Skip to content

Installation

The project supports both uv and pip. A pinned uv.lock is committed to the repository, so uv sync produces a reproducible install across machines.

Basic Install

git clone git@github.com:ramanathanlab/deepdrivewe-academy.git
cd deepdrivewe-academy
uv sync

uv sync creates a .venv/ in the repo, installs the project in editable mode, and resolves all dependencies from uv.lock. Activate the environment with source .venv/bin/activate, or prefix commands with uv run (for example, uv run python -m deepdrivewe ...).

With pip

git clone git@github.com:ramanathanlab/deepdrivewe-academy.git
cd deepdrivewe-academy
pip install -e .

Full Install with MD Dependencies

For running molecular dynamics simulations with OpenMM and AmberTools, use conda to install the simulation backends first, then install the Python package with either uv or pip:

git clone git@github.com:ramanathanlab/deepdrivewe-academy.git
cd deepdrivewe-academy

conda create -n deepdrivewe python=3.11 -y
conda activate deepdrivewe
conda install -c conda-forge openmm=8.1
conda install omnia::ambertools -y
pip install -e .          # or: uv pip install -e .

Deep Learning Models

To use the built-in AI models (convolutional VAE, adversarial autoencoder), install the correct version of PyTorch for your system and GPU drivers:

pip install torch

Note

The mdlearn dependency may require an earlier version of PyTorch. If you encounter compatibility issues, try:

pip install torch==1.12

Development Setup

For contributing or running the test suite:

# uv (recommended)
uv sync --extra dev --extra docs
uv run pre-commit install

# pip
python -m venv venv
source venv/bin/activate
pip install -U pip setuptools wheel
pip install -e '.[dev,docs]'
pre-commit install

Verify the setup:

uv run pre-commit run --all-files   # or: pre-commit run --all-files
uv run pytest                       # or: pytest

Documentation Build

To build the documentation locally:

# uv
uv sync --extra docs
uv run properdocs serve

# pip
pip install -e '.[dev,docs]'
properdocs serve

Then open http://localhost:8000 in your browser. For a production build with strict checking:

uv run properdocs build --strict   # or: properdocs build --strict

See the Contributing guide for more details.