======================================
SDSS Spectroscopic Templates with PyDL
======================================

One of the original motivations for creating the PyDL package was to reproduce
and, ultimately, improve the method for generating spectroscopic templates
for the Sloan Digital Sky Survey (SDSS_).  The spectroscopic templates are
used by the "1D" portion of the SDSS spectroscopic pipeline
(Bolton *et al.* 2012 [1]_; See also :mod:`pydl.pydlspec2d`).  Historically,
those were generated by:

1. Selecting individual objects (galaxies, QSOs, stars, white dwarfs, etc.)
   with good SDSS spectra and reliable classifications.
2. For a given set of objects, resample the spectra onto a common wavelength
   grid and redshift to rest frame.
3. Perform an iterative PCA analysis on the flux versus wavelength matrix.
4. Select the most significant eigenspectra as the templates.  Typically the
   first four eigenspectra were chosen.

For SDSS-I/II, the input spectra were all high signal-to-noise, but as the BOSS_
survey in SDSS-III started to target more distant galaxies, there was interest
in improving the robustness of the PCA fitting to handle low signal-to-noise
inputs.  In addition, full alternatives to PCA, such has
Heteroscedastic Matrix Factorization (HMF) [2]_, [3]_ could be compared.

Here are the algorithmic steps taken in the code to implement the
procedure described above.

1. :func:`~pydl.pydlspec2d.spec1d.template_input` reads a configuration file
   that includes a list of individual spectra.  There is one configuration
   file for each type of object (galaxy, QSO, etc.).
2. :func:`~pydl.pydlspec2d.spec1d.readspec` reads each spectrum.  Masked
   areas are identified by setting the per-spectrum inverse variance to
   zero as needed.
3. A new wavelength grid is created.  SDSS typically uses wavelength grids
   that are evenly spaced in :math:`\log_{10}` wavelength.
4. :func:`~pydl.pydlspec2d.spec1d.preprocess_spectra` calls
   :func:`~pydl.pydlspec2d.spec2d.combine1fiber` to perform the resampling
   and rest frame shift.
5. :func:`~pydl.pydlspec2d.spec1d.pca_solve` or
   :class:`~pydl.pydlspec2d.spec1d.HMF` are used to obtain the eigenspectra.
6. The eigenspectra are written to a FITS file.

At least historically, there was no test data set on which to (unit) test
these algorithms.  The best one could do was to take the inputs and compare them,
function-by-function, to the equivalent `IDL®`_ code, ensuring that the
results were the same, to some numerical precision.

.. _SDSS: https://www.sdss.org
.. _BOSS: https://www.sdss.org/surveys/boss/
.. _`IDL®`: http://www.harrisgeospatial.com/SoftwareTechnology/IDL.aspx
.. [1] `Bolton, Adam, et al. 2012 AJ 144, 144 <http://adsabs.harvard.edu/abs/2012AJ....144..144B>`_.
.. [2] `Tsalmantza, P., Decarli, R., Dotti, M., Hogg, D. W., 2011 ApJ 738, 20
    <http://adsabs.harvard.edu/abs/2011ApJ...738...20T>`_
.. [3] `Tsalmantza, P., Hogg, D. W., 2012 ApJ 753, 122
    <http://adsabs.harvard.edu/abs/2012ApJ...753..122T>`_
