Welcome to cogwheel’s documentation!

cogwheel is a code for parameter estimation of gravitational wave sources. It implements a convenient system of coordinates for sampling, a “folding” algorithm to reduce the multimodality of posteriors, and the relative binning algorithm for fast likelihood evaluation (generalized to waveforms with higher modes). It supports likelihood marginalization over distance, as well as over all extrinsic parameters describing a merger. It interfaces with third-party routines for downloading public data (GWOSC, GWpy), generating waveforms (lalsuite) and sampling distributions (PyMultiNest, dynesty, zeus, nautilus). The source code is available on GitHub.

Contents:

• Installation
  • Via conda
  • From source
• Crash course
• Overview
  • Sampler
  • Posterior
  • Prior
  • Likelihood
  • Event data
  • Waveform generator
• Full API
  • cogwheel
• Acknowledgment

References

• Coordinate system and folding algorithm: https://arxiv.org/abs/2207.03508 (poster (PDF))

• Marginalization over extrinsic parameters for quadrupolar, aligned-spin signals: https://arxiv.org/abs/2210.16278

• Marginalization over extrinsic parameters for signals with precession and higher modes: https://arxiv.org/abs/2404.02435

• IMRPhenomXODE waveform approximant: https://arxiv.org/abs/2306.08774