Note that conda activate will not work on Discovery with this version.
To activate your Conda environment, type source activate. Follow the prompts to complete the Conda install. To save space, also use the -prefix=/work// flag to build in your work directory. Tip: to see a list of all of your conda environments, type conda info -e. 
To create your environment, type conda create -n python=3.7 anaconda, where is the name you want to give your environment. To load anaconda, type module load anaconda3/2022.01. To check what version of Python you have installed, type which python. In the above example, we request for 1 node with 1 cpu core, but you can request for additional resources as per your requirements. If you are on a login node, move to a compute node by typing srun -partition=short -nodes=1 -cpus-per-task=1 -pty /bin/bash. This procedure uses the Anaconda module already loaded on Discovery. You can also have more than one environment with different packages for testing purposes. 
Using a locally installed Conda virtual environment is highly recommended so that you can install the specific packages that you need. Northeastern University Research ComputingĬreating a Conda virtual environment with Anaconda ¶
Message Passing Interface (MPI) Overview. Creating a Conda virtual environment with Anaconda.
0 kommentar(er)
