The paper describing the code was published in MNRAS with reference Baldry et al. 2014, MNRAS, 441, 2440. The tar file distribution includes, the IDL code, the arXiv version of the paper, as well as a readme.txt file with more details.
Please let me know your email address if you use the code (for emails regarding updates). Also, if you can suggest generalizations, e.g. keywords and spectral templates to add then please let me know. At the moment, the code is specific for AAOmega spectra (370nm to 890nm spliced together from the blue and red arms) cross-correlated against stellar templates and galaxy templates out to z < 0.9.
The GAMA survey has obtained spectra of over 230000 targets using the Anglo-Australian Telescope. To homogenise the redshift measurements and improve the reliability, a fully automatic redshift code was developed (autoz). The measurements were made using a cross-correlation method for both absorption-line and emission-line spectra. Large deviations in the high-pass filtered spectra are partially clipped in order to be robust against uncorrected artefacts and to reduce the weight given to single-line matches. A single figure of merit (FOM) was developed that puts all template matches onto a similar confidence scale. The redshift confidence as a function of the FOM was fitted with a tanh function using a maximum likelihood method applied to repeat observations of targets. The method could be adapted to provide robust automatic redshifts for other large galaxy redshift surveys. For the GAMA survey, there was a substantial improvement in the reliability of assigned redshifts and in the lowering of redshift uncertainties with a median velocity uncertainty of 33 km/s.
Links: GAMA survey - my home page.