Installation

CDOpt is compatible with Python 3.6+, and depends on NumPy and SciPy. CDOpt is available to install via the Python Package Index, and can be installed via pip,

pip install cdopt --upgrade

Moreover, to use the advanced features of CDOpt, including the automatic differentiation, customized manifolds, CUDA supports and training neural networks, the users need to install some numerical packages, such as autograd, PyTorch, or JAX. We strongly recommend the users install the following packages

• Autograd >=1.0

• PyTorch >= 1.7.0

• JAX >= 0.3.12 (optional since JAX is not available on windows platform)

Please follow the installation guide for autograd, PyTorch, and JAX to install these packages.