Possible to have a view of a MomentumObject4D with pt, eta, phi, m ?

[matthewfeickert]

Somewhat motivated by poking on scikit-hep/fastjet#182 I've noticed that if I construct a MomentumObject4D object I'm able to access its .pt, .eta, .phi, and .m attributes. I'm see that there's also methods to convert coordinates like .to_rhophithetatau() to convert to $\rho$ - $\phi$ - $\theta$ - $\tau$ coordinates but there doesn't seem to be any way to convert to $p_{T}$ - $\eta$ - $\phi$ - $m$ (i.e. no .to_ptetaphim() API).

Is there another API to do this? If not, is this intentional and I'm missing something?

import awkward as ak
import fastjet
import vector

jetdef = fastjet.JetDefinition(fastjet.kt_algorithm, 0.6)

array = ak.Array(
    [
        {"px": 1.2, "py": 3.2, "pz": 5.4, "E": 2.5, "ex": 0.78},
        {"px": 32.2, "py": 64.21, "pz": 543.34, "E": 24.12, "ex": 0.35},
        {"px": 32.45, "py": 63.21, "pz": 543.14, "E": 24.56, "ex": 0.0},
    ],
)

cluster = fastjet.ClusterSequence(array, jetdef)
inclusive_jets = cluster.inclusive_jets()
print(f"{inclusive_jets=}")

four_vector = vector.obj(**inclusive_jets[0].tolist())
print(f"{four_vector=}")

# Have ability to get pt, eta, phi, m as attributes
print(
    f"(pt, eta, phi, m): {four_vector.pt, four_vector.eta, four_vector.phi, four_vector.m}"
)

print(f"{four_vector.to_rhophithetatau()=}")  # This exists
# print(f"{four_vector.to_ptetaphim()=}")  # This API doesn't exist so can't get something

# Nonsense numbers
# but this works with pt eta phi m keys
# c.f. https://github.com/scikit-hep/fastjet/issues/182
pt_eta_phi_m_record_array = ak.Array(
    [{"pt": 1.1, "eta": 2.1, "phi": 3.1, "m": 4.1, "ex": 0.78}],
    with_name="Momentum4D",
    behavior=vector.backends.awkward.behavior,
)

$ python example.py
#--------------------------------------------------------------------------
#                         FastJet release 3.4.0
#                 M. Cacciari, G.P. Salam and G. Soyez                  
#     A software package for jet finding and analysis at colliders      
#                           http://fastjet.fr                           
#	                                                                      
# Please cite EPJC72(2012)1896 [arXiv:1111.6097] if you use this package
# for scientific work and optionally PLB641(2006)57 [hep-ph/0512210].   
#                                                                       
# FastJet is provided without warranty under the GNU GPL v2 or higher.  
# It uses T. Chan's closest pair algorithm, S. Fortune's Voronoi code,
# CGAL and 3rd party plugin jet algorithms. See COPYING file for details.
#--------------------------------------------------------------------------
inclusive_jets=<Array [{px: 64.7, py: 127, pz: ..., ...}, ...] type='2 * Momentum4D[px: fl...'>
four_vector=MomentumObject4D(px=64.65, py=127.41999999999999, pz=1086.48, E=48.68)
(pt, eta, phi, m): (142.88274528437645, 2.726117171791057, 1.1012644074821902, -1094.7531808129172)
four_vector.to_rhophithetatau()=MomentumObject4D(pt=142.88274528437645, phi=1.1012644074821902, theta=0.1307594047706852, mass=-1094.7531808129172)

[Saransh-cpp]

vector automatically converts generic type coordinates to momentum type and vice versa when needed. So, for example -

In [1]: import vector

In [2]: v = vector.obj(px=1, py=2, pz=3, E=4)

In [3]: v
Out[3]: MomentumObject4D(px=1, py=2, pz=3, E=4)

In [4]: v.to_rhophietat()
Out[4]: MomentumObject4D(pt=2.23606797749979, phi=1.1071487177940904, eta=1.103586841560145, E=4)

In [5]: v.to_rhophietatau()
Out[5]: MomentumObject4D(pt=2.23606797749979, phi=1.1071487177940904, eta=1.103586841560145, mass=1.4142135623730951)

Similar to your example, vector understands that if a user calls to_rhophietatau or to_rhophietat on a MomentumVector, it must automatically use pt instead of phi and mass/E instead of tau/t.

The missing API is not an intentional behavior. In fact, there is an issue open for this - #168, but users can still accomplish the conversion without explicitly using the Momentum type methods.

This behavior is not limited to to_{coordinate_names} methods. For example, notice how I can (and must) provide generic coordinates in to_Vector{X}D constructors, but vector automatically converts them to momentum type coordinates if the operation is being performed on a momentum vector -

In [1]: import vector

In [2]: v = vector.obj(px=1, py=2)

In [3]: v
Out[3]: MomentumObject2D(px=1, py=2)

In [4]: v.to_Vector3D(z=4)
Out[4]: MomentumObject3D(px=1, py=2, pz=4)

In [5]: v.to_Vector4D(z=4, tau=5)
Out[5]: MomentumObject4D(px=1, py=2, pz=4, mass=5)