Some lightweight tools to grab data from the EXFOR database using the x4i3 library, and organize it for visualization and use in model calibration and uncertainty quantification.
Currently, exfor_tools supports most reactions in EXFOR, but only a small subset of the observables/quantities. Feel free to contribute! If it doesn't meet your needs check out the project it's built on, which is far more complete: x4i3.
pip install exfor-tools
Package hosted at pypi.org/project/exfor-tools/. Otherwise, for development, simply clone the repo and install locally:
git clone git@github.com:beykyle/exfor_tools.git --recurse-submodules
pip instal exfor_tools -e
You can run the notebooks in the examples/ directory to see how to use the package.
To run the notebooks, some additional dependencies are required:
pip install -r examples/requirements.txt
There are two directories in examples/: examples/examples_2025_release/ and examples/examples_2023_release/. The former contains notebooks with stored outputs valid for the latest EXFOR release, while the latter contains notebooks with stored outputs valid for the 2023-04-29 release. You can run either one, the only difference is the stored outputs (which are used for testing).
The examples include:
- examples/examples_2025_release/introductory_tutorial.ipynb
- examples/examples_2025_release/data_curation_tutorial.ipynb
These demonstrate how to query for and parse exfor entries, and curate and plot data sets. In the first one, you will produce this figure:
The tests and the examples are one and the same. To run the tests, first install the dependencies for the notebooks:
pip install -r examples/requirements.txt
Then, to test that the notebooks run, use:
pytest --nbmake examples/examples_2025_release/
To test that they produce the expected results, use:
pytest --nbval-lax examples/examples_2025_release/
Note that there may be some difference in your installation, e.g. if you're using a different version of the EXFOR database, so the expected results may not be exactly the same as those in the tutorials.
By default, x4i3 ships with the 2023-04-29 EXFOR release. There are a set of notebooks with stored outputs valid for that release in examples/examples_2023_release/. These are used in the github actions. If you haven't updated to a more recent release but you would like to run the tests, then simply run:
pytest --nbval-lax examples/examples_2023_release/
First, download your desired version <exfor-YYYY.zip> from here: https://nds.iaea.org/nrdc/exfor-master/list.html. The latest is recomended. Then:
bash update_database.sh </path/to/exfor-XXXX.zip> --db-dir </path/where/db/should/go/>This will extract and process the data to </path/where/db/should/go/unpack_exfor-YYYY/X4-YYYY-12-31>, setting the environment variable $X43I_DATAPATH accordingly. x4i3 uses this environment variable to find the database on import, so you should add this to your environment setup. If you use bash, this will look something like this:
echo export X43I_DATAPATH=$X43I_DATAPATH >> ~/.bashrcThis functionality for modifying the database used by x4i3 is provided in x4i3_tools, which is included as a submodule.