Devcontainer for Data Science (Python & R)

This is a devcontainer configuration for setting up a comprehensive Data Science development environment with Python and R in Visual Studio Code.

Main Features

Base Environment

• Image: Based on the quay.io/jupyter/minimal-notebook image.
• User: Runs as the jovyan user with sudo NOPASSWD privileges.
• OS: Ubuntu with a full build environment (build-essential).
• Locale: Configured for en_US.UTF-8.

Python Environment (via Conda)

• Core Packages: scanpy, leidenalg, pyarrow, hdf5.
• Terminal: radian is configured as the default R terminal for an enhanced console experience.

R Environment

• Installation: A comprehensive set of R packages for bioinformatics and data analysis is pre-installed using pak for efficiency.
• Key Packages:
  • Core: tidyverse, Seurat, DESeq2, harmony
  • Workflow: targets, gittargets, crew
  • Single-cell: tidyseurat, tidySummarizedExperiment, tidybulk, scuttle, scDblFinder, SingleCellExperiment, sccomp, mojaveazure/seurat-disk, scCustomize, cellgeni/schard
  • Utilities: skimr, httpgd, languageserver, qs, cmdstanr
  • Plotting: EnhancedVolcano, RColorBrewer, ggalign
  • Community/Custom: huayc09/SeuratExtend, saezlab/liana

Development Tools

• Quarto: Pre-installed version 1.6.39 for scientific and technical publishing.
• System Tools: htop, vim, parallel, gh (GitHub CLI), x-cmd.
• R Configuration (.Rprofile & .lintr):
  • VSCode Integration: R language server support is automatically configured.
  • Default Packages: tidyverse, targets, skimr, and gittargets are loaded by default.
  • Formatting: Custom styler settings are configured for the R language server.
  • Linting: A .lintr file is created to disable specific linters for a cleaner coding experience.
• Git: User name (h4rvey-g) and email are pre-configured globally.

Host-Container Integration

• SSH Access to Host: An initialize-host.sh script runs on the host to create and authorize an SSH key (~/.ssh/devcontainer_id_rsa). This allows the container to securely connect back to the host via ssh ${HOST_USER}@host.docker.internal.
• Docker Socket: The Docker socket is mounted, enabling Docker-in-Docker capabilities.
• SSH Key Mount: The host's SSH key is mounted into the container at /home/jovyan/.ssh/id_rsa.

VSCode Integration

• Default Settings: Pre-configured for Python (formatter, path) and R (terminal, path).
• Extensions: A curated list of extensions for Python, R, Jupyter, Docker, Git, and general productivity is automatically installed. See devcontainer.json for the full list.

Usage Instructions

1. Open in Dev Container: Open this repository in Visual Studio Code and use the "Reopen in Container" command.
2. Initialization:
   • On the host machine, the initialize-host.sh script will run once to set up SSH keys.
   • After the container is created, the postCreateCommand.sh script runs inside the container to finalize the environment setup (configuring Git, R profiles, SSH permissions, etc.).
3. Start Coding: The environment is ready for use. All tools and packages are pre-installed and configured.