MEModel

class measure_extinction.model.MEModel(modinfo=None, obsdata=None, logf=False)

Bases: object

Model object for the measure_extinction fitting. Includes all parameters and useful functions. Inspired by astropy modeling.

Initialize the object, optionally using the min/max of the input model info to set the value and bounds on the stellar parameters

Parameters:

modinfoModelData object

all the information about the model spectra

obsdataStarData object

observed data for a reddened star

Attributes Summary

Av
B3
C2
C4
Rv
exclude_regions
fore_Av
fore_Rv
fore_sampling
g23_all_ext
gamma
lnp_bignum
logHI_MW
logHI_exgal
logTeff
logZ
logg
norm
paramnames
rng
vel_MW
vel_exgal
velocity
vturb
windalpha
windamp
xo

Methods Summary

add_exclude_region(exreg)	Add an exclude region to the list of such regions.
check_param_limits()	Check the parameters are within the parameter bounds
dust_extinguished_sed(moddata, sed)	Dust extinguished sed given the extinction parameters
fit_minimizer(obsdata, modinfo[, maxiter])	Run a minimizer (formally an optimizer) to find the best fit parameters by finding the minimum chisqr solution.
fit_sampler(obsdata, modinfo[, nsteps, ...])	Run a samplier (specifically emcee) to find the detailed parameters including uncertainties.
fit_to_parameters(fit_params[, uncs])	Set the parameter values based on a vector of the non-fixed values.
fit_weights(obsdata)	Compute the weight to be used for fitting.
get_nonfixed_paramnames()	Get the non-fixed parameter names.
hi_abs_sed(moddata, sed)	HI abs sed given the HI columns
lnlike(obsdata, modeldata)	Compute the natural log of the likelihood that the data fits the model.
lnprior()	Compute the natural log of the priors.
lnprob(obsdata, modeldata)	Compute the natural log of the probability
parameters()	Give all the parameters values in a vector (fixed or not).
parameters_to_fit()	Give the non-fixed parameters values in a vector.
plot(obsdata, modinfo[, resid_range, ...])	Standard plot showing the data and best fit.
plot_sampler_chains(sampler)	Plot the samplier chains.
plot_sampler_corner(flat_samples)	Plot the standard corner plot.
pprint_parameters()	Print the parameters with names and values
save_parameters([filename])	Save the parameters and uncertainties to a table.
set_initial_norm(obsdata, modeldata)	Set the initial normalization that puts the current model at the average level of the observed data.
stellar_sed(moddata)	Compute the stellar SED from the model parameters.

Attributes Documentation

Av = <measure_extinction.model.MEParameter object>

B3 = <measure_extinction.model.MEParameter object>

C2 = <measure_extinction.model.MEParameter object>

C4 = <measure_extinction.model.MEParameter object>

Rv = <measure_extinction.model.MEParameter object>

exclude_regions = <Quantity [[ 8.18,  8.28],            [ 8.7 , 10.  ],            [ 3.55,  3.6 ],            [ 3.8 ,  3.9 ],            [ 4.15,  4.3 ],            [ 6.4 ,  6.6 ],            [ 7.1 ,  7.3 ],            [ 7.45,  7.55],            [ 7.65,  7.75],            [ 7.9 ,  7.95],            [ 8.05,  8.1 ]] 1 / micron>

fore_Av = <measure_extinction.model.MEParameter object>

fore_Rv = <measure_extinction.model.MEParameter object>

fore_sampling = False

g23_all_ext = False

gamma = <measure_extinction.model.MEParameter object>

lnp_bignum = -inf

logHI_MW = <measure_extinction.model.MEParameter object>

logHI_exgal = <measure_extinction.model.MEParameter object>

logTeff = <measure_extinction.model.MEParameter object>

logZ = <measure_extinction.model.MEParameter object>

logg = <measure_extinction.model.MEParameter object>

norm = <measure_extinction.model.MEParameter object>

paramnames = ['logTeff', 'logg', 'logZ', 'vturb', 'velocity', 'windamp', 'windalpha', 'Av', 'Rv', 'C2', 'B3', 'C4', 'xo', 'gamma', 'vel_MW', 'logHI_MW', 'fore_Av', 'fore_Rv', 'vel_exgal', 'logHI_exgal', 'norm']

rng = Generator(PCG64) at 0x72499D65AEA0

vel_MW = <measure_extinction.model.MEParameter object>

vel_exgal = <measure_extinction.model.MEParameter object>

velocity = <measure_extinction.model.MEParameter object>

vturb = <measure_extinction.model.MEParameter object>

windalpha = <measure_extinction.model.MEParameter object>

windamp = <measure_extinction.model.MEParameter object>

xo = <measure_extinction.model.MEParameter object>

Methods Documentation

add_exclude_region(exreg)

Add an exclude region to the list of such regions.

Parameters:

exreglist

2 element list with min/max given in 1/micron

check_param_limits()

Check the parameters are within the parameter bounds

dust_extinguished_sed(moddata, sed)

Dust extinguished sed given the extinction parameters

Parameters:

moddataModelData object

all the information about the model spectra

seddict

fluxes for each spectral piece

Returns:

extinguished seddict

SED with {‘bands’: band_sed, ‘spec’: spec_sed, …}

fit_minimizer(obsdata, modinfo, maxiter=1000)

Run a minimizer (formally an optimizer) to find the best fit parameters by finding the minimum chisqr solution.

Parameters:

obsdataStarData object

observed data for a reddened star

moddataModelData object

all the information about the model spectra

maxiterint

maximum number of iterations for the minimizer [default=1000]

Returns:

fitmod, resultlist

fitmod is a MEModel with the best fit parameters result is the scipy minimizer output

fit_sampler(obsdata, modinfo, nsteps=1000, burnfrac=0.5, initfrac=0.01, save_samples=None, multiproc=False, resume=False)

Run a samplier (specifically emcee) to find the detailed parameters including uncertainties.

Parameters:

obsdataStarData object

observed data for a reddened star

moddataModelData object

all the information about the model spectra

nstepsint

number of steps for the samplier chains [default=1000]

burnfracfloat

fraction of nsteps to discard as the burn in [default=0.1]

initfracfloat

fraction for walker ball around initial position

save_samplesfilename

name of hd5 file to save the MCMC samples

multiprocboolean

set to run the emcee in parallel (does not speed up things much) [default=False]

resumeboolean

resume from previous run (requires save_samples to be an existing emcee hd5 save file)

Returns:

fitmod, resultlist

fitmod is a MEModel with the best fit parameters result is the scipy minimizer output

fit_to_parameters(fit_params, uncs=None)

Set the parameter values based on a vector of the non-fixed values. Needed for most fitters/samplers.

Parameters:

fit_paramsnp array

non-fixed parameters values

fit_weights(obsdata)

Compute the weight to be used for fitting. Observed data for the base weights 1/unc (expected by fitters). Weights in regions known to have data issues or that the models do not include are set to zero (e.g., stellar wind lines, geocoronal Ly-alpha)

Parameters:

obsdataStarData object

observed data for a reddened star

get_nonfixed_paramnames()

Get the non-fixed parameter names. Useful for plotting.

hi_abs_sed(moddata, sed)

HI abs sed given the HI columns

Parameters:

moddataModelData object

all the information about the model spectra

seddict

fluxes for each spectral piece

Returns:

hi absorbed seddict

SED with {‘bands’: band_sed, ‘spec’: spec_sed, …}

lnlike(obsdata, modeldata)

Compute the natural log of the likelihood that the data fits the model.

Parameters:

obsdataStarData object

observed data for a reddened star

moddataModelData object

all the information about the model spectra

Returns:

lnpfloat

natural log of the likelihood

lnprior()

Compute the natural log of the priors. Only Gaussian priors currently supported.

Returns:

lnpfloat

natural log of the prior

lnprob(obsdata, modeldata)

Compute the natural log of the probability

Parameters:

obsdataStarData object

observed data for a reddened star

moddataModelData object

all the information about the model spectra

parameters()

Give all the parameters values in a vector (fixed or not).

Returns:

paramsnp array

parameters values

parameters_to_fit()

Give the non-fixed parameters values in a vector. Needed for most fitters/samplers.

Returns:

paramsnp array

non-fixed parameters values

plot(obsdata, modinfo, resid_range=10.0, lyaplot=False, xticks=[0.1, 0.2, 0.3, 0.5, 0.7, 1.0, 2.0])

Standard plot showing the data and best fit.

Parameters:

obsdataStarData object

observed data for a reddened star

moddataModelData object

all the information about the model spectra

resid_rangefloat

percentage value for the +/- range for the residual plot

lyaplotboolean

set to add two panels giving the Ly-alpha fit and residuals

xticksvector

set to a vector of floats giving the values for the xticks

plot_sampler_chains(sampler)

Plot the samplier chains.

Parameters:

samplerobject

emcee sampler object

Returns:

figobject

returns the standard matplotlib fig info

plot_sampler_corner(flat_samples)

Plot the standard corner plot.

Returns:

figobject

returns the standard matplotlib fig info

pprint_parameters()

Print the parameters with names and values

save_parameters(filename=None)

Save the parameters and uncertainties to a table. Include if they were fixed, their bounds, and their priors.

Parameters:

filenamestr

name of the file for the saved info

Returns:

otabastropy table

table giving the results, often output with the computed extinction curve

set_initial_norm(obsdata, modeldata)

Set the initial normalization that puts the current model at the average level of the observed data. The normalization is a fit parameter, so is included fully in the fitting.

Parameters:

obsdataStarData object

observed data for a reddened star

moddataModelData object

all the information about the model spectra

stellar_sed(moddata)

Compute the stellar SED from the model parameters.

If foreground dust extinction included, then also includes the foreground dust extinction. Including this here results in extinction curves that do not include the foreground extinction.

Parameters:

moddataModelData object

all the information about the model spectra

Returns:

seddict

SED with {‘bands’: band_sed, ‘spec’: spec_sed, …}