logo Another tool from the A* Vienna software team

Pyckles is a super simple, light-weight interface to the Pickles (1998) catalogue of stellar spectra

https://travis-ci.org/astronomyk/Pyckles.svg?branch=master https://readthedocs.org/projects/pyckles/badge/?version=latest

Minimum requirements:

pip install pyckles


07.07.2020 Update to Brown et al. (2014) catalogue

Some referencing discrepancies were found in the Pyckles version of the Brown+ (2014) galaxy spectra catalogue. Pyckles will automatically re-download the catalogue if there is nothing in the astropy cache.

You can clear the astropy cache using astropy.utils.data.clear_download_cache()


July 2022: The downloadable content server was retired and the data migrated to a new server.

Pyckles v0.2 and above have been redirected to a new server URL.

For older versions, please either upgrade to the latest version (pip install --upgrade pyckles), or follow these instructions to update the server URL in the config file.

Which spectra are available


The package was originally intended only for the Pickles catalogue, but it now also has access to the Brown (2014) galaxy spectra catalogue

To list which catalogues are available, call get_catalog_list:

>>> import pyckles
>>> pyckles.get_catalog_list()
<Table length=2>
 name    type        filename
 str7    str7         str19
------- ------- -------------------
pickles stellar pickles98_full.fits
 brown  galaxy   brown14_full.fits

To access any of the spectra in the library, we need to create a SpectralLibrary object using the name of one of the available libraries:

>>> spec_lib = pyckles.SpectralLibrary("pickles")

To see a list of all the spectra in the catalogue, use the available_spectra attribute:

>>> spec_lib.available_spectra
<Column name='name' dtype='str5' length=131>

Accessing spectra

The spectra can be accessed both as items and as attributes of the SpectralLibrary.

As an item:

>>> a0v = spec_lib["A0V"]

or as an attribute:

>>> a0v = spec_lib.A0V


If there there is a space in the name, the attribute call will not work - use the item call: spec_lib["my spec"]

By default Pyckles returns spectra as astropy.fits.BinTableHDU objects, however depending on the use case, Pyckles can return spectra in any one of the following formats. To change the way spectra are returned, we need to set the meta parameter in the SpectralLibrary object:

>>> spec_lib.meta["return_style"] = "synphot"      # default is "fits"

Acceptable settings are:

  • fits: returns spectra as fits.BinTableHDU objects. wavelength and flux information is contained in the .data attribute of the HDU

  • synphot: returns the spectra as synphot.SourceSpectrum objects

  • quantity: returns the spectra as a tuple of two astropy.Quantity array objects in the format (wavelength, flux)

  • array: returns the spectra as a tuple of two numpy.ndarray objects in the format (wavelength, flux). The units here are those which were used in the FITS library (normally angstrom and erg s-1 cm-2 angstrom-1)

from matplotlib import pyplot as plt
import pyckles
spec_lib = pyckles.SpectralLibrary("pickles")

plt.plot(spec_lib.A0V.data["wavelength"], spec_lib.A0V.data["flux"])
plt.plot(spec_lib["G2V"].data["wavelength"], spec_lib["G2V"].data["flux"])

Adding libraries

If you have a spectral library that you would like added to pyckles, please open an issue on github and we can talk about about adding in.

Library file formatting

Libraries are stored in a single FITS files with a contents table in the first extension, and the spectra in further BinTableHDU extensions


HDU type






Index table mapping spectra to extensions



[n additional tables of spectra indexes]



First of m spectra



m-th spectra

Index table extensions

The first extension should contain a table with (at a minimum) the following columns:









Additional spectra index tables can be included, but this means that all index tables must be updated to keep the extension column pointing to the correct spectra. For example in the pickles library, ext 1 indexes all the spectra contained in the original Pickles (1998) library, however for ease of use, I have added a further extension index in ext 2 which only includes the main sequence stars. Additionally ext 3 contains a table with physical properties of each of the spectral types, such as temperature and evolutionary stage. The actual spectra are only contained onwards from ext 4.

Spectra extensions

The spectra are simple BinTableHDU objects containing 2 columns: wavelength and flux.



If astropy is used to create these tables the headers will be taken care of. If not, the following header information should be included:

TTYPE1  = 'wavelength'
TFORM1  = 'D       '
TUNIT1  = 'Angstrom'
TTYPE2  = 'flux    '
TFORM2  = 'D       '
TUNIT2  = 'Angstrom-1 cm-1 erg s-1'

where the TUNITn keywords are in readable by astropy.units.


If you find any bugs with the code, or have suggestions for how the code can be improved, please open an issue on github: https://github.com/astronomyk/Pyckles

If you are feeling adventurous, you’re more than welcome to play with the code and submit a pull request with any changes you deem necessary.

We’re more than happy for this to be incorporated into the astropy framework, if anyone feels like taking on that challenge…

Written by Kieran Leschinski