fastar-astro 0.1.2

Stellar population synthesis tools

FASTAR

FASTAR is a fully differentiable, evolutionary stellar population synthesis code. It generates single stellar population (SSP) model predictions for both spectroscopic (from ~3,500 Å to ~7,400 Å, at a 2.51 Å FWHM resolution) and spectral energy distribution (SED, from ~2,000 Å to ~11,500 Å) measurements.

Unlike traditional SSP models, FASTAR produces continuous predictions for any age, metallicity and initial mass function (IMF) value. It also supports SSP model uncertainty estimation through Monte Carlo sampling.

Critically, FASTAR can also generate on the fly semi-resolved SSP models assuming the IMF is not fully sampled, i.e., when the number of stars per resolution element is not large enough.

---

Ingredients

• MILES — empirical stellar library
• BOSZ — theoretical stellar library
• Worthey & Lee (2011) — bolometric correction tables
• BaSTI-IAC — isochrones

---

Features

• NN-based predictions of stellar spectra.
• SSP synthesis for photometric and spectroscopic data.
• Integrated and semi-resolved SSP predictions.

---

Installation

git clone https://github.com/inavarro/fastar.git
cd fastar/
pip install .

This will install it in the current environment, but it is recommended to use a virtual environment for the installation (using venv, conda, uv or the tool of your choosing).

Development installation

If you wish to modify the fastar code, we recommend to use the provided Makefile to setup your environment. It uses uv to setup a virtual environment and it should be transparent to the developer.

git clone https://github.com/inavarro/fastar.git
cd fastar/
make install-dev
# ... do your development and test them
make tests
# ... add your changes and do a commit
# git add ...
# git commit ...
# This will trigger some automatic checks (format, linting,...) that you must
# address before committing (some of them are fixed automatically).
git push

Changes are welcome! But remember that it is always safer to work on your local fork and only when your development is finished you can proceed to do a PR.

---

Quick Start

from fastar import IntegratedSSPSynthesizer
from fastar.imf import kroupa

ssp = IntegratedSSPSynthesizer(imf_function=kroupa)
flux = ssp.synthesize(age=10.0, met=0.0)
wave = ssp.wave