Python scripts to reproduce MESA inlists for "Machine Learning Methods for Stellar Collisions. I. Predicting Outcomes of SPH Simulations" by González Prieto et al. 2026

These scripts were used to generate the MESA stellar models described in Section 2.2 of the paper. All models were computed using MESA version 24.08.1 with the MESA SDK x86_64-linux-23.7.3. inlist_1Msun_prems_to_tams illustrates the stellar evolution setup for stars with masses below 4 M☉ (the initial_massparameter can be adjusted to the desired value). inlist_8Msun_prems_to_tams provides the corresponding setup for stars above 8 M☉. inlist_6Msun_prems_to_tams demonstrates an intermediate case in which the overshooting parameters are interpolated across the 4–8 M☉ mass range. write_inlist_many.py generates MESA inlists consistent with POSYDON. It takes as input the stellar metallicity, the adopted solar abundance definition, and the stellar mass. make_inlists.sh runs write_inlist_many.py across all directories matching the pattern *Msun (where * denotes a numerical value) and copies the resulting inlist into each corresponding directory. hr.py generates Hertzsprung–Russell (HR) diagrams. get_mesa_profiles_from_sph.py extracts MESA profiles from simulation logs and copies them into a profiles/ directory for use in StarSmasher relaxation runs. inlist_template provides a template MESA inlist for collAIder stellar evolution models. mesa_job is an example batch submission script and must be adapted to the user’s local directory structure and computing environment. Files ending in _ages perform the same tasks as their corresponding non-_ages counterparts, but output results at specified stellar ages.