Photometric redshifts are an estimate of the redshift of galaxies (or AGNs, quasars) using only large/medium band photometry instead of spectroscopy. The efficiency of the method relies in the identification of spectral breaks, i.e. strong spectral features, still recognizable after the integration of the Spectral Energy Distribution (hereafter SED) below the filter's transmission function. The precision of this estimate is worse than the spectroscopic redshift one, depending on the filters set and on the photometric accuracy, but for many cosmological and extragalactic applications the photometric redshift represents a sufficient information. Thus, the photometric redshift tool can be extensively applied and will rapidly become a crucial tool of observational cosmology. This is the reason inducing us to develop a public code, trying to introduce some improvement in the ``classical'' method, as the inclusion of the redshift probability distributions and of error bars rather than singular values.
Between the two methods most largely used, the empirical training set
(e.g. Connolly et al. 1995, Brunner et al. 1997) and the SED fitting (e.g.
Lanzetta et al. 1996, Fernández-Soto et al. 1999) we chose the latter,
because of its flexibility when different sets of data are used.