Installation

Prerequisites

• Python 3.11 (required; the project pins requires-python = "==3.11.*")
• uv for dependency management

Install uv

curl -LsSf https://astral.sh/uv/install.sh | sh

Install cdr_bench

git clone https://github.com/AxelRolov/cdr_bench.git
cd cdr_bench
uv sync

This installs all required dependencies including RDKit, UMAP, openTSNE, ChemographyKit, Numba, and DGL-Life.

Verify installation

python -c "from src.cdr_bench.dr_methods.dimensionality_reduction import DimReducer; print('OK')"

Optional: GPU Support

ChemDist embedding generation (via DGL-Life and PyTorch) benefits from GPU acceleration. If you have a CUDA-capable GPU, ensure your PyTorch installation includes CUDA support. The device option in bench_configs/features.toml controls whether to use "cuda" or "cpu".

Development Setup

To install with development dependencies (includes JupyterLab):

uv sync --group dev

To install documentation dependencies:

uv sync --group docs

To install all dependency groups:

uv sync --all-groups