Configuration

Lazy.jl uses TOML parameter files with four sections. Generate an annotated example with lazy params > example.toml.

[io] – Input/Output

Parameter	Type	Default	Description
input_catalog	String	required	Path to input FITS catalog
output_file	String	required	Output file path (.fits, .h5, or .hdf5)
output_format	String	inferred	Output format: 'fits', 'hdf5', or 'both'. If omitted, inferred from output_file extension
missing_data_format	String	required	How missing data is encoded in the catalog: 'nan', 'zero', or a numeric value (e.g., '-99')
output_pz	Bool	false	Include P(z) distributions in output
output_templates	Bool	false	Include IGM-attenuated template SEDs in output (needed for SED reconstruction)

Rest-Frame Magnitudes

Parameter	Type	Default	Description
output_restframe_mags	Bool	false	Compute rest-frame absolute magnitudes (M_UV, M_U, M_V, M_J)
flux_units	String	–	Required if output_restframe_mags = true. Flux units of the input catalog: 'uJy', 'nJy', 'Jy', or 'cgs'
H0	Float	70.0	Hubble constant in km/s/Mpc (flat LCDM cosmology)
Om	Float	0.3	Matter density parameter (flat LCDM cosmology)

Forced Low-Redshift Fitting

Parameter	Type	Default	Description
output_forced_lowz	Bool	false	Perform a second fit restricted to low redshifts
forced_lowz_zmax	Float	–	Required if output_forced_lowz = true. Maximum redshift for the forced low-z fit

When enabled, the output includes z_best_lowz, chi2_lowz, delta_chi2 (= chi2_lowz - chi2_best), low-z P(z) quantiles, and low-z model photometry. This is useful for validating high-redshift candidates.

[fitting] – Scientific Parameters

Parameter	Type	Default	Description
template_set	String or Array	required	Template set name (e.g., 'sfhz') or list of custom template file paths
template_error	String	required	Template error model. Currently only 'template_error' is supported
template_error_scale	Float	required	Scale factor for the template error function
igm_model	String	required	IGM attenuation model. Currently only 'inoue14' is supported
nphot_min	Int	required	Minimum number of valid photometric points (SNR > 2) required to run the fit
sys_err	Float	required	Systematic error added in quadrature to photometric errors
z_min	Float	required	Minimum redshift for the fitting grid
z_max	Float	required	Maximum redshift for the fitting grid
z_step	Float	required	Redshift step size
template_cache	Bool	true	Cache template grids for reuse in subsequent runs

Spectroscopic Redshifts

Parameter	Type	Default	Description
use_zspec	Bool	false	Fix the fit at the spectroscopic redshift when available

When enabled, objects with a valid spectroscopic redshift (not NaN, not negative) are fit only at the nearest grid point to their z_spec. Objects without z_spec are fit normally across the full grid. Requires a zspec entry in the [translate] section.

CGM Damping Wing

Parameter	Type	Default	Description
add_cgm	Bool	true	Enable CGM Lyman-alpha damping wing absorption (Asada+2024)
cgm_A	Float	3.5918	Sigmoid amplitude for HI column density evolution
cgm_a	Float	1.8414	Sigmoid slope for HI column density evolution
cgm_c	Float	18.001	Sigmoid baseline for log10(N_HI)

The CGM model adds damping wing absorption redward of Lyman-alpha at z >= 6, applied multiplicatively on top of the IGM attenuation. The HI column density evolves with redshift as:

log10(N_HI) = A / (1 + exp(-a * (z - 6))) + c

The default parameters are from Asada et al. (2024) and generally do not need to be changed.

[runtime] – Processing Options

Parameter	Type	Default	Description
chunked_processing	Bool	false	Enable memory-efficient chunked processing
target_memory_gb	Float	0.5	Target memory usage per chunk in GB
preserve_work_file	Bool	false	Keep intermediate .work.h5 file after completion

See Advanced Usage for details on chunked processing and resume capability.

[translate] – Column Mapping

Maps filter names to flux/error columns in your input catalog.

[translate]
    f606w = {flux = 'f_f606w', error = 'e_f606w'}
    f814w = {flux = 'f_f814w', error = 'e_f814w'}
    f115w = {flux = 'f_f115w', error = 'e_f115w'}

The keys are filter names or nicknames (see Filters for the full list). The values must match column names in the input FITS catalog.

Spectroscopic Redshift Column

When use_zspec = true in [fitting], add a zspec entry:

[translate]
    zspec = 'z_spec'

This maps to the column containing spectroscopic redshifts. Objects with NaN, negative, or missing values are fit normally.