Module builder

source code

SED builder program.

To construct an SED, use the SED class. The functions defined in this module are mainly convenience functions specifically for that class, but can be used outside of the SED class if you know what you're doing.

Table of contents:

1. Retrieving and plotting photometry of a target
2. Where/what is my target?
3. SED fitting using a grid based approach
   • Binar star ### W.I.P ###
   • Saving SED fits
   • Loading SED fits ### BROKEN ###
4. Accessing the best fitting full SED model
5. Radii, distances and luminosities
   • Relations between quantities
   • Parallaxes
   • Seismic constraints
   • Reddening constraints
   • Evolutionary constraints

Section 1. Retrieving and plotting photometry of a target

>>> mysed = SED('HD180642')
>>> mysed.get_photometry()
>>> mysed.plot_data()

and call Pylab's show function to show to the screen:

]]include figure]]ivs_sed_builder_example_photometry.png]

Catch IndexErrors and TypeErrors in case no photometry is found.

You can give a search radius to get_photometry via the keyword radius. The default value is 10 arcseconds for stars dimmer than 6th magnitude, and 60 arcseconds for brighter stars. The best value of course depends on the density of the field.

>>> mysed.get_photometry(radius=5.)

If your star's name is not recognised by any catalog, you can give coordinates to look for photometry. In that case, the ID of the star given in the SED command will not be used to search photometry (only to save the phot file):

>>> mysed.get_photometry(ra=289.31167983,dec=1.05941685)

Note that ra and dec are given in degrees.

You best switch on the logger (see ivs.aux.loggers.get_basic_logger) to see the progress: sometimes, access to catalogs can take a long time (the GATOR sources are typically slow). If one of the gator, vizier or gcpd is impossibly slow or the site is down, you can include/exclude these sources via the keywords include or exclude, which take a list of strings (choose from gator, vizier and/or gcpd). For ViZieR, there is an extra option to change to another mirror site via

>>> vizier.change_mirror()

The vizier.change_mirror function cycles through all the mirrors continuously, so sooner or later you will end up with the default one and repeat the cycle.

>>> mysed.get_photometry(exclude=['gator'])

The results will be written to the file HD180642.phot. An example content is:

 #  meas    e_meas flag unit photband          source                        _r   _RAJ2000    _DEJ2000   cwave       cmeas     e_cmeas cunit       color include
 #float64   float64 |S20 |S30 |S30              |S50                     float64    float64     float64 float64     float64     float64 |S50         bool    bool
   7.823      0.02 nan  mag  WISE.W3           wise_prelim_p3as_psd    0.112931  2.667e-05   2.005e-05  123337 4.83862e-17 8.91306e-19 erg/s/cm2/AA     0       1
   7.744     0.179 nan  mag  WISE.W4           wise_prelim_p3as_psd    0.112931  2.667e-05   2.005e-05  222532 4.06562e-18 6.70278e-19 erg/s/cm2/AA     0       1
    7.77     0.023 nan  mag  WISE.W1           wise_prelim_p3as_psd    0.112931  2.667e-05   2.005e-05 33791.9   6.378e-15  1.3511e-16 erg/s/cm2/AA     0       1
   7.803      0.02 nan  mag  WISE.W2           wise_prelim_p3as_psd    0.112931  2.667e-05   2.005e-05   46293 1.82691e-15 3.36529e-17 erg/s/cm2/AA     0       1
   8.505     0.016 nan  mag  TYCHO2.BT         I/259/tyc2                 0.042   7.17e-06    1.15e-06  4204.4 2.76882e-12 4.08029e-14 erg/s/cm2/AA     0       1
   8.296     0.013 nan  mag  TYCHO2.VT         I/259/tyc2                 0.042   7.17e-06    1.15e-06 5321.86 1.93604e-12 2.31811e-14 erg/s/cm2/AA     0       1
    8.27       nan nan  mag  JOHNSON.V         II/168/ubvmeans             0.01    1.7e-07    3.15e-06 5504.67 1.80578e-12 1.80578e-13 erg/s/cm2/AA     0       1
    0.22       nan nan  mag  JOHNSON.B-V       II/168/ubvmeans             0.01    1.7e-07    3.15e-06     nan     1.40749    0.140749 flux_ratio      1       0
   -0.66       nan nan  mag  JOHNSON.U-B       II/168/ubvmeans             0.01    1.7e-07    3.15e-06     nan     1.21491    0.121491 flux_ratio      1       0
    8.49       nan nan  mag  JOHNSON.B         II/168/ubvmeans             0.01    1.7e-07    3.15e-06 4448.06 2.54162e-12 2.54162e-13 erg/s/cm2/AA     0       1
    7.83       nan nan  mag  JOHNSON.U         II/168/ubvmeans             0.01    1.7e-07    3.15e-06 3641.75 3.08783e-12 3.08783e-13 erg/s/cm2/AA     0       1
   2.601       nan nan  mag  STROMGREN.HBN-HBW J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685     nan     1.66181    0.166181 flux_ratio      1       0
  -0.043       nan nan  mag  STROMGREN.M1      J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685     nan    0.961281   0.0961281 flux_ratio      1       0
   8.221       nan nan  mag  STROMGREN.Y       J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685 5477.32 1.88222e-12 1.88222e-13 erg/s/cm2/AA     0       1
   0.009       nan nan  mag  STROMGREN.C1      J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685     nan     0.93125    0.093125 flux_ratio      1       0
   0.238       nan nan  mag  STROMGREN.B-Y     J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685     nan     1.28058    0.128058 flux_ratio      1       0
   8.459       nan nan  mag  STROMGREN.B       J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685  4671.2 2.41033e-12 2.41033e-13 erg/s/cm2/AA     0       1
   8.654       nan nan  mag  STROMGREN.V       J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685 4108.07 2.96712e-12 2.96712e-13 erg/s/cm2/AA     0       1
   8.858       nan nan  mag  STROMGREN.U       J/A+A/528/A148/tables       0.54 -5.983e-05 -0.00013685 3462.92 3.40141e-12 3.40141e-13 erg/s/cm2/AA     0       1
    7.82      0.01 nan  mag  JOHNSON.J         J/PASP/120/1128/catalog     0.02  1.017e-05    3.15e-06 12487.8 2.36496e-13 2.36496e-15 erg/s/cm2/AA     0       1
    7.79      0.01 nan  mag  JOHNSON.K         J/PASP/120/1128/catalog     0.02  1.017e-05    3.15e-06 21951.2 3.24868e-14 3.24868e-16 erg/s/cm2/AA     0       1
    7.83      0.04 nan  mag  JOHNSON.H         J/PASP/120/1128/catalog     0.02  1.017e-05    3.15e-06 16464.4 8.64659e-14 3.18552e-15 erg/s/cm2/AA     0       1
  8.3451    0.0065 nan  mag  HIPPARCOS.HP      I/239/hip_main             0.036   2.17e-06    1.15e-06 5275.11 1.91003e-12 1.91003e-14 erg/s/cm2/AA     0       1
   8.525     0.011 nan  mag  TYCHO2.BT         I/239/hip_main             0.036   2.17e-06    1.15e-06  4204.4 2.71829e-12   2.754e-14 erg/s/cm2/AA     0       1
   8.309     0.012 nan  mag  TYCHO2.VT         I/239/hip_main             0.036   2.17e-06    1.15e-06 5321.86   1.913e-12 2.11433e-14 erg/s/cm2/AA     0       1
    8.02     0.057 nan  mag  COUSINS.I         II/271A/patch2               0.2 -4.983e-05   2.315e-05 7884.05 7.51152e-13 3.94347e-14 erg/s/cm2/AA     0       1
   8.287     0.056 nan  mag  JOHNSON.V         II/271A/patch2               0.2 -4.983e-05   2.315e-05 5504.67 1.77773e-12 9.16914e-14 erg/s/cm2/AA     0       1
    8.47       nan nan  mag  USNOB1.B1         I/284/out                  0.026   7.17e-06     1.5e-07 4448.06 2.57935e-12 7.73805e-13 erg/s/cm2/AA     0       1
    8.19       nan nan  mag  USNOB1.R1         I/284/out                  0.026   7.17e-06     1.5e-07 6939.52 1.02601e-12 3.07803e-13 erg/s/cm2/AA     0       1
   8.491     0.012 nan  mag  JOHNSON.B         I/280B/ascc                0.036   2.17e-06    2.15e-06 4448.06 2.53928e-12 2.80651e-14 erg/s/cm2/AA     0       1
   8.274     0.013 nan  mag  JOHNSON.V         I/280B/ascc                0.036   2.17e-06    2.15e-06 5504.67 1.79914e-12 2.15419e-14 erg/s/cm2/AA     0       1
   7.816     0.023 nan  mag  2MASS.J           II/246/out                 0.118 -2.783e-05   1.715e-05 12412.1 2.28049e-13 4.83093e-15 erg/s/cm2/AA     0       1
   7.792     0.021 nan  mag  2MASS.KS          II/246/out                 0.118 -2.783e-05   1.715e-05 21909.2 3.26974e-14 6.32423e-16 erg/s/cm2/AA     0       1
   7.825     0.042 nan  mag  2MASS.H           II/246/out                 0.118 -2.783e-05   1.715e-05 16497.1 8.48652e-14 3.28288e-15 erg/s/cm2/AA     0       1
   8.272     0.017 nan  mag  GENEVA.V          GCPD                         nan        nan         nan  5482.6 1.88047e-12 2.94435e-14 erg/s/cm2/AA     0       1
     1.8     0.004 nan  mag  GENEVA.G-B        GCPD                         nan        nan         nan     nan    0.669837  0.00669837 flux_ratio      1       0
   1.384     0.004 nan  mag  GENEVA.V1-B       GCPD                         nan        nan         nan     nan    0.749504  0.00749504 flux_ratio      1       0
    0.85     0.004 nan  mag  GENEVA.B1-B       GCPD                         nan        nan         nan     nan     1.05773   0.0105773 flux_ratio      1       0
   1.517     0.004 nan  mag  GENEVA.B2-B       GCPD                         nan        nan         nan     nan    0.946289  0.00946289 flux_ratio      1       0
   0.668     0.004 nan  mag  GENEVA.V-B        GCPD                         nan        nan         nan     nan    0.726008  0.00726008 flux_ratio      1       0
   0.599     0.004 nan  mag  GENEVA.U-B        GCPD                         nan        nan         nan     nan     1.13913   0.0113913 flux_ratio      1       0
   7.604 0.0174642 nan  mag  GENEVA.B          GCPD                         nan        nan         nan 4200.85 2.59014e-12 4.16629e-14 erg/s/cm2/AA     0       1
   9.404 0.0179165 nan  mag  GENEVA.G          GCPD                         nan        nan         nan 5765.89 1.73497e-12   2.863e-14 erg/s/cm2/AA     0       1
   8.988 0.0179165 nan  mag  GENEVA.V1         GCPD                         nan        nan         nan 5395.63 1.94132e-12 3.20351e-14 erg/s/cm2/AA     0       1
   8.454 0.0179165 nan  mag  GENEVA.B1         GCPD                         nan        nan         nan 4003.78 2.73968e-12 4.52092e-14 erg/s/cm2/AA     0       1
   9.121 0.0179165 nan  mag  GENEVA.B2         GCPD                         nan        nan         nan 4477.56 2.45102e-12  4.0446e-14 erg/s/cm2/AA     0       1
   8.203 0.0179165 nan  mag  GENEVA.U          GCPD                         nan        nan         nan 3421.62 2.95052e-12 4.86885e-14 erg/s/cm2/AA     0       1
    8.27       nan nan  mag  JOHNSON.V         GCPD                         nan        nan         nan 5504.67 1.80578e-12 1.80578e-13 erg/s/cm2/AA     0       1
    0.22       nan nan  mag  JOHNSON.B-V       GCPD                         nan        nan         nan     nan     1.40749    0.140749 flux_ratio      1       0
   -0.66       nan nan  mag  JOHNSON.U-B       GCPD                         nan        nan         nan     nan     1.21491    0.121491 flux_ratio      1       0
    8.49       nan nan  mag  JOHNSON.B         GCPD                         nan        nan         nan 4448.06 2.54162e-12 2.54162e-13 erg/s/cm2/AA     0       1
    7.83       nan nan  mag  JOHNSON.U         GCPD                         nan        nan         nan 3641.75 3.08783e-12 3.08783e-13 erg/s/cm2/AA     0       1
  -0.035       nan nan  mag  STROMGREN.M1      GCPD                         nan        nan         nan     nan    0.954224   0.0954224 flux_ratio      1       0
   0.031       nan nan  mag  STROMGREN.C1      GCPD                         nan        nan         nan     nan     0.91257    0.091257 flux_ratio      1       0
   0.259       nan nan  mag  STROMGREN.B-Y     GCPD                         nan        nan         nan     nan     1.25605    0.125605 flux_ratio      1       0
  -0.009 0.0478853 nan  mag  2MASS.J-H         II/246/out                 0.118 -2.783e-05   1.715e-05     nan     2.68719    0.118516 flux_ratio      1       0
  -0.033 0.0469574 nan  mag  2MASS.KS-H        II/246/out                 0.118 -2.783e-05   1.715e-05     nan    0.385286   0.0166634 flux_ratio      1       0
   -0.01 0.0412311 nan  mag  JOHNSON.J-H       J/PASP/120/1128/catalog     0.02  1.017e-05    3.15e-06     nan     2.73514    0.103867 flux_ratio      1       0
   -0.04 0.0412311 nan  mag  JOHNSON.K-H       J/PASP/120/1128/catalog     0.02  1.017e-05    3.15e-06     nan    0.375718    0.014268 flux_ratio      1       0
   0.209 0.0206155 nan  mag  TYCHO2.BT-VT      I/259/tyc2                 0.042   7.17e-06    1.15e-06     nan     1.43014    0.027155 flux_ratio      1       0

Once a .phot file is written and get_photometry is called again for the same target, the script will not retrieve the photometry from the internet again, but will use the contents of the file instead. The purpose is minimizing network traffic and maximizing speed. If you want to refresh the search, simply manually delete the .phot file or set force=True when calling get_photometry.

The content of the .phot file is most easily read using the ivs.inout.ascii.read2recarray function. Be careful, as it contains both absolute fluxes as flux ratios.

>>> data = ascii.read2recarray('HD180642.phot')

Notice that in the .phot files, also a comment column is added. You can find translation of some of the flags here (i.e. upper limit, extended source etc..), or sometimes just additional remarks on variability etc. Not all catalogs have this feature implemented, so you are still responsible yourself for checking the quality of the photometry.

The references to each source are given in the bibtex column. Simply call

>>> mysed.save_bibtex()

to convert those bibcodes to a .bib file.

Using SED.plot_MW_side and SED.plot_MW_top, you can make a picture of where your star is located with respect to the Milky Way and the Sun. With SED.plot_finderchart, you can check the location of your photometry, and also see if proper motions etc are available.

Section 2. Where/what is my target?

To give you some visual information on the target, the following plotting procedure might be of some help.

To check whether the downloaded photometry is really belonging to the target, instead of some neighbouring star (don't forget to set radius when looking for photometry!), you can generate a finderchart with the location of the downloaded photometry overplotted. On top of that, proper motion data is added when available, as well as radial velocity data. When a distance is available, the proper motion velocity will be converted to a true tangential velocity.

>>> p = pl.figure();mysed.plot_finderchart(window_size=1)

]]include figure]]ivs_sed_builder_finderchart.png]

To know the location of your target wrt the Milky Way (assuming your target is in the milky way), you can call

>>> p = pl.figure();mysed.plot_MW_side()
>>> p = pl.figure();mysed.plot_MW_top()

]]include figure]]ivs_sed_builder_MWside.png]

]]include figure]]ivs_sed_builder_MWtop.png]

Section 3. SED fitting using a grid based approach

Subsection 3.1 Single star

We make an SED of HD180642 by simply exploring a whole grid of Kurucz models (constructed via fit.generate_grid, iterated over via fit.igrid_search and evaluated with fit.stat_chi2). The model with the best parameters is picked out and we make a full SED with those parameters.

>>> mysed = SED('HD180642')
>>> mysed.get_photometry()

Now we have collected all fluxes and colors, but we do not want to use them all: first, fluxes and colors are not independent, so you probably want to use either only absolute fluxes or only colors (plus perhaps one absolute flux per system to compute the angular diameter) (see SED.set_photometry_scheme). Second, some photometry is better suited for fitting an SED than others; typically IR photometry does not add much to the fitting of massive stars, or it can be contaminated with circumstellar material. Third, some photometry is not so reliable, i.e. the measurements can have large systematic uncertainties due to e.g. calibration effects, which are typically not included in the error bars.

Currently, four standard schemes are implemented, which you can set via SED.set_photometry_scheme:

1. absolute: use only absolute values
2. colors: use only colors (no angular diameter values calculated)
3. combo: use all colors and one absolute value per photometric system
4. irfm: (infrared flux method) use colors for wavelengths shorter than infrared wavelengths, and absolute values for systems in the infrared. The infrared is defined as wavelength longer than 1 micron, but this can be customized with the keyword infrared=(value,unit) in SED.set_photometry_scheme.

Here, we chose to use colors and one absolute flux per system, but exclude IR photometry (wavelength range above 2.5 micron), and some systems and colors which we know are not so trustworthy:

>>> mysed.set_photometry_scheme('combo')
>>> mysed.exclude(names=['STROMGREN.HBN-HBW','USNOB1','SDSS','DENIS','COUSINS','ANS','TD1'],wrange=(2.5e4,1e10))

You can include/exclude photoemtry based on name, wavelength range, source and index, and only select absolute photometry or colors (include_abs,include_colors). When working in interactive mode, in particular the index is useful. Print the current set of photometry to the screen with

>>> print(photometry2str(mysed.master,color=True,index=True))

and you will see in green the included photometry, and in red the excluded photometry. You will see that each column is preceded by an index, you can use these indices to select/deselect the photometry.

Speed up the fitting process by copying the model grids to the scratch disk

>>> model.copy2scratch(z='*')

Start the grid based fitting process and show some plots. We use 100000 randomly distributed points over the grid:

>>> mysed.igrid_search(points=100000)

Delete the model grids from the scratch disk

>>> model.clean_scratch()

and make the plot

>>> p = pl.figure()
>>> p = pl.subplot(131);mysed.plot_sed()
>>> p = pl.subplot(132);mysed.plot_grid(limit=None)
>>> p = pl.subplot(133);mysed.plot_grid(x='ebv',y='z',limit=None)

]]include figure]]ivs_sed_builder_example_fitting01.png]

The grid is a bit too coarse for our liking around the minimum, so we zoom in on the results:

>>> teffrange = mysed.results['igrid_search']['CI']['teffL'],mysed.results['igrid_search']['CI']['teffU']
>>> loggrange = mysed.results['igrid_search']['CI']['loggL'],mysed.results['igrid_search']['CI']['loggU']
>>> ebvrange = mysed.results['igrid_search']['CI']['ebvL'],mysed.results['igrid_search']['CI']['ebvU']
>>> mysed.igrid_search(points=100000,teffrange=teffrange,loggrange=loggrange,ebvrange=ebvrange)

and repeat the plot

>>> p = pl.figure()
>>> p = pl.subplot(131);mysed.plot_sed(plot_deredded=True)
>>> p = pl.subplot(132);mysed.plot_grid(limit=None)
>>> p = pl.subplot(133);mysed.plot_grid(x='ebv',y='z',limit=None)

]]include figure]]ivs_sed_builder_example_fitting02.png]

You can automatically make plots of (most plotting functions take colors=True/False as an argument so you can make e.g. the 'color' SED and 'absolute value' SED):

1. the grid (see SED.plot_grid)
2. the SED (see SED.plot_sed)
3. the fit statistics (see SED.plot_chi2)

To change the grid, load the ivs.sed.model module and call ivs.sed.model.set_defaults with appropriate arguments. See that module for conventions on the grid structure when adding custom grids.

To add arrays manually, i.e. not from the predefined set of internet catalogs, use the SED.add_photometry_fromarrays function.

Warning: Be careful when interpreting the Chi2 results. In order to always have a solution, the chi2 is rescaled so that the minimum equals 1, in the case the probability of the best chi2-model is zero. The Chi2 rescaling factor f mimicks a rescaling of all errorbars with a factor sqrt(f), and does not discriminate between systems (i.e., all errors are blown up). If the errorbars are underestimated, it could be that the rescaling factor is also wrong, which means that the true probability region can be larger or smaller!

Subsection 3.2 Binary star - ### W.I.P ###

The SED class can create SEDs for multiple stars as well. There are 2 options available, the multiple SED fit which in theory can handle any number of stars, and the binary SED fit which is for binaries only, and uses the mass of both components to restrict the radii when combining two model SEDs.

As an example we take the system PG1104+243, which consists of a subdwarf B star, and a G2 type mainsequence star. The photometry from the standard catalogues that are build in this class, is of to low quality, so we use photometry obtained from the subdwarf database.

>>> mysed = SED('PG1104+243')

>>> meas, e_meas, units, photbands, source = ascii.read2array('pg1104+243_sddb.phot', dtype=str)
>>> meas = np.array(meas, dtype=float)
>>> e_meas = np.array(e_meas, dtype=float)
>>> mysed.add_photometry_fromarrays(meas, e_meas, units, photbands, source)

We use only the absolute fluxes

>>> mysed.set_photometry_scheme('abs')

For the main sequence component we use kurucz models with solar metalicity, and for the sdB component tmap models. And we copy the model grids to the scratch disk to speed up the process:

>>> grid1 = dict(grid='kurucz',z=+0.0)
>>> grid2 = dict(grid='tmap')
>>> model.set_defaults_multiple(grid1,grid2)
>>> model.copy2scratch()

The actual fitting. The second fit starts from the 95% probability intervals of the first fit.

>>> teff_ms = (5000,7000)
>>> teff_sdb = (25000,45000)
>>> logg_ms = (4.00,4.50)
>>> logg_sdb = (5.00,6.50)
>>> mysed.igrid_search(masses=(0.47,0.71) ,teffrange=(teff_ms,teff_fix),loggrange=(logg_ms,logg_sdb), ebvrange=(0.00,0.02), zrange=(0,0), points=2000000, type='binary')
>>> mysed.igrid_search(masses=(0.47,0.71) ,points=2000000, type='binary')

Delete the used models from the scratch disk

>>> model.clean_scratch()

Plot the results

>>> p = pl.figure()
>>> p = pl.subplot(131); mysed.plot_sed(plot_deredded=False)
>>> p = pl.subplot(132); mysed.plot_grid(x='teff', y='logg', limit=0.95)
>>> p = pl.subplot(133); mysed.plot_grid(x='teff-2', y='logg-2', limit=0.95)

]]include figure]]ivs_sed_builder_example_fittingPG1104+243.png]

Subsection 3.3 Saving SED fits

You can save all the data to a multi-extension FITS file via

>>> mysed.save_fits()

This FITS file then contains all measurements (it includes the .phot file), the resulting SED, the confidence intervals of all parameters and also the whole fitted grid: in the above case, the extensions of the FITS file contain the following information (we print part of each header):

   EXTNAME = 'DATA    '
   XTENSION= 'BINTABLE'           / binary table extension
   BITPIX  =                    8 / array data type
   NAXIS   =                    2 / number of array dimensions
   NAXIS1  =                  270 / length of dimension 1
   NAXIS2  =                   67 / length of dimension 2
   TTYPE1  = 'meas    '
   TTYPE2  = 'e_meas  '
   TTYPE3  = 'flag    '
   TTYPE4  = 'unit    '
   TTYPE5  = 'photband'
   TTYPE6  = 'source  '
   TTYPE7  = '_r      '
   TTYPE8  = '_RAJ2000'
   TTYPE9  = '_DEJ2000'
   TTYPE10 = 'cwave   '
   TTYPE11 = 'cmeas   '
   TTYPE12 = 'e_cmeas '
   TTYPE13 = 'cunit   '
   TTYPE14 = 'color   '
   TTYPE15 = 'include '
   TTYPE16 = 'synflux '
   TTYPE17 = 'mod_eff_wave'
   TTYPE18 = 'chi2    '

   EXTNAME = 'MODEL   '
   XTENSION= 'BINTABLE'           / binary table extension
   BITPIX  =                    8 / array data type
   NAXIS   =                    2 / number of array dimensions
   NAXIS1  =                   24 / length of dimension 1
   NAXIS2  =                 1221 / length of dimension 2
   TTYPE1  = 'wave    '
   TUNIT1  = 'A       '
   TTYPE2  = 'flux    '
   TUNIT2  = 'erg/s/cm2/AA'
   TTYPE3  = 'dered_flux'
   TUNIT3  = 'erg/s/cm2/AA'
   TEFFL   =    23000.68377498454
   TEFF    =    23644.49138963689
   TEFFU   =    29999.33189058337
   LOGGL   =    3.014445328877565
   LOGG    =    4.984788506855546
   LOGGU   =    4.996095055525759
   EBVL    =   0.4703171900728142
   EBV     =   0.4933185398871652
   EBVU    =   0.5645144879211454
   ZL      =   -2.499929434601112
   Z       =   0.4332336811329144
   ZU      =   0.4999776537652627
   SCALEL  = 2.028305798068863E-20
   SCALE   = 2.444606991671813E-20
   SCALEU  = 2.830281842143698E-20
   LABSL   =    250.9613352757437
   LABS    =     281.771013453664
   LABSU   =    745.3149766772975
   CHISQL  =    77.07958742733673
   CHISQ   =    77.07958742733673
   CHISQU  =    118.8587169011471
   CI_RAWL =   0.9999999513255379
   CI_RAW  =   0.9999999513255379
   CI_RAWU =   0.9999999999999972
   CI_RED  =   0.5401112973063139
   CI_REDL =   0.5401112973063139
   CI_REDU =   0.9500015229597392

   EXTNAME = 'IGRID_SEARCH'
   XTENSION= 'BINTABLE'           / binary table extension
   BITPIX  =                    8 / array data type
   NAXIS   =                    2 / number of array dimensions
   NAXIS1  =                   80 / length of dimension 1
   NAXIS2  =                99996 / length of dimension 2
   TTYPE1  = 'teff    '
   TTYPE2  = 'logg    '
   TTYPE3  = 'ebv     '
   TTYPE4  = 'z       '
   TTYPE5  = 'chisq   '
   TTYPE6  = 'scale   '
   TTYPE7  = 'escale '
   TTYPE8  = 'Labs    '
   TTYPE9  = 'CI_raw  '
   TTYPE10 = 'CI_red  '

Subsection 3.4 Loading SED fits ### BROKEN ###

Unfortunately this is not yet working properly!

Once saved, you can load the contents of the FITS file again into an SED object via

>>> mysed = SED('HD180642')
>>> mysed.load_fits()

and then you can build all the plots again easily. You can of course use the predefined plotting scripts to start a plot, and then later on change the properties of the labels, legends etc... for higher quality plots or to better suit your needs.

Section 4. Accessing the best fitting full SED model

You can access the full SED model that matches the parameters found by the fitting routine via:

>>> wavelength,flux,deredded_flux = mysed.get_best_model()

Note that this model is retrieved after fitting, and was not in any way used during the fitting. As a consequence, there could be small differences between synthetic photometry calculated from this returned model and the synthetic fluxes stored in mysed.results['igrid_search']['synflux'], which is the synthetic photometry coming from the interpolation of the grid of pre-interpolated photometry. See the documentation of SED.get_model for more information.

Section 5. Radii, distances and luminosities

Subsection 4.1. Relations between quantities

Most SED grids don't have the radius as a tunable model parameter. The scale factor, which is available for all fitted models in the grid when at least one absolute photometric point is included, is directly propertional to the angular diameter. The following relations hold:

   >> distance = radius / np.sqrt(scale)
   >> radius = distance * np.sqrt(scale)

Where radius and distance have equal units. The true absolute luminosity (solar units) is related to the absolute luminosity from the SED (solar units) via the radius (solar units):

   >> L_abs_true = L_abs * radius**2

Finally, the angular diameter can be computed via:

   >> 2*conversions.convert('sr','mas',scale)

Subsection 4.2. Seismic constraints

If the star shows clear solar-like oscillations, you can use the nu_max give an independent constraint on the surface gravity, given the effective temperature of the model (you are free to give errors or not, just keep in mind that in the former case, you will get an array of Uncertainties rather than floats):

   >> teff = 4260.,'K'
   >> nu_max = 38.90,0.86,'muHz'
   >> logg_slo = conversions.derive_logg_slo(teff,nu_max,unit='[cm/s2'])

If you have the large separation (l=0 modes) and the nu_max, you can get an estimate of the radius:

   >> Deltanu0 = 4.80,0.02,'muHz'
   >> R_slo = conversions.derive_radius_slo(numax,Deltanu0,teff,unit='Rsol')

Then from the radius, you can get both the absolute luminosity and distance to your star.

Subsection 4.3. Parallaxes

From the parallax, you can get an estimate of the distance. This is, however, dependent on some prior assumptions such as the shape of the galaxy and the distribution of stars. To estimate the probability density value due to a measured parallax of a star at particular distance, you can call distance.distprob:

   >> gal_latitude = 0.5
   >> plx = 3.14,0.5
   >> d_prob = distance.distprob(d,gal_latitude,plx)

Using this distance, you can get an estimate for the radius of your star, and thus also the absolute luminosity.

Subsection 4.4: Reddening constraints

An estimate of the reddening of a star at a particular distance may be obtained with the extinctionmodels.findext function. There are currently three reddening maps available: Drimmel, Marshall and Arenou:

   >> lng,lat = 120.,-0.5
   >> dist = 100. # pc
   >> Rv = 3.1
   >> EBV = extinctionmodels.findext(lng,lat,model='drimmel',distance=dist)/Rv
   >> EBV = extinctionmodels.findext(lng,lat,model='marshall',distance=dist)/Rv
   >> EBV = extinctionmodels.findext(lng,lat,model='arenou',distance=dist)/Rv