Generating test signals for gravitational-wave searches

One of the most important tools in a gravitational-wave astronomer’s toolbox is the ability to simulate signals: to inject a known signal into detector data, run a search algorithm over the data, and check that the search recovers what was put in. This process — called software injection — is fundamental to characterising the sensitivity of detection algorithms, estimating their systematic uncertainties, and validating pipelines before they are deployed on real data.

Minke provides a unified Python interface for generating the variety of signal morphologies needed for these tests. These range from simple parameterised burst waveforms (sine-Gaussians, Gaussians, white noise bursts) used to measure search sensitivity in a model-independent way, to physically motivated waveforms for specific astrophysical scenarios.

The package is designed to integrate with the gravitational-wave Python ecosystem, including gwpy for handling detector data and astropy for coordinate handling. It manages the generation of signals in a detector-agnostic way, then handles the projection of waveforms onto the detector network given the source sky position, and the creation of the frame files used as standard input by LIGO data analysis software.

Minke originated as a tool to support burst searches during the early LIGO observing runs, and has continued to evolve alongside changes in the wider software ecosystem. Ongoing development adds support for compact binary coalescence waveforms via modern waveform libraries, giving both the burst and parameter estimation communities a common framework for signal simulation.