NR-SCFT-ML: Neutron Reflectivity SCFT Machine Learning

A Python package for Neutron Reflectivity data analysis using Machine Learning. Including Chi parameters prediction from SLD profile data and SLD profile prediction from NR curves.

Overview

NR-SCFT-ML is a PyPI package for efficient processing and analysis of Neutron Reflectivity (NR) data. It provides a streamlined pipeline for:

• Preprocessing NR datasets
• Predicting SLD Profile by training a CNN
• Predicting Chi parameters by training a combined Autoencoder and MLP model.

Prerequest

Conda
Python >=3.10

Quick Start Guide

Prepare the Package Environment on HPC or Your Local Machine (Large Memory Required). The setup.sh creates a Jupyter kernel environment named PyreflectEnvironment.

run in cell:

!curl -fsSLo setup.sh https://raw.githubusercontent.com/williamQyq/pyreflect/main/setup.sh
!bash setup.sh

To install from TestPyPI, run:

%pip install -i https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple pyreflect==1.3.1

🧑🏻‍💻Example Notebooks

To learn how to use it, check the example notebooks in pyreflect/example_notebooks or watch the tutorial video. Click below:

1️⃣ Initialize Configuration

To initialize the default configuration for model path, hyperparameters, and data reading paths, run:

python -m pyreflect init --force

Use the --force flag if the configuration has already been created and you want to overwrite it.

2️⃣ Run Training & Prediction

Open settings.yml and update the file paths for your SLD profile data and Chi parameters

**2️⃣.1️⃣ Training datasets and Train model

Generate training datasets using the example in example_notebook_autoencoder.ipynb.
The memory usage should be properly managed by controlling the number of curves generated for training, as it will consume a large amount of memory.

Train the model using the example in example_notebook_SLD_prediction.ipynb.

**2️⃣.2️⃣ A simple Experimental Datasets - from real world lab

datasets contains processed experimental NR, its manual fit SLD profile data, and AE denoised experimental NR data.

3️⃣ Run interaction chi parameters prediction

python -m pyreflect run --enable-chi-prediction

4️⃣ Run sld profile prediction from nr curves(or import package lib in notebook)

python -m pyreflect run --enable-sld-prediction

Credits

This project builds on work by:

• Brian Qu (NR-SLD-CNN)
• Dr. Rajeev Kumar
• Prof. Miguel Fuentes-Cabrera NR-SCFT-ML

Author

• Yuqing Qiao (William) – Maintainer and developer of this PyPI package