The ABEL simulation framework is a particle-tracking framework for multi-element particle accelerators (such as plasma-accelerator linacs, colliders, experimental test facilities, etc.), implemented at varying levels of complexity, for fast investigations or optimizations. As a systems code, it can be used for physics simulations as well as generating (and optimizing for) cost estimates.
The project name is abel-framework, which contains the abel package with various sub-packages.
The project homepage is: https://github.com/abel-framework/ABEL/
ABEL needs Python v3.11, due to requirements from some dependencies.
Make and activate a new Python (version 3.11) virtual environment for ABEL:
python3.11 -m venv your_abel_venv and source your_abel_venv/bin/activate
If not already done, setup conda-forge:
conda config --add channels conda-forge and conda config --set channel_priority strict
Make and activate a new Python conda environment for ABEL:
conda create -p abel-2026 python=3.12 and conda activate /home/kyrsjo/CONDA/abel-2026
You may consider to then manually install some packages:
conda install numpy scipy numba matplotlib tqdm PyQt6 pyqtgraph toml dill
After preparing the Python environment, simply run:
pip install abel-framework
This will also install any missing dependencies in the right versions, including Wake-T, RF-Track, CLICopti, ax-platform, and impactx.
To remove ABEL from the python environment, run pip uninstall abel-framework.
To install a specific version from GitHub, e.g. for development, you first need to clone the repository to a local folder:
git clone https://github.com/abel-framework/ABEL
Note that if you intend to contribute (push) to ABEL, cloning with SSH instead of HTTPS is reccomended.
Once the repository has been cloned, in your target python environment, run
pip install -e path-to-ABEL where path-to-ABEL is the folder you have cloned ABEL to.
If you want to be able to modify ABEL without uninstalling and reinstalling, you can instead run
pip install -e path-to-abel
Changes to the files in the folder src\abel will be reflected in the installed package, as if you had put it into your $PYTHONPATH.
To remove ABEL, run pip uninstall abel-framework.
To use ABEL, you must configure it. This is done with the file .abelconfig.toml, which is automatically created in your home directory the first time you import ABEL.
Edit this file with your text editor to tell ABEL where to find tools such as ELEGANT, HIPACE++, and GUINEA-PIG, as well as configure it for your computing cluster, if needed.
Comments in the file explain how to edit it. The configuration file uses the TOML format, which is a simple text file similar to .ini, but more well defined.
Please do not edit the template file abelconfig.toml or CONFIG.py in the source code folder.
The loaded configuration of ABEL is printed to the terminal when ABEL starts, along with the name of the config file it has loaded.
Unit tests are implemented with pytest; to run the tests on an installed version of ABEL please run
pytest -v
from the root folder of ABEL.
It is also possible to run single tests by name, for example: pytest -v tests/test_init.py::testCore_init
The tests are stored in in the tests subdirectory, in files with names starting with test_.
In these files, the functions with names starting with test_ are represent one test; if it makes it to the end without any of the asserts triggering and all the expected exceptions happening, the test has PASSED.
The test functions are also annotated with @pytest.mark.MARKNAME.
Pytest is configured in the [tool.pytest.ini_options] section of pyproject.toml. This especially defines the "markers", which are named groups of tests that can be specified to run using pytest -v -m MARKNAME.
When the tests succeed, no output (except testfile::testfile PASSED) is printed. If a test fails, a traceback and the printouts of that test are printed. If many tests fail, this can be very verbose.
Please cite the following when referring to ABEL or using ABEL simulations for publications:
[1] J. B. B. Chen et al., ABEL: The Adaptable Beginning-to-End Linac simulation framework, Proceedings of IPAC 2025 (Taipei, Taiwan, 2025), pp. 1438-1441.
This work was supported by the European Research Council (project SPARTA, Grant No. 101116161) and the Research Council of Norway (Grant No. 313770 and 353317).