Modelling a gravity-mode pulsator

The neural network for Computing Pulsation Periods and Photospheric Observables (C-3PO) is trained on a grid of MESA stellar evolution models and GYRE pulsation models, covering the entire γ Doradus instability strip. C-3PO is integrated into a forward modelling scheme designed by Mombarg et al. (2021) to estimate the stellar mass and age from the observed prograde dipole gravity-mode pulsations, and optionally also the luminosity, effective temperature, and surface gravity. The C-3PO network can be downloaded here.

Prerequisites

To run C-3PO, you need the keras-enviroment.yml and requirements.txt file. Then, run the following commands (make sure you have Conda installed first):

conda env create -f keras-environment.yml
conda activate keras-env
pip install -r ./requirements.txt

You can then simply run Model_star.py once you have set all the correct paths in this file. The neural network modules are relatively large. If your system cannot handle this, you can try to load only a few of the C-3PO_Zext_npg15-91_P_*.h5 modules.

An online tutorial on asteroseismic modelling with C-3PO that I gave at the TASC6/KASC13 conference can be found here.