brutus

brutus#

Et tu, Brute?

“Brute-force” Bayesian inference for stellar photometry

brutus logo

Derive distances, reddenings, and stellar properties from photometry

v1.1.0

Installation

Get started quickly with pip install. Download required model grids and data.

Installation

Quick Start Guide

Learn the basic workflow in 5 minutes. Fit your first star with BruteForce.

Quick Start Guide

Glossary

Key terms and concepts including EEP, isochrones, and extinction.

Glossary

What’s New

See the latest changes in v1.1.0. Release notes and migration guides.

Changelog

Key Features#

Individual Star Fitting

Fit distances, reddenings, and stellar properties for individual stars using Bayesian inference with pre-computed model grids.

How do I fit stellar parameters?

  • Large sample (>100 stars) with standard filters: Use BruteForce + StarGrid

    Fastest approach. Pre-computed grids enable fitting thousands of stars quickly. See Quick Start Guide - “Fitting with BruteForce”.

  • Small sample or non-standard setup: Use StarEvolTrack

    More flexible but slower. Interpolates along stellar evolution tracks. See Quick Start Guide - “Individual Star Modeling”.

Cluster Analysis

Model stellar clusters with consistent ages, metallicities, and distances using isochrone-based population models.

How do I model a stellar cluster?

Use Isochrone + StellarPop with your preferred sampler or optimizer.

Fits shared age, metallicity, distance, and extinction for coeval populations. Handles cluster membership probabilities, binary stars, and outlier rejection. See Quick Start Guide.

3D Dust Mapping

Integrate with 3D dust maps (Bayestar) and model extinction along lines of sight for improved distance estimates.

How do I use dust maps?

  • Extinction priors: Use Bayestar to query 3D dust maps and constrain extinction in your fitting. See Prior Distributions.

  • Dust estimation: Derive extinction estimates from fitted stellar samples along lines of sight.

  • Line-of-sight modeling: Fit and visualize dust distributions using the los_clouds_* functions. See Quick Start Guide.

Synthetic Photometry

Generate synthetic SEDs for individual stars or full stellar populations using MIST isochrones and neural network models.

How do I generate synthetic photometry?

  • For individual stars: Use StarEvolTrack.get_seds()

  • For populations: Use StellarPop.get_seds()

Useful for simulations, testing, or understanding model predictions. See Stellar Models and Photometry.

Code Examples#

Install the latest stable release from PyPI:

pip install astro-brutus

Or install the development version (v1.1.0) directly from GitHub:

pip install git+https://github.com/joshspeagle/brutus.git

Note

Windows users: brutus requires WSL (Windows Subsystem for Linux) due to the healpy dependency. See Installation for details.

Download required data files:

from brutus.data import fetch_grids, fetch_isos, fetch_offsets

# Download stellar model grid (~750 MB)
grid_path = fetch_grids(grid='mist_v9')

# Download isochrone tables (~250 MB)
iso_path = fetch_isos(iso='MIST_1.2_vvcrit0.0')

# Download photometric offsets (recommended for fitting)
offset_path = fetch_offsets(grid='mist_v9')

Fit stellar parameters with BruteForce:

from brutus.data import load_models, load_offsets
from brutus.core import StarGrid
from brutus.analysis import BruteForce
from brutus.utils import inv_magnitude

# Load model grid with specific filters
filters = ['Gaia_G_MAW', 'Gaia_BP_MAWf', 'Gaia_RP_MAW',
           'PS_g', 'PS_r', 'PS_i', 'PS_z', 'PS_y',
           '2MASS_J', '2MASS_H', '2MASS_Ks']
models, labels, label_mask = load_models(
    'grid_mist_v9.h5', filters=filters
)
grid = StarGrid(models, labels)

# Load photometric offsets
offsets = load_offsets('offsets_mist_v9.txt', filters=filters)

# Convert magnitudes to linear flux densities
flux, flux_err = inv_magnitude(mag, mag_err)

# Fit photometry
fitter = BruteForce(grid)
results = fitter.fit(
    data=flux,              # (N, Nfilt) linear flux densities
    data_err=flux_err,      # (N, Nfilt) flux errors
    data_mask=flux_mask,    # (N, Nfilt) boolean mask
    data_labels=obj_ids,    # (N,) object identifiers
    phot_offsets=offsets,   # Photometric calibration
    parallax=plx,           # (N,) parallax in mas
    parallax_err=plx_err,   # (N,) parallax error
    data_coords=coords,     # (N, 2) Galactic (l, b) in degrees
    save_file='results.h5'
)

Generate synthetic photometry for stellar populations:

from brutus.core import Isochrone, StellarPop

# Create population generator with specific filters
filters = ['Gaia_G_MAW', 'PS_g', 'PS_r', 'PS_i', '2MASS_J']
iso = Isochrone()
pop = StellarPop(isochrone=iso, filters=filters)

# Generate photometry for a 1 Gyr solar-metallicity population
seds, params, params2 = pop.get_seds(
    feh=0.0,        # Solar metallicity
    loga=9.0,       # log10(age/yr) = 1 Gyr
    av=0.5,         # V-band extinction
    rv=3.3,         # Extinction law R(V)
    dist=2000.0     # Distance in parsecs
)