UB CCR maintains a "stock" version of Anaconda Python that is updated a couple times each year.  You will not be able to add packages to this.  You can request through CCR help the installation of packages but if it is not deemed useful for the majority of CCR users or it might conflict with other packages, we will need to decline the request.  You can install your own packages using CCR's Anaconda installation; however, when we update it might affect your package installations.  More details can be found here.

If CCR's installation is not suitable for your work you  should install and maintain your own version of the software.


NOTE: If you're using this installation within OnDemand, leave out the 'module load' command in the .bashrc and make sure to follow the instructions in #2 in this article



We do not recommend installing Anaconda in your home directory as it takes up the majority of your quota.  Create a directory in your project space - for example: /projects/academic/<group_name>/<your_user_name> or if you'd like it shared with all group members: /projects/academic/<group_name/anaconda)


Download the Anaconda installer and move it into your installation directory.  


Installation instructions can vary depending on version and can be found on the Anaconda website.  Here we'll demonstrate an installation of Anaconda3 (the script name you download may not be exactly the same).  You'll run the installation script and direct it to install in the directory you just created in your projects area using the --prefix option:


./Anaconda3-2020.02-Linux-x86_64.sh --prefix=/projects/academic/<group_name>/<install_dir>


Setting a temporary directory:
During installation you may see an error such as
"conda.exe: error while loading shared libraries: libz.so.1: failed to map segment from shared object: Operation not permitted"
This relates to an issue storing temporary files created during installation. Please create a temporary directory within your project directory and specify that location in the installation command. For example:

mkdir /projects/academic/<group_name>/condatemp
TMPDIR=/projects/academic/<group_name>/condatemp ./Anaconda3-2020.02-Linux-x86_64.sh --prefix=/projects/academic/<group_name>/<install_dir>



Step through all the installation prompts until completed.  Now create a directory for the anaconda python module file.  If you want to share this with everyone in your group, we recommend:

mkdir -p /projects/academic/<group_name>/modulefiles/python 



Now copy over an existing UB CCR anaconda module file (the version may differ, generally you want something recent, which you can check with "module avail python"):

cd /projects/academic/<group_name>/modulefiles/python

cp /util/common/modulefiles/Core/python/recent-python.lua  my-python-3.lua


(where recent-python may be something like py38-anaconda-2021.05.lua). It is important to give your module a new name.  If you're using this with the OnDemand Custom Jupyter Notebook app, make sure to name your module 'jupyter'



Now edit the anaconda module file so that it references your installation rather than the one in the "/util/" area.  You can edit the file in the OnDemand file manager or on the command line:

 gedit my-python-3.lua



Finally, use the following sequence to load your anaconda module: 


module use /projects/academic/<group_name>/modulefiles

module load python/my-python-3 (see note below)


NOTE: If you're using this installation within OnDemand, leave out the 'module load' command in the .bashrc and make sure to follow the instructions in #2 in this article



We provide detailed guides on how to use your Anaconda environment on our systems