Julia API
This page explains how to install the Julia version of HPR-QP.
1. Prerequisites
Julia ≥ 1.10.4
CUDA: NVIDIA GPU with proper CUDA driver
System tools: Standard C/C++ toolchain (for dependencies)
Check that CUDA is available in Julia:
using CUDA
CUDA.versioninfo()
2. Pick the Right Solver and Clone the Repository
If you need to solve a LASSO problem (with an \(l_1\) regularizer) or a QAP instance where the matrix form of \(Q\) is unavailable, please refer to the HPR-QP_QAP_LASSO module.
git clone https://github.com/PolyU-IOR/HPRQLP.git
cd HPR-QP_QAP_LASSO
otherwise, for convex QP (COP) problems where the matrix form of \(Q\) is available, please refer to the HPR-QP module.
git clone https://github.com/PolyU-IOR/HPRQLP.git
cd HPR-QP
Or download and extract the ZIP file from GitHub.
3. Install Dependencies
Inside the project root, run:
julia --project -e 'using Pkg; Pkg.instantiate()'
This will install all required Julia dependencies (including CUDA.jl).
Next Steps
See the Interfaces section for Julia usage.