The NNPDF2.2 parton set

We present a method developed by the NNPDF Collaboration that allows the inclusion of new experimental data into an existing set of parton distribution functions without the need for a complete refit. A Monte Carlo ensemble of PDFs may be updated by assigning each member of the ensemble a unique weight determined by Bayesian inference. The reweighted ensemble therefore represents the probability density of PDFs conditional on both the old and new data. This method is applied to the inclusion of W-lepton asymmetry data into the NNPDF2.1 fit producing a new PDF set, NNPDF2.2.