Install & Run¶
The source installation is the recommended workflow for most bayRing
analyses because catalogue paths, fits and waveform-model experiments often
need local edits.
Developer Installation¶
Clone the bayRing repository and
expose its location through BAYRING_PREFIX:
git clone https://github.com/GCArullo/bayRing.git
cd bayRing
export BAYRING_PREFIX=$PWD
For a permanent setup, add the export line to the shell startup file used on
your machine, for example ~/.bashrc or ~/.zshrc.
Install Python dependencies and the package:
pip install -r requirements.txt
pip install -e .
The editable install is useful for development because changes to files under
bayRing/ are picked up without reinstalling the package.
Pip Installation¶
Tagged releases can be installed from PyPI:
pip install bayRing
The PyPI release can be behind the source repository. Use it only when the tracked release contains the model and catalogue features required by the analysis.
pyRing Dependency¶
bayRing uses pyRing waveform classes and utilities. The README
recommends installing the hm_nc_fit branch of pyRing for source-based
development; see the pyRing installation documentation.
First Run¶
From the repository root, run the simplest tracked example:
bayRing --config-file config_files/config_SXS_0305_Kerr_220_quick.ini
This quick example fits the SXS 0305 waveform with a single Kerr
(2,2,0) QNM, constant NR error, and deliberately small sampler settings:
[Model]
template = Kerr
QNM-modes = 220
[Inference]
nlive = 32
maxmcmc = 32
t-start = 30.0
t-end = 80.0
Use this run to verify installation and output structure. It is not a production sampler configuration.
Running A Config File¶
Every run uses the same syntax:
bayRing --config-file path/to/config.ini
Missing options use defaults.
Run Types¶
run-type controls how much work is performed:
Value |
Behaviour |
|---|---|
|
Load NR data, run inference, save results and make diagnostics. |
|
Re-read previous results from |
|
Load the NR waveform and model object, make NR-only plots, then exit. |