We explore the accuracy of the clustering-based redshift estimation proposed by ( ) when applied to VIPERS and CFHTLS real data .
This method enables us to reconstruct redshift distributions from measurement of the angular clustering of objects using a set of secure spectroscopic redshifts .
We use state-of-the art spectroscopic measurements with i _ { \text { AB } } < 22.5 from the VIMOS Public Extragalactic Redshift Survey ( VIPERS ) as reference population to infer the redshift distribution of galaxies from the Canada-France-Hawaii Telescope Legacy Survey ( CFHTLS ) T0007 release .
VIPERS provides a nearly representative sample to a flux limit of i _ { \text { AB } } < 22.5 at a redshift of > 0.5 which allows us to test the accuracy of the clustering-based redshift distributions .
We show that this method enables us to reproduce the true mean colour-redshift relation when both populations have the same magnitude limit .
We also show that this technique allows the inference of redshift distributions for a population fainter than the reference and we give an estimate of the colour-redshift mapping in this case .
This last point is of great interest for future large redshift surveys which require a complete faint spectroscopic sample .