kwsp / mcml-py

Port of MCML (Monte Carlo Multi-Layered) to Python.

MCML Python

Port of MCML (Monte Carlo Multi-Layered) to Python.

Requires Python 3.10, numba, and tqdm. As of Nov 2022, numba doesn't yet support Python 3.11.

Getting started

1. Clone the repo (or download the zip file and extract).

2. Create a virtual environment. Use Miniconda if you don't know what that means. After installing miniconda, open "Anaconda Powershell Prompt" and run

conda create -n mcml python=3.10 numba tqdm

This creates a new virtual environment named mcml, and installs Python 3.10 along with numba and tqdm.

4. Activate the mcml virtual environment:

conda activate mcml

5. cd into the working directory. This is where you cloned this git repo, or where you unzipped the downloaded zip file.

cd [path to mcml-py]

6. Run the MCML main program with a sample input file.

On Windows:

python main.py sample.mci

On Unix

python3 main.py sample.mci

Accuracy

The program is checked against the original MCML program. Errors should be <1%.

Speed

Most of this program is JIT compiled with numba, hence the speed is reasonable. Simulating 1,000,000 photons in a 3 layer medium takes <20s on my laptop.