Assessing the Impact of Fitting Methodology at aN$^3$LO with FPPDF: an Open Source Tool for Extracting Parton Distribution Functions in the Hessian Approach

We present a new public code, FPPDF, to perform global fits of parton distribution functions (PDFs). The fitting methodology follows that implemented by the MSHT collaboration, namely applying a fixed polynomial parameterisation of the PDFs and Hessian approach to error propagation, while for data and theory settings the libraries used by the NNPDF collaboration are taken. This therefore complements the already publicly available NNPDF fitting code to enable fits with both neural network and fixed polynomial PDF parameterisations to be performed by the community, with otherwise identical theoretical and experimental inputs. As a first application, we use the new code to compare the PDFs found from fits at both NNLO and aN$^3$LO perturbative orders, but applying these two fitting approaches. We assess the impact of the two different methodologies on the PDFs and their uncertainties, providing results that complement previous comparisons between published PDF sets at NNLO and aN$^3$LO. We in particular find that the relative impact of going to the higher perturbative order and/or including missing higher order uncertainties is rather insensitive to which of these PDF parameterisation methodologies are used.

💡 Research Summary

The paper introduces FPPDF, a new open‑source framework that implements the MSHT‑style fixed‑polynomial parameterisation of parton distribution functions (PDFs) together with a Hessian error‑propagation scheme, while re‑using the public NNPDF libraries for data handling and theory calculations. By doing so, the authors provide a tool that enables a direct, apples‑to‑apples comparison between the two dominant PDF fitting philosophies: the polynomial‑based Hessian approach (used by MSHT and CT) and the neural‑network based Monte‑Carlo replica approach (used by NNPDF).

The authors first describe the technical implementation. PDFs are parameterised at the input scale Q₀ using a Chebyshev polynomial basis identical to that employed in the MSHT20 fit:
 xf(x,Q₀²)=A(1−x)^η x^δ