Many next-to-leading order QCD predictions are available through Monte Carlo (MC) simulations.
Usually, multiple CPU hours are needed to calculate predictions at a required precision, which
is unfeasible for global PDF analyses. This problem is solved by a process known as "gridding":
The values of the hard-scattering cross-section are calculated only once with the MC program,
and then interpolated and stored in look-up tables (grids) of the kinematical variables. To obtain
the physical predictions, they are convolved with the PDFs (e.g. during the fitting stage in a
PDF global analysis), which takes a tiny fraction of the time needed to calculate the MC results.
This is possible with PineAPPL, a library tackling the aforementioned process of grid creation
and convolution. In this work, we use PineAPPL to grid the predictions for open heavy-flavor
production in the general-mass variable-flavor-number scheme (GM-VFNS). In the GM-VFNS,
the differential cross-section interpolates between the fixed-flavor-number scheme (FFNS) and
the zero-mass variable-flavor-number scheme (ZM-VFNS). These are each only valid in different
kinematical regions, in which the GM-VFNS cross-section reproduces the FFNS and ZM-VFNS as
the limiting cases of high energies and small masses, respectively. Better than permille agreement
is achieved between the grids and the MC predictions, while at the same time not substantially
increasing the time of the MC calculations.