This repo is the official project repository of the paper Rapid Structure-agnostic Single Crystal X-Ray Diffraction Indexing via Reflection Cloud Learning [ arxiv ] [ RC-400K ]
Authors:
我们建议对数据模拟部分和模型训练部分使用分别的环境避免潜在环境冲突问题，建议使用Nvidia BlackWell架构GPU，如GeForce RTX 50 Series。。如果只是复现论文结果，你可以跳过数据模拟步骤直接使用RC-400K数据集。数据集将在论文建刊后开源。
We build our simulator based on DATAD. Please first apply for datad.whl file at here.
mamba create -n rct python=3.12
mamba activate rct
mamba install -c nvidia/label/cuda-12.8.0 cuda-toolkit -y
mamba install cudnn gcc gxx -y
uv pip install ninja
uv pip install torch==2.8.0 torchvision==0.23.0 torchaudio==2.8.0 --index-url https://download.pytorch.org/whl/cu128
uv pip install h5py pyyaml
uv pip install sharedarray tensorboard tensorboardx wandb yapf addict einops scipy plyfile termcolor timm matplotlib
uv pip install torch-cluster torch-scatter torch-sparse -f https://data.pyg.org/whl/torch-2.8.0+cu128.html
uv pip install spconv_cu126-2.3.8-cp312-cp312-manylinux_2_28_x86_64.whl
uv pip install flash_attn-2.8.2+cu128torch2.8-cp312-cp312-linux_x86_64.whl
cd libs/pointops
python setup.py install
cd ../..
Produce simulated reflection cloud dataset
RCT is based on PointTransformerV3, 并使用了pointcept的代码实现。
If you find our work helpful, please consider citing the following BibTeX entry.