SSC-RAFT: Reconstruction Algorithms for Tomography

Authors

• Eduardo X. Miqueles, LNLS/CNPEM
• Paola Ferraz, LNLS/CNPEM

Contributors

• Larissa M. Moreno, LNLS/CNPEM
• João F. G. de Albuquerque Oliveira, LNLS/CNPEM
• Alan Zanoni Peixinho, LNLS/CNPEM
• Yuri Rossi Tonin, LNLS/CNPEM

Past contributors

• Otávio M. Paiano
• Gilberto Martinez Jr.
• Giovanni Baraldi
• Janito Vaqueiro Ferreira Filho
• Fernando S. Furusato
• Matheus F. Sarmento, LNLS/CNPEM
• Nikoley Koshev
• Elias Helou

Acknowledgements

We would like to acknowledge the Brazilian Ministry of Science, Technology, and Innovation MCTI for supporting this work through the Brazilian Center for Research in Energy and Materials (CNPEM). We want to thank Petrobras (Manhattan Project) - ID 2021/00018-3 for funding part of this project - and Mogno beamline from Sirius, for many fruitful discussion about the reconstruction strategies. We also thank FAPESP/CEPID 2013/07375-0 for funding Nikolay Koshev, and FAPESP 2016/16238-4 for funding Gilberto Martinez Jr.

Contact

Sirius Scientific Computing Team: [email protected]

Documentation

The package documentation can be found on the GCC website https://gcc.lnls.br/ssc/ssc-raft/index.html inside the CNEPM network. Also, the HTML documentation can be found in the source directory ./docs/build/index.html and can be opened with your preferred brownser.

Install

This package uses C, C++, CUDA and Python3. See bellow for full requirements.

The library sscRaft can be installed with form the source code or by pip/git if inside the CNPEM network.

Source code from Zenodo

The source code can be downloaded from zenodo website under the DOI:10.5281/zenodo.10988343.

After download the ssc-raft-v<version>.tar.gz with the source files, one can decompress by

    tar -xvf ssc-raft-v<version>.tar.gz

To compile the source files, enter the follwing command inside the folder

    make clean && make

PIP

---

Warning: This installation option is compiled for the LNLS Tepui cluster DGX machines.

---

If one is inside the CNPEM network, they can install the latest version of sscRaft directly from the pip server

    pip install sscRaft==<version> --index-url https://gitlab.cnpem.br/api/v4/projects/1978/packages/pypi/simple

Where <version> is the version number of the sscRaft

    pip install sscRaft==3.0.3 --index-url https://gitlab.cnpem.br/api/v4/projects/1978/packages/pypi/simple

GIT

---

Warning: For this installation option is necessary to be on the CNPEM network.

---

One can clone our gitlab repository and install the latest version by:

git clone https://gitlab.cnpem.br/GCC/ssc-raft.git 
cd ssc-raft
make clean && make

For a specific version, one can use:

    git clone --recursive https://gitlab.cnpem.br/GCC/ssc-raft.git --branch v<version> --single-branch
    cd ssc-raft 
    make clean && make

The <version> is the version of the sscRaft to be installed. Example, to install version 3.0.0

    git clone --recursive https://gitlab.cnpem.br/GCC/ssc-raft.git --branch v3.0.0 --single-branch
    cd ssc-raft 
    make clean && make

Memory

Be careful using GPU functions due to memory allocation.

Requirements

Before installation, you will need the following packages installed:

• CUDA >= 10.0.0
• C
• C++
• Python >= 3.8.0
• PIP
• libcurl4-openssl-dev

This package supports nvidia GPUs with capabilities 7.0 or superior and a compiler with support to c++17.

The following modules are used:

• CUBLAS
• CUFFT
• PTHREADS
• CMAKE>=3.10

The following Python3 modules are used:

• scikit-build>=0.17.0
• setuptools>=64.0.0
• numpy
• scikit-image >=0.19.3
• scipy
• matplotlib
• SharedArray
• h5py

Uninstall

To uninstall sscRaft use the command, independent of the instalation method,

    pip uninstall sscRaft