Base data science container

A container for data science things.

Based on the jupyter/datascience-notebook.

Running

Use a run label

podman container runlabel -n jupyter run ghcr.io/lysogeny/jupyter-ds-base:master

Or manually

podman run --rm \
    -v "${PWD}":/home/jovyan/work:Z \
    --name jupyter \
    -p 8888:8888 \
    --userns=keep-id:uid=1000,gid=100 \
    ghcr.io/lysogeny/jupyter-ds-base:master

Build the image and run the container

podman build -f base/Containerfile -t extended-jupyter 
podman run --rm \
    -v "${PWD}":/home/jovyan/work \
    --name jupyter \
    -p 8888:8888 \
    --userns=keep-id:uid=1000,gid=100 \
    extended-jupyter

Contents

Since this builds on the jupyter/datascience-notebook:

• Jupyter Lab
• Python with common packages (matplotlib, pandas, numpy, ...)
• R with common packages (tidyverse)
• Julia with some packages

Additionally:

• jupyter-lsp Plugin for LSP support
• LanguageServer.jl, r-languageserver and pyright language servers
• jupyter_vim for vim-like key binds
• jupyterlab_execute_time for cell execution times

Additional Julia packages:

• CSV.jl
• DataFrames.jl
• DataFramesMeta.jl
• Plots.jl
• StatsPlots.jl

Layering

To layer this you may want to do something like this

FROM ghcr.io/lysogeny/jupyter-ds-base:master

USER ${NB_USER}

RUN mamba install -c conda_forge scanpy
RUN julia -e 'using Pkg; Pkg.add("DifferentialEquations"); Pkg.precompile()'