:orphan:
.. _custom_ssp_libraries:
Acquiring and using alternative SSP libraries
==============================================
This section of the docs describes how to use DSPS with alternatives to the
default SSP libraries supplied at the DSPS data URL.
Whether you use the `python-fsps `__ wrapper
within DSPS, or whether you download the SSP libraries from an external location,
DSPS has a few convenience functions that you can use to load your SPS data.
See :ref:`dsps_drn_config` for more information.
Using python-fsps
----------------------------------
DSPS includes a convenience function for using
`python-fsps `__
to generate customizable versions of these SED libraries.
The following function implements a simple standalone
wrapper around python-fsps that returns SSP information
in a convenient form of arrays and matrices for use with the rest of DSPS.
.. code-block:: Python
from dsps.data_loaders import retrieve_ssp_data_from_fsps
With python-fsps there are many other options available for
generating a library of SSP spectra.
The code in this function is straightforward
to adapt to generate alternative SSP templates.
Using spectra from another library
----------------------------------
.. Important:: Mind your units and logarithms when using your own SSP library.
As we will see in the demo notebook and the docstrings of the source code,
stellar ages should be specified in units of Gyr (not in yr!)
and metallicity as the mass fraction of elements heavier than helium
(not noramlized by Zsun!), and both quantities should be supplied in base-10 log.
The SSP spectra are defined in units of Lsun/Hz as a function of
wavelength λ in Angstroms.
There are many other sources of SED libraries that are publicly available,
and you can use DSPS with whatever SED library is most appropriate for
your science application. As described in more detail :ref:`here `,
there is no need to use the data-loading functions in DSPS,
since these are just convenience functions that return a sequence of plain ndarrays.