Download this notebook (.ipynb + data files)
6. AutoHF Crash Course#
Goal: Learn the basics of using AutoHF to generate Hartree-Fock trial wavefunctions for SAFIRE.
6.1. What You Will Learn#
How to find the Hartree-Fock solution of a lattice model Hamiltonian with AutoHF
How to turn the AutoHF result into a SAFIRE trial wavefunction
How to check the charge and spin densities of the Hartree-Fock solution
6.2. Software Preliminaries#
For this tutorial, you need to install AutoHF (see Installation).
pip install sfqmc-autohf
6.3. Set Up the Model#
In this tutorial, we will perform Hartree-Fock calculations using AutoHF for the Hubbard model on a 4×4 lattice, which we learned to set up in Building Lattice Model Hamiltonians.
# create scratch directory in notebook path
%mkdir -p {_dh[0]}/data
%cd -q {_dh[0]}/data
from safiretools import Lattice, LatticeHamiltonian
from safiretools import vis
lattice = Lattice.from_type('square', (4, 4), boundaries=('periodic', 'periodic'), twists=(0.0, 0.0))
vis.plot_lattice(lattice)
hamiltonian = LatticeHamiltonian.from_lattice(
lattice,
t=1.0,
U=4.0,
)
6.4. Obtaining a Hartree-Fock Solution with AutoHF#
AutoHF takes two inputs: the Hamiltonian, converted from the safiretools format by AutoHFHamiltonian, and a dictionary of settings.
Here, nelec puts 5 spin-up and 5 spin-down electrons on the 16 sites, and ansatz='SD' together with noncollinear=True optimizes a single Slater determinant whose orbitals may mix spin up and spin down.
Because the optimization can get stuck in local minima, AutoHF runs batch_size independent trials from randomly perturbed starting points and keeps the one with the lowest energy.
from autohf import solve_hf, AutoHFHamiltonian
# convert SAFIRE Hamiltonian to the AutoHF Hamiltonian format
hamiltonian_autohf = AutoHFHamiltonian(hamiltonian)
# define settings
hf_settings = dict(
ansatz = 'SD',
steps = 100,
nelec = (5,5),
batch_size = 30,
seed = 42,
measure_spin = True,
random_initial_phi0 = False,
noncollinear = True,
verbose = False
)
results = solve_hf(
hamiltonian=hamiltonian_autohf,
settings=hf_settings
)
print("E_HF =", results[0]["E"])
You should get a Hartree-Fock energy of \(-17.75\,t\) (to within about \(10^{-8}\,t\)).
6.5. Turning the AutoHF Result into a SAFIRE Wavefunction#
import numpy as np
from safiretools import Wavefunction
# the noncollinear orbitals are spinors, spin-up components first
orbitals, = results[0]["orbitals"]
wavefunction = Wavefunction.from_single_determinant(
np.asarray(orbitals).reshape(2, lattice.N_sites, sum(hf_settings["nelec"]))
)
6.6. Analyzing the HF Wavefunction#
It is very important to analyze the HF solution before using it as a trial wavefunction to ensure that it has the desired properties. Here, we plot its charge and spin densities.
# get the density from autohf
state_results, functions = results
makeRDMs = functions['makeRDMs']
state = state_results['state']
rdm = makeRDMs(state)
# get lattice dimensions from the Lattice instance
L = lattice.L
# the noncollinear 1RDM has the spin-up block first
M = lattice.N_sites
rho_up = rdm[:M,:M].diagonal().reshape(*L)
rho_down = rdm[M:,M:].diagonal().reshape(*L)
rho_charge = rho_up + rho_down
rho_spin = rho_up - rho_down
vis.plot_lattice(
lattice,
density=rho_charge.real,
density_label="charge density",
vmin=0.0,
vmax=2.0
)
vis.plot_lattice(
lattice,
density=rho_spin.real,
density_label="spin density",
vmin=-0.5,
vmax=0.5
)
6.7. Summary#
In this tutorial, you learned the basics of using AutoHF to generate Hartree-Fock trial wavefunctions for SAFIRE.
6.7.1. What You Learned#
How to find the Hartree-Fock solution of a lattice model Hamiltonian with AutoHF
How to turn the AutoHF result into a SAFIRE trial wavefunction
How to check the charge and spin densities of the Hartree-Fock solution
6.8. Further Reading#
The examples in Lattice Models contain many more advanced applications of AutoHF, including models that show stripe ground states and how to apply pinning fields. The AutoHF documentation is a good resource to learn more about the detailed settings of AutoHF.