Architecting MESA Simulation Grids with POSYDON
You can use POSYDON framework to run your own, customized MESA simulations of single and binary systems. Here we will go through POSYDON’s dedicated MESA configuration repository and infrastructure for one or even big grids of MESA runs.
Getting Started Tutorials
The POSYDON MESA simulation parameters are stored in POSYDON-MESA-INLISTS Github repository. To get the latest version of the POSYDON-MESA-INLISTS repository, you need to clone the POSYDON-MESA-INLISTS repository.
git clone https://github.com/POSYDON-code/POSYDON-MESA-INLISTS.git
cd POSYDON-MESA-INLISTS
This repository contains the MESA inlists used by POSYDON to run single and binary star simulations. At the time of writing, the POSYDON v2.0.0 code is compatible with MESA r11701 and the inlists are formatted as such. Additionally, multiple changes were made to the MESA source code to fix bugs, including reverse mass transfer in binaries (where the secondary star fills its Roche lobe).
Warning
Please contact us if you want to run MESA grids with this adapted version of MESA.
Each branch and commit in the POSYDON-MESA-INLISTS repository contains a specific set of MESA inlists associated to a POSYDON rerun.
To follow the next step of the tutorial, you will need to have a MESA version compatible with POSYDON installed on your machine. If you do not have MESA installed, please follow the instructions on the MESA website.
Warning
The POSYDON v2.0.0 code is compatible with MESA r11701 inlists. Support might not be available for the latest MacOS version.
I. Running your first MESA simulation using POSYDON
The following Jupyter Notebook guides you through our dedicated MESA submission API, showcasing how to simulate a HMS-HMS binary systems.
The first version of this tutorial showcases how to run the simulation using SLURM on HPC.
In case you want to run the simulation locally, you can follow the second version of this tutorial.
To gather more information about the MESA simulation submission API ini file, please refer to the MESA Grids API section of the documentation.
II. Running your first MESA grid simulation using POSYDON
The following tutorial will show you how to run a grid of MESA simulations using POSYDON. The grid will consist of 100 simulations of HMS-HMS binary systems at 0.1Zsun metallicity in the mass ratio slice 0.7 each with different primary masses and orbital periods. The simulations will be run on a HPC using SLURM.
Now that you have run your first grid, you can process, visualize and explore the results using the POSYDON post-processing pipeline API. TODO: link
III. Running single stars with POSYDON
This tutorials shows how to run a grid of single star simulations using POSYDON.
We show how to export EEPs and make a PSyGrid object in this tutorial TODO: link the advanced tutorial of step 2.
Advanced Tutorials
Creating and Customizing MESA Grids layer by layer
Step-by-step guide on how to leverage our MESA simulation submission API for creating and customizing MESA grids to suit your needs.
Dive deeper into advanced techniques for grid creation, by layering MESA and POSYDON default inlist parameters and customizing your inlists variables on top to your needs.
Warning
This feature is only partially implemented. Please contact us if you require more information.
Running a dynamically sampled grid of MESA simulations
Warning
This feature is experimental. Please contact us if you encounter any issues.
Provided a sparse, rectilinearly sampled MESA grid, POSYDON allows to dynamically sample the parameter space to enhance the coverage of the parameter space to improve classification and interpolation accuracy.
Support & Feedback
Encountered obstacles or have questions? Ensure you consult our [FAQ](link-to-FAQ) or drop by the [contact-information.rst](contact-information.rst) page.