The setup file

All details regarding the setup for the SED fit are provided to Speedyfit using a setup file in yaml format. Yaml was chosen because it is easy to read by a human and by a computer. The default setup file with all its options are given below:

# photometry file with index to the columns containing the photbands, observations and errors
objectname: Vega
photometryfile: Vega.phot
photband_index: band
obs_index: flux
err_index: eflux
photband_include: ['GAIA', 'APASS', '2MASS']
photband_exclude: ['GALEX']
# parameters to fit and the limits on them in same order as parameters
pnames: [teff, logg, rad, ebv]
limits:
- [3500, 10000]
- [4.31, 4.31]
- [0.01, 2.5]
- [0, 0.10]
# constraints on distance, mass ratio or any of the fitted parameters when known
constraints:
    parallax: [130.23, 0.36]
# added constraints on derived properties as mass, luminosity, luminosity ratio
derived_limits:
    mass: [1.5, 2.0]
# path to the model grids with integrated photometry
grids:
- kurucz
# setup for the MCMC algorithm
nwalkers: 100    # total number of walkers
nsteps: 1000     # steps taken by each walker (not including burn-in)
nrelax: 250      # burn-in steps taken by each walker
a: 10            # relative size of the steps taken
# set the percentiles for the error determination
percentiles: [16, 50, 84] # 16 - 84 corresponds to 1 sigma
# output options
resultfile: Vega_results.csv   # filepath to write results
plot1:
 type: sed_fit
 result: pc
 path: Vega_sed.png
plot2:
 type: distribution
 show_best: true
 path: Vega_distribution.png
 parameters: ['teff', 'rad', 'L', 'ebv', 'd', 'mass']

Photometry setup

objectname (str)

Name of the system being fitted. Can be whatever you want, and doesn’t need to be resolvable by Simbad. It is used for the figures and can help in recognizing what model this file belongs too, but doesn’t serve any other purpose.

photometryfile (str)

The path to the file where the photometry of your object is stored.

photband_index (str or int)

The name or index of the column in the photometry file that contains the names of the photometric bands. See The photometry file for more details.

obs_index (str or int)

The name or index of the column in the photometry file that contains the flux. See The photometry file for more details.

err_index (str or int)

The name or index of the column in the photometry file that contains the error on the flux. See The photometry file for more details.

photband_include (list)

A list of all the photometric bands to include in the fit. Any band not mentioned in this list is excluded. Bands are named as: SYSTEM.BAND (e.g. GAIA2.G for the G band from Gaia DR2). When the system is used (e.g. GAIA2) all bands from that system are included. If a specific band (e.g. GAIA2.G) is used, only that band is included. When both photband_exclude and photband_include are included in the setup file, photband_include is processed first, after which bands to be excluded are removed.

photband_exclude (list)

A list of which photometric bands to exclude from the fit. All bands mentioned here will be excluded. Bands are named as: SYSTEM.BAND (e.g. GAIA2.G for the G band from Gaia DR2). When the system is used (e.g. GAIA2) all bands from that system are excluded. If a specific band (e.g. GAIA2.G) is mentioned, only that band is excluded. When both photband_exclude and photband_include are included in the setup file, photband_include is processed first, after which bands to be excluded are removed.

Fit and constraints setup

pnames (list)

A list of the parameters to include in the fit. If you are performing a single fit, this will be:

[teff, logg, rad, ebv]

If you are performing a binary fit this will be:

[teff, logg, rad, teff2, logg2, rad2, ebv]

Note that for a binary fit, the reddening parameter (ebv) will only be included once. Both components in the fit need to have the same reddening.

limits (list of lists)

The limits on the parameters included in the fit. Limits have to be given in the same order as the parameters are given in pnames.

For example:

pnames = [teff, logg]

limits = [(3500, 10000), (3.5, 4.5)]

means that effective temperature is limited between 3500 and 10000 K, and logg is limited between 3.5 and 4.5 dex.

constraints (dict)

A dictionary containing all the constraints that you want to apply in the fit. These are treated as priors in the Bayesian fit. You can apply a prior on the distance/parallax, any of the parameters you are fitting (teff, logg, rad, ebv), any of the derived parameters (mass and luminosity) and in case of a binary fit also the mass ratio.

A constraint is not the same as a limit. A constraint consists of a value and an error. You can provide a double sided error as well. e.g.

parallax: [130.23, 0.36] -> parallax = 130.23 +- 0.36

teff: [5630, 150, 270] -> teff = 5630 -150 +270

Note: there constraints can be valuable if for example the effective temperature and surface gravity are known from a spectroscopic fit, and you want to derive the radius from the SED and the parallax while correctly propagating the errors.

derived_limits (dict)

A dictionary containing all the limits that you want to apply to derived parameters. These are also treated as priors, but as limiting priors were the prior is uniform between the lower and upper limit. You can apply limits on all derived parameters (mass and luminosity), on the distance and in case of a binary fit also on the mass ratio (q). e.g:

q: [0.5, 0.9]

means that the mass ratio is limited between 0.5 and 0.9

grids (list)

List of the model atmosphere grids used in the fit. For a single star fit only one grid name should be provided, for a binary fit, two grid names should be provided. You can provide the name of a grid as recognized by speedyfit, or the direct path to the grid. Recognized names are:

• kurucz
• munari
• tmap
• koester
• blackbody

For more info on the different available grids see Model atmosphere grids.

percentiles (list)

The percentiles to use when calculating the error on the fitted parameters. By default the percentiles are set to:

[16, 50, 85]

The best fitting value is the median of all values, and the lower and upper error given by speedyfit will then correspond to 1 sigma. This means that 68% of all steps taken by the walkers are within the error given by Speedyfit. See the 68–95–99.7 rule for more information on this. Quick summary:

1 sigma: [16, 50, 85]

2 sigma: [2.5, 50, 97.5]

3 sigma: [0.15, 50, 99.85]

If you don’t do any burn-in (via option nrelax), setting the percentiles to correspond with for example 3 sigma will result in very large unrealistic errors.

MCMC setup

The MCMC chain is implemented using the EMCEE package. See emcee for more information on the parameters governing the MCMC fitting process.

nwalkers (int)

Number of walkers used in the MCMC chain.

nsteps (int)

Number of steps that each walker takes in the MCMC chain.

nrelax (int)

Number of burn in steps that each walker takes. Walkers will start of at random locations in the parameter space and will slowly converge to an optimal solution. The nrelax steps are the first number of steps that will not be taken into account when determining the errors on the parameters from the walker chains.

a (int)

Relative step size taken by each walker.

Output and figures

resultfile (str)

The name of the file to store the results of the fit. This file will contain the best fitting value and error of each fitted and derived parameters in csv format.

plot<n> (dict)

In the result file you can also define what figures you want speedyfit to make. There are 3 types of figures that speedyfit can make:

• sed_fit
• constraint
• distribution

In the setup file you can define up to 10 different configurations of those 3 figures. All figures will follow the same structure:

plot1: # number of the plot, up to 10
    type: <plot type> # which type of plot you want
    path: <figurename>.png # filename in which to store the figure
    parameter: value # any other parameters to configure this plot

For more information on available plots and their configuration see Making figures