Photometric redshifts using Artificial Neural Networks (Collister & Lahav 2004)
ANNz is a freely available software package for photometric redshift estimation using Artificial Neural Networks. ANNz learns the relation between photometry and redshift from an appropriate training set of galaxies for which the redshift is already known. Where a large and representative training set is available ANNz is a highly competitive tool when compared with traditional template-fitting methods.
The following plots show a comparison between the performance of ANNz (left) and the popular photometric redshift code Hyperz (right). Hyperz was applied with the widely used Coleman, Wu and Weedman template spectra. Each plot shows, for 10,000 Sloan Digital Sky Survey galaxies, the predicted photometric redshift (on the vertical axis) against the actual (spectroscopic) redshift.
Further details of the ANNz method and package are given in:
* Firth, A.E., Lahav, O. & Somerville, R.S., 2003, MNRAS, 339, 1195
* Collister, A. A. & Lahav, O., 2004, PASP, 116, 345 (astro-ph/0311058)
Source code: annz.src.tar.gz (1.0 Mb)
Decompress and unpack the downloaded file using:
> gtar -xzf annz.src.tar.gz
This will create the directory annz, with subdirectories docs , example and src.
The ANNz user guide can be found in the docs directory.
The example directory contains an example data set consisting of real data from the Sloan Digital Sky Survey Data Release 1, to allow you to quickly test the functionality of ANNz. This has dereddened ugriz photometry and spectroscopic redshifts for each galaxy (including those in the testing set). The sample is a selection of galaxies from the Main and Luminous Red Galaxy samples, and has been split at random into training, testing and validation sets.
The ANNz source code is found in the src directory. To compile, give the command make all from this directory. The three executables annz_net, annz_train and annz_test can then be moved to your ~/bin directory (or any other directory in your PATH) so you can access the programs easily from the command line.