speedyfit-wrapper

A Python wrapper around SpeedyFit for fitting binary SEDs with an infrared disc component. The pipeline iteratively fits a blackbody to the IR residuals of a binary SED fit, subtracts the disc contribution, and re-fits the binary model until the IR residuals converge to zero.

Overview

SpeedyFit fits stellar SEDs using Bayesian inference and MCMC. This wrapper extends that to a three-component system (primary + companion + disc) by:

1. Running SpeedyFit on the binary model (primary + companion)
2. Converting IR mag residuals to flux excess
3. Fitting a blackbody disc model to the IR flux excess
4. Subtracting the disc flux from the IR bands and re-running SpeedyFit
5. Repeating until convergence

Requirements

• Python 3.8
• SpeedyFit 0.2.4

Package	Version
astropy	5.2.2
astroquery	0.4.7
corner	2.2.1
emcee	3.1.6
h5py	3.11.0
matplotlib	3.7.5
numpy	1.24.4
pandas	2.0.3
scipy	1.10.1
pyaml	25.1.0
tqdm	4.67.1

Install into a conda environment:

conda create -n sed_fitting python=3.8
conda activate sed_fitting
pip install speedyfit==0.2.4 astropy astroquery numpy scipy matplotlib pandas emcee corner h5py tqdm pyaml

File Structure

File	Description
main.py	pipeline script
SF_phot.py	Reads SpeedyFit observation and model output files
third_comp_model.py	Planck function, mag-to-flux conversion, blackbody disc model
third_comp_params.py	Initial disc parameter estimates via Wien's law and Stefan-Boltzmann
plotting.py	SED and disc residual plots
logger.py	Rotating file + console logger

Usage

Edit the inputs block at the top of main.py:

STAR_NAME    = "your_target"      # used for output file names
OBS_FILE     = "phot.csv"         # SpeedyFit photometry file
SETUP_YAML   = "setup.yaml"       # SpeedyFit binary setup file
MODEL_FILE   = "model_out.csv"    # SpeedyFit model output file
DISTANCE_PC  = 500.0              # Gaia distance in pc
IR_THRESHOLD = 10000.0            # wavelength (AA) above which is treated as IR
MAX_ITER     = 20                 # maximum iterations
CONV_RMS     = 0.05               # convergence threshold (mag)

Then run:

python main.py

Outputs

• logs/<STAR_NAME>.log — pipeline log with disc parameters per iteration
• <STAR_NAME>_sed.pdf/png — full binary SED plot
• <STAR_NAME>_disc_sed.pdf/png — disc blackbody fit to IR residuals

Module Reference

SF_phot.unpack_sf

Reads SpeedyFit obs and model files.

sf = unpack_sf(obs_file="phot.csv", model_file="model.csv")
obs_wave, obs_flux, obs_err, obs_band = sf.observations()
mod_wave, mod_flux = sf.model_fluxes()

third_comp_model.mag_to_residuals

Converts SpeedyFit mag residuals to flux excess with propagated errors.

conv = mag_to_residuals(res_mag, res_err, mod_flux)
delta_flux, sigma_flux = conv.mag_conversion()

third_comp_params.disc_params

Estimates initial disc temperature and radius from the IR flux excess.

dp = disc_params(ir_wave, ir_flux_exc, ir_flux_exc_err, distance_pc)
T_disc = dp.wien_temp()
R_disc, R_err = dp.r_disc()

third_comp_model.bb_model

Evaluates the observed blackbody disc flux at given wavelengths.

flux = bb_model(wavelength, T_disc, R_disc, distance_pc).blackbody_flux()