MAGIX: Generic tool for fitting models to astrophysical data

Bernst, Irina

MAGIX (Modelling and Analysis Generic Interface for eXternal numerical codes) is a model optimizer developed under the framework of the CATS (Coherent set of Astrophysical Tools for Spectroscopy) project, funded by the European ASTRONET. The goal is to use existing astrophysical models, and optimize their parameters to fit a given astrophysical data set. A number of algorithms can be used to explore the parameter space (i.e. Levenberg-Marquardt, simulated annealing, nested sampling, the Bees algorithm), to find the best-fitting values of the parameters to be optimized, and the corresponding confidence intervals. More algorithms are to be included and the whole system of MAGIX is being parallelized, while a heuristics module is under way, which will assist on choosing the best combination of algorithms based on user-defined priorities.

A GUI based frontend is being developed to make the registration of new models and creation/editing of new instances of initial conditions user-friendly. MAGIX complies with the data structures and reduction tools of ALMA (Atacama Large Millimeter Array), and is developed to be a tool accompanying observations assembled with the ALMA interferometer, but can be used on any, even non-astronomical, data sets. So far, MAGIX has been successfully used to fit Herschel/HIFI data from line surveys.

Return to poster list