ChronoRay

PyChrono + Ray Integration for Distributed Simulation Workflows

---

Installation

There are two installation options depending on whether you already have PyChrono installed.

---

Option 1 — You already have PyChrono

Use this if PyChrono is already installed in your environment.

pip install -e .

This will install ChronoRay along with its remaining dependencies (ray, numpy).

---

AMD ROCm + Ray (PyTorch HIP)

If you use PyTorch ROCm and Ray on AMD GPUs, read docs/AMD_ROCM.md. The ChronoRay package sets the recommended RAY_EXPERIMENTAL_NOSET_* environment variables on import; you can also call prepare_rocm_ray_env() from ChronoRay before importing Ray in standalone scripts.

---

Option 2 — You need everything (including PyChrono)

Use this if you are starting fresh. This creates a full conda environment with PyChrono and all dependencies.

1. Create the environment (miniconda or equivalent required):

conda env create --file environment.yml -n chrono-ray

2. Activate it:

conda activate chrono-ray

3. Install ChronoRay:

pip install -e .

---

More Info To Come Later