Scripts for "Stranger Things: A Grid-based Survey of Strange Modes in Post-Main Sequence Models"

This repository contains the Python scripts used to generate the models and data presented in "Stranger Things: A Grid-based Survey of Strange Modes in Post-Main Sequence Models" (Tarczay-Nehéz et al., 2026). The scripts automate the creation of MESA input directories for a grid of stellar masses and metallicities, including overshoot prescriptions and inlist configurations.

⚠️ Important compatibility note
These scripts have been tested and are known to work with:

• MESA version 23.05.1
• GYRE version 7.0

Other versions may lead to unexpected behavior or incompatibilities.

⚠️ Disclaimer
It is assumed that the user is already familiar with MESA, has it properly installed, and knows how to run it. Installation instructions, tutorials, and general usage guides are not part of this repository.

Contents

• generate_dirs/ – this includes the grid generating pys.
• include/ – this direcotry contains MESA src/ and make dirs for the current runs.
• exec/ - this directory contains the MESA executables (clean, rn, star, etc) for the script.
• example/ – sample run directory structure.

Requirements

• Python 3.8 or newer
• NumPy ≥ 1.18
• MESA version 23.05.1
• GYRE version 7.0

Usage

1. To recreate your grid of MESA runs, clone the repository:

git clone https://github.com/username/Strange-mode-cepheids.git
cd Strange-mode-cepheids

2. Run the scripts in one of two ways:

Option A — generate all grids at once (recommended):

python initialize_grid.py

This will run all generator scripts (initialize_convos_low_grid.py, initialize_convos_mid_grid.py, initialize_convos_high_grid.py, initialize_no_os_grid.py) and create the corresponding directories in the repository root.

Option B — generate grids individually:

cd generata_dirs
python initialize_convos_high_grid.py

... or the corresponding script (initialize_convos_mid_grid.py, initialize_convos_high_grid.py) depending on the grid you want to generate.

3. After successful execution, a set of run directories will be created under the chosen base_dir in the root directory (i.e., no_os, nad_convos_low, nad_convos_mid, nad_convos_high). Each run directory contains:

• inlist and inlist_project input files
• copied executables from exec/
• copied source/build directories from include/

👉 See example/run_nad_convos_mid_8.0MSUN_z0.0075/ for a reference of how a correctly generated run directory should look.

4. Run the models

You can either run all the generated models for all the four directories, or set it by hand.

Option A — Run all the models:

python run_grid.py

This will automatically run all the models parallel on 4 CPU cores.

Option B — Run only one model set:

To run only specific grids, use --base_dir:

python run_grid.py --base_dir no_os nad_convos_mid

⚠️ Important note on CPU usage:

You can control the number of parallel runs with --cores. For example:

python run_grid.py --cores 8

The driver will keep your cores busy by starting new runs as soon as others finish. Keep in mind that MESA itself often uses multiple threads internally (commonly 2 by default via OpenMP). Since threads share physical CPU cores, oversubscribing can slow everything down. As a rule of thumb, set --cores to at most half of your available physical cores unless you explicitly configure MESA to use only one thread per run (export OMP_NUM_THREADS=1).

Citation

In case you use these scripts, please cite the associated paper:

@article{TarczayNehez2026,
  author  = {Tarczay-Nehéz, D. and Coauthors},
  title   = {Stranger Things: A Grid-based Survey of Strange Modes in Post-Main Sequence Models},
  journal = {Astronomy & Astrophysics},
  year    = {2026},
  volume  = {XXX},
  pages   = {YYY-ZZZ},
  doi     = {10.1234/zenodo.XXXXX}
}

Additionally, you can cite the Zenodo DOI for this repository:
https://doi.org/10.5281/zenodo.1722995