McFine: A tool for hyperfine spectral line fitting
Much of radio astronomy is using magic to turn line intensities into gas conditions. Magic because many molecules are complicated, and there are often multiple distinct components down the line of sight. McFine attempts to do this wizardry in a fully automated, Bayesian way so you can turn your spectra into science without too much hassle.
McFine uses an iterative approach, fitting and comparing increasingly complex models until it deems them sufficiently complicated. It does this through the Bayesian and Akaike Information Criterion, and specifically the change in these metrics between models. If given a data cube, it can also use the neighbouring information to attempt a better fit. For more details about the philosophy and maths of McFine, have a read of Williams & Watkins (2024), or see fitting philosophy. They’re both quite short!
Documentation
Developers
This package has been developed by:
Thomas Williams (Manchester)
Elizabeth Watkins (Manchester)
Citing McFine
If you use McFine in your work, please drop us a citation (Williams & Watkins, 2024), and provide a footnote to the package repository (https://github.com/thomaswilliamsastro/mcfine).