Outputs¶
MCSED can produce a number of output ascii files and plots, which
summarize the fitted parameters, show their uncertainties, and compare
the fits to the input data. All these data will be stored in a directory
entitled output. The primary output file, which summarizes the
model parameter estimates for each object, is saved in an output file
specified by the -o argument on the command line (the default filename
is test.dat). All ancillary outputs are saved with filenames that
include the type of output, the field and object ID, and the star formation
history and dust attenuation laws used in the calculation. The user can select
which files to return and which files to skip using the output_dict
dictionary defined in config.py.
Summary File (parameters = True)¶
If the parameters keyword in the output dictionary is set to
True, a summary output file is generated. In this file, one row will
be returned for each object. The columns of this file include at a
minimum, the field, the galaxy ID, \(z\), and the fitted model parameters,
including the current log SFR (in \(M_\odot\) yr\(^{-1}\)), the
stellar mass (in \(M_\odot\)), the amount of dust attenuation as
parameterized by \(E(B-V)\), and variables associated with the
description of the SFR history, attenuation law, dust emissivity, and
stellar metallicity. Each of these parameter will be printed with its 5,
16, 50, 84, and 95 percentile measurements. (The percentiles that are returned
can be changed in config.py using the array param_percentiles.) The
output file from test mode will also return columns of truth values for
each parameter. The default is parameters = True.
Settings File (settings = True)¶
If the settings keyword in the output dictionary is set to True,
a file will be generated that lists all the user-defined fitting
assumptions, including the choice of star formation rate history,
attenuation law, dust emission law, and all associated parameters specified
in the config.py file or on the command line. This file will have the same
base name as the file described in Summary File (parameters = True), with an additional
extension .args. The default is settings = True.
Posterior Probability Distributions (fitposterior = True)¶
For each object, a file will be returned which contains the calculated
parameters and the corresponding log posterior probability distribution
for each MCMC chain. The number of chains depends on the number of
walkers and steps defined in config.py or on the command line using
the -nw and -ns arguments. The parameters listed are the same as
those in the main output file.
The default for this option is fitposterior = False.
Full 3D Posterior Probability Distributions (fullposterior = True)¶
A single file, titled fullposterior.hdf5, will contain multiple datasets
containing a 3D array containing each parameter estimate for each walker
at each step. Enabling this option is useful for detailed analyses of
convergence. This option utilize the h5py module to save data.
The default for this option is fullposterior = False.
Best Fit SEDs (bestfitspec = True)¶
For each object, a file containing the best fit SED will be returned,
with columns of wavelength and flux density. If the separate_stars_gas
keyword in config.py is set to True, the output file will contain
additional columns for the best fit stellar and nebular components, in
addition to the combined SED. The default wavelength
range is from rest-frame 500 Å to 350 \(\mu\)m. If dust emission
is not being fit (dust_em = False), the spectrum will extend to the
rest-frame wavelength specified by the wave_dust_em keyword in config.py.
The flux densities are given in units of \(\mu\)Jy. The default is
bestfitspec = True.
Observed and Modeled Flux Densities (fluxdensity = True)¶
For each object, a file containing the input and best-fit photometric
flux densities is returned. The columns are the name of the emission
line, its wavelength (in Å), the measured flux density, the error on
this flux density, and the best-fit flux density, all in
\(\mu\)Jy. The default for this option is fluxdensity = True.
Observed and Modeled Emission Line Fluxes (lineflux = True)¶
For each object, a file containing the observed and modeled emission
line fluxes will be returned. The list of lines are those defined by
emline_list_dict in config.py, and the columns returned are the
name of line, the observed monochromatic flux, the uncertainty in the
measurement, and the modeled flux. The default for this option is
lineflux = True. The file will only be saved if emission line
fluxes are provided in the input file and used in constraining the SED fit.
Observed and Modeled Absorption Indices (absindx = True)¶
For each object, a file containing the observed and modeled absorption
line indices will be returned. The list of features are those defined by
absorption_index_dict in config.py, and the columns returned are
the name of index, it observed value, the uncertainty in the
measurement, and the modeled value. The default for this option is
absindx = True. The file will only be saved if absorption line indices are provided in
the input file and used in constraining the SED fit.
Sample Plot (sample plot = True)¶
Plot of all realizations of the posterior distribution for all model
parameters. Each MCMC walker is represented by an individual curve and
the \(x\)-axis tracks the number of steps of each walker. The
default for this option is sample plot = False. The file format of the figure
can be changed via the image format keyword.
SSP template spectra (template spec = True)¶
Save a figure of the SSP spectra used in the fit. The spectra are
weighted by the amount of time between each SSP age gridpoint and the curves
are assigned colors progressing from blue to red as the age increases. The SSP
grid will only carry ages that are relevant for the calculation: for
example, if the input galaxy sample is at redshifts \(z > 3\), the
SSP grid will exclude ages older than \(\sim 2\) Gyr. The default
for this option is template spec = True. The file format of the figure
can be changed via the image format keyword.
Triangle Plots (triangle Plot = True)¶
For each object, a figure summarizing the results of the fits will be
returned. Each figure will contain of a number of different plots,
including illustrations showing 200 randomly-drawn MCMC solutions
for the fitted spectrum, the dust attenuation curve, and the galaxy’s star
formation rate history. Each figure will also contain contour plots illustrating
the co-variances of each parameter along with the parameters’ final
probability distribution functions. The user may select the image format
for these figures using the dictionary output_dict in config.py,
with the available options being
image format = .eps, .pdf, .pgf, .png, .ps, .raw, .rgba, .svg and
’.svgz’. By default, image format = .png.
The default for this option is triangle plot = True.