PycWB is a modular Python implementation of the coherent WaveBurst (cWB/cWB-2G) algorithms for gravitational-wave burst searches. The documentation can be found at docs.pycwb.org.
To run the bundled example from a source checkout:
conda create -n pycwb -c conda-forge python=3.13 pip nds2-client python-nds2-client lalsuite python-ligo-lw
conda activate pycwb
git clone https://git.ligo.org/yumeng.xu/pycwb.git
cd pycwb
python -m pip install .
pycwb --version
pycwb run examples/demo/user_parameters.yaml --work-dir my_first_search
pycwb progress --work-dir my_first_searchThe CLI example in examples/demo/ generates synthetic data containing a loud
injected burst and writes the normal pipeline outputs. It needs no detector-data account. The first run may
download a roughly 53 MiB cross-talk catalog and compile numerical kernels.
Use a fresh directory for each run. See Your First Search
for expected outputs and troubleshooting.
A normal python -m pip install pycwb selects a stable release; --pre allows
prereleases. Older releases have different dependencies and may require ROOT.
Match pycwb --version to the documentation version. Development documentation
can describe features not yet available on PyPI.
The current native path requires Python >=3.11 and does not require ROOT. Linux x86_64/Python 3.13 is the current CI environment. Other platform coverage and optional ROOT/PyCBC/XGBoost setup are described in the installation guide.
pycwb run user_parameters.yamlfrom pycwb.workflow.run import search
search("user_parameters.yaml")For production configuration templates, follow Setup Config Templates. Read Understanding Your Results, Reproducibility, and Validation Scope before interpreting an analysis.
- Support and bug reports, including an email route without a GitLab account.
- Contribution instructions and changes.
- Citation metadata and scientific citation guidance.
- Release and compatibility process.
Current native notebooks and optional integrations are listed in the example guide. The CLI synthetic example is the maintained beginner path.