Astronomers measure distance to galaxies in terms of 'redshift' — the far-off reddening of a galaxy's light as it zooms away from us. Current spectroscopic techniques measure at most a few hundred redshifts at once, but a new computer code can analyse larger batches of galaxies from digital images. It was developed by James Wray of Cornell University in Ithaca, New York, and Jim Gunn of Princeton University in New Jersey.
The algorithm works with general properties such as the colour and distribution of light across each galaxy. It gives good estimates of redshift when tested on 221,617 galaxies from the Sloan Digital Sky Survey, and could soon be used in the Sloan and other digital surveys to create three-dimensional galactic plots such as the one pictured.