Fitting Options
fit_reg¶
Type: [tuple | list] of length (2,)
Default: (4400,5500)
Description: The minimum and maximum desired fitting wavelength in angstroms.
fit_area¶
Type: dict
Default: {}
Description: Defines the area to be fit for data cubes. See examples/muse_examples.py for usage.
mask_bad_pix¶
Type: bool
Default: False
Description: Mask pixels which the specified instrument has flagged as bad due to sky line subtraction or cosmic rays.
mask_emline¶
Type: bool
Default: False
Description: Mask any significant absorption and emission features relative to the continuum. This uses an automated iterative moving median filter of various sizes to detect significant flux differences between window sizes. Good for continuum fitting but tends to over mask lots of features near the edges of the spectrum.
fit_stat¶
Type: str
Default: "ML"
Description: The fit statistic used for the likelihood. Options:
"ML"for standard maximum likelihood (pixels weighted by noise with no noise scaling)."OLS"for ordinary least-squares fitting (all pixels weighted by same amount).
n_basinhop:¶
Type: int
Default: 25
Description: Number of successive niter_success times the basinhopping algorithm needs to achieve a solution. The fit becomes much better with more success times, however this can increase the time to a solution significantly. Recommended 5-10.
reweighting¶
Type: bool
Default: True
Description: If true, BADASS will reweight the noise vector to achieve a reduced chi-squared ~ 1. This is done after the initial basinhopping fit, and applied to any bootstrapped uncertainties and MCMC fitting performed afterward. This does not affect the chi-squared ratio metric used in line and configuration testing, but does effect the amplitude-over-noise and SNR calculations in BADASS.
test_lines¶
Type: bool
Default: False
Description: Performs tests for lines. Options are specified in test_options.
max_like_niter¶
Type: int
Default: 10
Description: Number of bootstrapping iterations to perform after the initial basinhopping fit. This is a means to obtain uncertainties on parameters without performing MCMC fitting, however, do not produce as robust uncertainties as MCMC.
output_pars¶
Type: bool
Default: False
Convenience feature that prints out all free parameters.
cosmology¶
Type: dict
Default: {"H0":70.0, "Om0": 0.30})
Description: The flat Lambda-CDM cosmology assumed for calculating luminosities from fluxes.