Conda

This page contains useful commands/tips for conda usage

1 Frequently used commands

Code

# Deletes the ENV_NAME along with all its packages
# (This one requiers you to confirm yes)
conda remove -n ENV_NAME --all 
conda remove -n ENV_NAME --all -y # This one do not ask for confirmation

Code

# List all environments
conda env list

Code

# creates ENV_NAME could use either with -n or --name
conda create -n | --name -y ENV_NAME

2 Combine usage with JupyterLab

JupyterLab enables you to use interactive kernels of conda, such that each of the kernel could have a conda manager and pip manager. And, there are two frequent kernels: - ipykernel - iRkernel

It is recommended to separate Python and R kernel usages, say have different conda environments for each individual Python/R kernel. And, having nb_conda_kernels in the environment that you launch Jupyter Notebook or JupyterLab.

For example, this is how I use it:

Install nb_conda_kernels in the conda environment that has jupyter lab, in my case, I use the base env.
Create new conda envs for the projects that I do:

conda create -n -y python_proj ipyerkernel For Python
conda craete -n -y r_proj r-irkernel For R

Launch jupyter lab from base (by staying at base env), then when it launches you will be able to see the following:
Lastly, just choose the environment you are going to use

If you want to install packages to specified environments, do the following:

conda activate ENV_NAME , replace ENV_NAME of the one you like to deal and mess around with
conda install -y PKG_NAME or pip install PKG_NAME, install the packages you need depending on where they are supported (NOTE: It supports most python packages, but not quite for R packages). If you like to use more R packages, I recommend Docker or renv.