Data science development environment

A ready-to-use data science environment for VS Code, designed for intro Python and ML bootcamp students. Covers data visualization, data cleaning, feature engineering, and traditional machine learning. Available in three configurations: NVIDIA GPU, CPU-only, and Mac (Apple Silicon).

What's included

Package	Purpose
numpy, pandas, scipy	Core data science stack
scikit-learn, xgboost, statsmodels	Machine learning and statistics
matplotlib, seaborn, plotly	Visualization
optuna	Hyperparameter optimization
jupyterlab	Interactive notebooks
cupy-cuda12x	GPU-accelerated arrays (NVIDIA only)

Devcontainer configurations

Configuration	Image	Use when
DataScience NVIDIA	gperdrizet/datascience-nvidia	You have an NVIDIA GPU
DataScience CPU	gperdrizet/datascience-cpu	CPU-only machine (any OS)
DataScience Mac	gperdrizet/datascience-mac	Apple Silicon Mac (M1/M2/M3)

Project structure

datascience-devcontainer/
├── .devcontainer/
│   ├── nvidia/
│   │   └── devcontainer.json   # NVIDIA GPU dev container configuration
│   ├── cpu/
│   │   └── devcontainer.json   # CPU dev container configuration
│   └── mac/
│       └── devcontainer.json   # Mac (ARM64) dev container configuration
├── data/                       # Store datasets here
├── notebooks/
│   └── environment_test.ipynb  # Verify your setup
├── .gitignore
├── LICENSE
└── README.md

Requirements

• Docker (Windows | Linux)
• VS Code with the Dev Containers extension

NVIDIA configuration (additional requirements)

• NVIDIA GPU (Pascal or newer) with driver ≥570
• NVIDIA Container Toolkit (Linux): install guide

Mac configuration

• Docker Desktop for Mac (Apple Silicon): install guide

Note: GPU acceleration is not available inside Docker containers on Apple Silicon. Metal/MPS is a macOS-only framework with no Docker passthrough. The Mac configuration provides native ARM64 CPU performance.

GPU compatibility (NVIDIA)

Architecture	Example GPUs	Compute Capability
Pascal	GTX 1050-1080, Tesla P100	6.0-6.1
Volta	Tesla V100, Titan V	7.0
Turing	RTX 2060-2080, GTX 1660	7.5
Ampere	RTX 3060-3090, A100	8.0-8.6
Ada Lovelace	RTX 4060-4090	8.9
Hopper	H100, H200	9.0
Blackwell	RTX 5070-5090, B100, B200	10.0

Quick start

1. Fork this repository (click "Fork" button above)

2. Clone your fork:

   git clone https://github.com/<your-username>/datascience-devcontainer.git

3. Open VS Code

4. Open Folder in Container from the VS Code command palette (Ctrl+Shift+P), start typing Open Folder in...

   VS Code will prompt you to choose a devcontainer configuration. Select the one that matches your hardware.

5. Verify by running notebooks/environment_test.ipynb

Using as a template for new projects

One-time setup: Make your fork a template

1. Go to your fork on GitHub
2. Click Settings → scroll to Template repository
3. Check the box to enable it

Creating a new project from your template

1. Go to your fork on GitHub
2. Click the green Use this template button → Create a new repository
3. Enter your new repository name and settings, click Create repository
4. Clone your new repository:

   git clone https://github.com/<your-username>/my-new-project.git

Now you have a fresh data science project with the dev container configuration ready to go!

Adding Python packages

Using pip directly

Install packages in the container terminal:

pip install <package-name>

Note: Packages installed this way will be lost when the container is rebuilt.

Using requirements.txt (recommended)

1. Create a requirements.txt file in the repository root:

   lightgbm
   shap
   

2. Update the appropriate devcontainer.json to install packages on container creation:

   "postCreateCommand": "pip install -r requirements.txt"

3. Rebuild the container (F1 → "Dev Containers: Rebuild Container")

Keeping your fork updated

# Add upstream (once)
git remote add upstream https://github.com/gperdrizet/datascience-devcontainer.git

# Sync
git fetch upstream
git merge upstream/main

Troubleshooting

Problem	Solution
Docker won't start	Enable virtualization in BIOS
Permission denied (Linux)	Add user to docker group, then log out/in
GPU not detected	Update NVIDIA drivers (≥570), install NVIDIA Container Toolkit
Container build fails	Check internet connection
Module not found	Rebuild container after adding to requirements.txt