jaxglitches

Differentiable LISA glitch waveforms, time-delay interferometry (TDI) responses, and parameter-estimation utilities built with JAX. Importing the package enables float64 arithmetic.

The package models a single-exponential test-mass glitch on link 12, with TDI generations 1 and 2, AET or XYZ channels, and equal or frozen unequal arms. It also provides Gaussian likelihoods, Fisher matrices, priors, and synthetic catalogues resampled from LISA Pathfinder glitch populations.

Start with the installation and waveform example, check units and response conventions, try the example notebook, or browse the API reference.

The core dependencies are JAX, NumPy, lisaglitch, and lisaorbits. Noise models used in the example notebooks live in notebooks/noise.py and are not part of the installed package. Galactic-binary waveforms, wavelet transforms, and simulation tools are optional extras.

Explore the research notebook gallery for longer validation and approximation studies with saved outputs.

References

• Source code and notebooks
• LISA glitch simulation package
• Baghi et al. (2022), LISA Pathfinder glitch populations

Logo assets

Download the wordmark (SVG), transparent waveform mark (PNG), or square icon (PNG) for use elsewhere.