This repository contains code for the construction and sampling of a Gaussian Process (GP) from an input Parton Distribution Function (PDF) set.
Aside from a python installation and some standard python packages (numpy, matplotlib, argparse) the only other requirement is an LHAPDF6 installation callable from python.
gppdf.py [PDF set] [N]
Generates a numpy archive containing N
samples of a GP defined according to
the mean- and covariance-functions of the input PDF set
. For the format of the
archive see the header of gppdf.py
.
To quickly plot the resulting samples, use
gppdf_plot.py [Output of gppdf.py]
to generate plots like the header image.
This code will only work with Monte Carlo PDF Sets.