() command from within R/RStudio will download and install the specified packages in your home folder by default. If the correct install folder is not already there, R/RStudio will create it for you:

install.packages('somepkg')

See the R documentation for additional options (install from another location besides CRAN, overwrite current installation, etc.)

For installation in project spaces: If you are working on a team and this is something for everyone to use, install the packages using the 'lib=' option:

install.packages('somepkg', lib='/path/to/project/R/')

In order for this to be seen when working on your project in this location, one can pre- or post-pend this location with a .libPaths() command as part of a startup script that runs when one opens a project or via new R session.

Note: as R and RStudio are managed via our central conda installation, one can also use conda and conda environments to manage our packages, which we don't recommend.

Installing Python Modules

Python modules can be installed via two methods: using pip install (recommended) or via creating conda environments. The latter is more heavyweight, as one is installing all the python infrasture and core, in addition to the modules you need. See our conda environments documentation if this is what is needed.

For installation in home folders: Include the --user option to perform this installation:

pip install --user some_module

To update/upgrade a module already installed, include also the --upgrade option: 

pip install --upgrade --user some_module

The installed modules will be located at $HOME/.local/lib/pythonxx.yy, with xx.yy being the python major.minor version.

For installation in project spaces: If you are working on a team and/or for production work, install the modules with pip install, though use the --prefix option instead:

pip install --prefix /path/to/project/.local some_module

Similar to R, one will need to include this location in the PYTHONPATH using sys.path as a part of one's first few lines of code or as part of a Python startup script.

Installing Compatible Software on the HBSGrid Cluster

Ideally, the application you need will have a version compatible with the cluster's OS (operating system; currently Red Hat Enterprise Linux 7, also known as RHEL 7.) Examples include the PyCharm Python IDE and fzf command-line fuzzy-finder. If you have a compatible application, using the terminal, download and extract the program; run any installer provided (if there is no installer just move the program wherever you like); and run the software directly from the terminal command line. If the software is available in multiple versions, pick one that mentions "RHEL", "RPM", "Linux x86-64 / AMD64", or just "Linux".

As an illustration, we can install PyCharm as follows:

  1. Connect to the HBSGrid via NoMachine, start Terminal from the menu, and download Pycharm: 

wget https://download.jetbrains.com/python/pycharm-community-2020.1.3.tar.gz
2. You will notice that the file has a '.tar.gz' extension, indicating that it needs to be decompressed with tar: 
tar -xvf pycharm-community-2020.1.3.tar.gz
3. Submit an interactive job to run Pycharm on a compute node: 
bsub -I pycharm-community-2020.1.3/bin/pycharm.sh
If all goes well you will see the Pycharm splash screen and you can start a new Python project using this popular IDE.

If the application you need does not have a pre-built version compatible with the HBSGrid system or you are having issues installing your compatible software, reach out to research@hbs.edu.

Note: Please do not use sudo, apt-get, dnf, or rpm programs, as these require elevated privileges, which are reserved for the cluster administrators and not mere mortals such as ourselves.