Bug report/Feature Request: Allow lattice_dyson_g0_wk to take bosonic Matsubara frequencies

[andrewkhardy]

Description

In the current unstable version, lattice_dyson_g0_wk does not take bosonic meshes

Steps to Reproduce

from triqs.gf import *
from triqs.gf.tools import *
from triqs.operators import *
from triqs.gf.block_gf import *
from triqs_tprf.lattice import *
from triqs_tprf.lattice_utils import *
from triqs_tprf.tight_binding import TBLattice

params = dict()
params["hopping"] = 1.0
params["n_orbitals"] = 1
params["n_k"] = 32
params["beta"] = 10.0
params["w_max"] = 10
params["err"] = 1e-6
params["n_iw"] = 2049
params["stat"] = "Boson"#"Fermion"
gf_struct = []
for o in range(0, -params["n_orbitals"]):
    gf_struct.append("up-" + str(o))
    gf_struct.append("dn-" + str(o))

H = TBLattice(
    units = [(1, 0, 0), (0, 1, 0)],
    hopping = {
        # nearest neighbour hopping -t
        ( 0,+1): -params["hopping"] * np.eye(2*params["n_orbitals"]),
        ( 0,-1): -params["hopping"] * np.eye(2*params["n_orbitals"]),
        (+1, 0): -params["hopping"] * np.eye(2*params["n_orbitals"]),
        (-1, 0): -params["hopping"] * np.eye(2*params["n_orbitals"]),
        },
orbital_positions = [(0,0,0)]*2*params["n_orbitals"],
orbital_names = gf_struct
    )
kmesh = H.get_kmesh(n_k=(32, 32, 1))

e_k = H.fourier(kmesh)


#w_mesh = MeshDLRImFreq(beta=params["beta"], statistic=params["stat"], w_max=params["w_max"], eps=1e-12, symmetrize=True)
w_mesh = MeshImFreq(beta = params["beta"],  statistic = params["stat"], n_iw = params["n_iw"])
g0_wk = lattice_dyson_g0_wk(mu=0., e_k=e_k, mesh=w_mesh)
#chi00_wk = imtime_bubble_chi0_wk(g0_wk, nw=len(w_mesh), save_memory=False)

crashes for both a DLR and traditional mesh when using bosonic statistics. This is perhaps a feature request in that you could construct bosonic propagators through chi or some other means (usually where they originate from physically) but seeing as Green's functions in the rest of TRIQS are statistic agnostic, it would be nice for this to be extended to TPRF.