Create a condo environment on JupyterHub and add it to JupyterHub

Setting up the environment

• In the JupyterLab launcher, click "Terminal" to launch the terminal.
• Change working directory to directory where you will keep the environment definition.

cd <directory_with_environment_definition>

• Create and edit the conda environment YAML description file to specify your environment. Make sure that it has ipykernel installed in it. For example

  
name: <your_environment_name>
channels:
  - bioconda
  - conda-forge
dependencies:
  - ipykernel
  - snakemake-minimal >=5.24.1
  - macs2 >=2.2.7.1
variables:
  PYTHONPATH:

Make sure to unset the PYTHONPATH environment variable as in the example above.

• Activate the base conda environment

conda activate

• Create a conda environment that contains the necessary tools, including jupyterlab

mamba env create --file <environment_description_file>.yaml

• Confirm the creation by listing your conda environments

conda env list

and ensuring that an environment with the name <your_environment_name> has been created.

Add environment to JupyterHub

• Make sure that your environment, besides the other things you need, also has ipykernel installed in it.

conda activate <your_environment_name>
conda list | grep ipykernel

• If it does not contain ipykernel, it may be installed like so

mamba install ipykernel -n <your_environment_name>

• Activate your environment

conda activate <your_environment_name>

• Install the kernel

python -m ipykernel install --user –-name <your_environment_name> –-display-name “Python (<your_environment_name>)” --env PATH ~/.conda/envs/<your_environment_name>/bin:${PATH}

• Reload the JupyterHub webpage in the browser to display the newly added kernel in your Launcher.