PartonDensity¶

class madspace.PartonDensity(*args, **kwargs)¶

Bases: FunctionGenerator

Parton-density interpolation as a compute-graph function.

Interpolates a PdfGrid at the sampled \((x, q)\), returning \(x f(x, q)\) for the requested flavours; the numerical output agrees with LHAPDF to double precision [1] (Sec. 3.2.3 of [2]). With a CustomPdf, the user-supplied library is called once per requested flavour instead.

batch is the leading batch dimension.

Arguments

x – float, shape (batch,) – momentum fraction.
q – float, shape (batch,) – factorization scale.
flavor_index – int, shape (batch,) – which flavour to return. Present only when dynamic_pid is true.

Returns

pdf – float, shape (batch,) when dynamic_pid is true, otherwise (batch, len(pids)) – the interpolated x f(x, q).

References

[1] A. Buckley et al., “LHAPDF6: parton density access in the LHC precision era”, https://arxiv.org/abs/1412.7420
[2] T. Heimel, O. Mattelaer, R. Winterhalder, “MadSpace”, https://arxiv.org/abs/2602.06895 (Sec. 3.2.3)

Overloaded function.

__init__(self: madspace._madspace_py.PartonDensity, grid: madspace._madspace_py.PdfGrid, pids: collections.abc.Sequence[typing.SupportsInt], dynamic_pid: bool = False, prefix: str = ‘’) -> None

Parameters:

grid – The parton-density grid to interpolate.
pids – PDG ids of the flavours to return.
dynamic_pid – If true, a single flavour selected per event by the flavor_index argument is returned.
prefix – Prefix for the grid global names.

__init__(self: madspace._madspace_py.PartonDensity, custom_pdf: madspace._madspace_py.CustomPdf, pids: collections.abc.Sequence[typing.SupportsInt], beam_index: typing.SupportsInt, dynamic_pid: bool = False) -> None

Parameters:

custom_pdf – The user-supplied parton density.
pids – PDG ids of the flavours to return.
beam_index – Index of the beam (0 or 1), passed to the library.
dynamic_pid – If true, a single flavour selected per event by the flavor_index argument is returned.