Skip to content

Install Conda and Python

  • Connect to Amarel Open OnDemand
  • Click Clusters
  • Choose Amarel Cluster Shell Access
  • Enter your password
  • In the terminal, do the following commands (one line each time). If you're using Windows, type 'Ctrl+c' to copy and 'Ctrl+Shift+v' to paste command. 
$ module use /projects/community/modulefiles
$ module load anaconda/2020.07-gc563
$ cd
$ source .bashrc
$ mkdir -p .conda/pkgs/cache .conda/envs
  • Test if conda is successfully installed: 
$ which conda
  • Install a conda environment called 'rcaes_env': 
$ conda create -n rcaes_env
  • Enter Y to proceed. Wait until you see the following. It may take a while. 
#                                                                                                                                                               
# To activate this environment, use                                                                                                                             
#                                                                                                                                                               
#     $ conda activate rcaes_env                                                                                                                              
#                                                                                                                                                               
# To deactivate an active environment, use                                                                                                                      
#                                                                                                                                                               
#     $ conda deactivate
  • Now let's activate the environment:
$ conda activate rcaes_env
  • First, let's install essential packages
conda install -c conda-forge python=3.9 jupyter jupyterlab notebook numpy scipy ipython
  • Next, let's install github CLI, a command-line interface to GitHub for use in your terminal or your scripts
conda install -c conda-forge gh

  • Next, go back to Amarel Open OnDemand. This time, we will launch a personal jupyter. Click on 'Interactive Apps', choose 'Personal Jupyter'.

  • Settings for Personal Jupyter: 

    Number of hours: 10 
    Number of cores: 1 
    Gigabytes of memory: 10 
    Partition: main
    Leave Reservation and slurm feature blank 
    conda path: /projects/community/anaconda/2020.07/gc563 
    conda environment: rcaes_env

Conda Alternative: Mamba

  • If you find Conda super slow at solving the environment, I'd recommend you try using an alternative package manager Mamba.

  • Install Miniforge using: 

wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
bash Miniforge3-$(uname)-$(uname -m).sh

  • Close and reopen the terminal window.

  • Create a new environment rcaes_env_new: 

mamba create -n rcaes_env_new python=3.9 jupyter jupyterlab notebook numpy scipy ipython pandas matplotlib cartopy geopandas xarray dask netCDF4 bottleneck gh

  • Next, go back to Amarel Open OnDemand. This time, we will launch a personal jupyter. Click on 'Interactive Apps', choose 'Personal Jupyter'.

  • Settings for Personal Jupyter: 

    Number of hours: 10 
    Number of cores: 1 
    Gigabytes of memory: 10 
    Partition: main
    Leave Reservation and slurm feature blank 
    conda path: /home/YOURNETID/miniforge3
    conda environment: rcaes_env_new