Photometric redshift via Gaussian processes with physical kernels.

Read the documentation here: http://delight.readthedocs.io

Warning: this code is still in active development and is not quite ready to be blindly applied to arbitrary photometric galaxy surveys. But this day will come.

Tests: pytest for unit tests, PEP8 for code style, coveralls for test coverage.

./paper/: journal paper describing the method
./delight/: main code (Python/Cython)
./tests/: test suite for the main code
./notebooks/: demo notebooks using delight
./data/: some useful inputs for tests/demos
./docs/: documentation
./other/: useful mathematica notebooks, etc

Python 3.5, cython, numpy, scipy, pytest, pylint, coveralls, matplotlib, astropy, mpi4py

Boris Leistedt (NYU)
David W. Hogg (NYU) (Flatiron)

Please cite [Leistedt and Hogg (2016)] (https://arxiv.org/abs/1612.00847) if you use this code your research. The BibTeX entry is:

@article{delight,
    author  = "Boris Leistedt and David W. Hogg",
    title   = "Data-driven, Interpretable Photometric Redshifts Trained on Heterogeneous and Unrepresentative Data",
    journal = "The Astrophysical Journal",
    volume  = "838",
    number  = "1",
    pages   = "5",
    url     = "http://stacks.iop.org/0004-637X/838/i=1/a=5",
    year    = "2017",
    eprint         = "1612.00847",
    archivePrefix  = "arXiv",
    primaryClass   = "astro-ph.CO",
    SLACcitation   = "%%CITATION = ARXIV:1612.00847;%%"
}

Copyright 2016-2017 the authors. The code in this repository is released under the open-source MIT License. See the file LICENSE for more details.