화학 Hamiltonian의 샘플 기반 양자 대각화

import math

import ffsim
import matplotlib.pyplot as plt
import numpy as np
import pyscf
import pyscf.cc
import pyscf.mcscf
from qiskit import QuantumCircuit, QuantumRegister
from qiskit.primitives import StatevectorSampler
from qiskit.providers.fake_provider import GenericBackendV2
from qiskit_ibm_runtime import QiskitRuntimeService
from qiskit_ibm_runtime import SamplerV2 as Sampler

# Specify molecule properties
spin_sq = 0

# Build N2 molecule
mol = pyscf.gto.Mole()
mol.build(
    atom=[["N", (0, 0, -0.5)], ["N", (0, 0, 0.5)]],
    basis="sto-6g",
    symmetry="Dooh",
)

# Define active space
n_frozen = 2
active_space = range(n_frozen, mol.nao_nr())

# Get molecular integrals
scf = pyscf.scf.RHF(mol).run()
norb = len(active_space)
n_electrons = int(sum(scf.mo_occ[active_space]))
n_alpha = (n_electrons + mol.spin) // 2
n_beta = (n_electrons - mol.spin) // 2
nelec = (n_alpha, n_beta)
cas = pyscf.mcscf.CASCI(scf, norb, nelec)
mo = cas.sort_mo(active_space, base=0)
hcore, nuclear_repulsion_energy = cas.get_h1cas(mo)
eri = pyscf.ao2mo.restore(1, cas.get_h2cas(mo), norb)

# Compute exact energy using FCI
reference_energy = cas.run().e_tot

print(f"norb = {norb}")
print(f"nelec = {nelec}")

# Get CCSD t2 amplitudes for initializing the ansatz
ccsd = pyscf.cc.CCSD(
    scf, frozen=[i for i in range(mol.nao_nr()) if i not in active_space]
).run()
t1 = ccsd.t1
t2 = ccsd.t2

import warnings

from qiskit.transpiler import CouplingMap

warnings.formatwarning = lambda msg, *args, **kwargs: f"Warning: {msg}\n"

# Set ansatz properties
n_reps = 4
pairs_aa = [(p, p + 1) for p in range(norb - 1)]

# Let generate_lucj_pass_manager determine the alpha-beta interactions
pairs_ab = None

# Initialize backend
coupling_map = CouplingMap.from_grid(norb, 2)
backend = GenericBackendV2(
    coupling_map.size(),
    coupling_map=coupling_map,
    basis_gates=["cp", "xx_plus_yy", "p", "x", "swap"],
)

# Create pass manager
pass_manager, pairs_ab = ffsim.qiskit.generate_lucj_pass_manager(
    backend=backend,
    norb=norb,
    connectivity="square",
    interaction_pairs=(pairs_aa, pairs_ab),
    optimization_level=3,
)

# Create the LUCJ ansatz operator
ucj_op = ffsim.UCJOpSpinBalanced.from_t_amplitudes(
    t2=t2,
    t1=t1,
    n_reps=n_reps,
    interaction_pairs=(pairs_aa, pairs_ab),
    # Setting optimize=True enables the "compressed" factorization
    optimize=True,
)

# create an empty quantum circuit
qubits = QuantumRegister(2 * norb, name="q")
circuit = QuantumCircuit(qubits)

# prepare Hartree-Fock state as the reference state and append it
# to the quantum circuit
circuit.append(ffsim.qiskit.PrepareHartreeFockJW(norb, nelec), qubits)

# apply the UCJ operator to the reference state
circuit.append(ffsim.qiskit.UCJOpSpinBalancedJW(ucj_op), qubits)
circuit.measure_all()

isa_circuit = pass_manager.run(circuit)
print(f"Gate counts: {isa_circuit.count_ops()}")

rng = np.random.default_rng(12345)
sampler = StatevectorSampler(seed=rng)
job = sampler.run([isa_circuit], shots=100_000)

primitive_result = job.result()
pub_result = primitive_result[0]

def is_valid_bitstring(
    bitstring: str, norb: int, nelec: tuple[int, int]
) -> bool:
    n_alpha, n_beta = nelec
    return (
        len(bitstring) == 2 * norb
        and bitstring[norb:].count("1") == n_alpha
        and bitstring[:norb].count("1") == n_beta
    )


bit_array = pub_result.data.meas
num_valid = sum(
    is_valid_bitstring(b, norb, nelec) for b in bit_array.get_bitstrings()
)
valid_fraction = num_valid / bit_array.num_shots
print(f"Fraction of sampled configurations that are valid: {valid_fraction}")

expected_fraction_random = (
    math.comb(norb, n_alpha) * math.comb(norb, n_beta) / 2 ** (2 * norb)
)
print(
    f"Expected fraction of valid configurations from uniformly random bitstrings: "
    f"{expected_fraction_random}"
)

from functools import partial

from qiskit_addon_sqd.fermion import (
    SCIResult,
    diagonalize_fermionic_hamiltonian,
    solve_sci_batch,
)

# SQD options
energy_tol = 1e-3
occupancies_tol = 1e-3
max_iterations = 5

# Eigenstate solver options
num_batches = 3
samples_per_batch = 1000
max_dim = 50
symmetrize_spin = True
carryover_threshold = 1e-4
max_cycle = 200

# Use the Hartree-Fock configuration as an initial guess for the orbital occupancies
initial_occupancies = (
    np.array([1] * n_alpha + [0] * (norb - n_alpha)),
    np.array([1] * n_beta + [0] * (norb - n_beta)),
)

# Pass options to the built-in eigensolver. If you just want to use the defaults,
# you can omit this step, in which case you would not specify the sci_solver argument
# in the call to diagonalize_fermionic_hamiltonian below.
sci_solver = partial(solve_sci_batch, spin_sq=0.0, max_cycle=max_cycle)

# List to capture intermediate results
result_history = []


def callback(results: list[SCIResult]):
    result_history.append(results)
    iteration = len(result_history)
    print(f"Iteration {iteration}")
    for i, result in enumerate(results):
        print(f"\tSubsample {i}")
        print(f"\t\tEnergy: {result.energy + nuclear_repulsion_energy}")
        print(
            f"\t\tSubspace dimension: {np.prod(result.sci_state.amplitudes.shape)}"
        )


result = diagonalize_fermionic_hamiltonian(
    hcore,
    eri,
    bit_array,
    samples_per_batch=samples_per_batch,
    norb=norb,
    nelec=nelec,
    num_batches=num_batches,
    energy_tol=energy_tol,
    occupancies_tol=occupancies_tol,
    max_iterations=max_iterations,
    sci_solver=sci_solver,
    symmetrize_spin=symmetrize_spin,
    max_dim=max_dim,
    initial_occupancies=initial_occupancies,
    carryover_threshold=carryover_threshold,
    callback=callback,
    seed=rng,
)

final_energy = result.energy + nuclear_repulsion_energy
energy_error = final_energy - reference_energy
print(f"Final energy: {final_energy}")
print(f"Final energy error: {energy_error}")

# Data for energies plot
x1 = range(len(result_history))
min_e = [
    min(result, key=lambda res: res.energy).energy + nuclear_repulsion_energy
    for result in result_history
]
e_diff = [abs(e - reference_energy) for e in min_e]

# Chemical accuracy (+/- 1.6 milli-Hartree)
chem_accuracy = 0.0016

# Data for avg spatial orbital occupancy
y2 = np.sum(result.orbital_occupancies, axis=0)
x2 = range(len(y2))

fig, axs = plt.subplots(1, 2, figsize=(12, 6))

# Plot energies
axs[0].plot(x1, e_diff, label="energy error", marker="o")
axs[0].set_xticks(x1)
axs[0].set_xticklabels(x1)
axs[0].set_yscale("log")
axs[0].set_ylim(1e-5, 1e-2)
axs[0].axhline(
    y=chem_accuracy,
    color="#BF5700",
    linestyle="--",
    label="chemical accuracy",
)
axs[0].set_title("Approximated Ground State Energy Error vs SQD Iterations")
axs[0].set_xlabel("Iteration Index", fontdict={"fontsize": 12})
axs[0].set_ylabel("Energy Error (Ha)", fontdict={"fontsize": 12})
axs[0].legend()

# Plot orbital occupancy
axs[1].bar(x2, y2, width=0.8)
axs[1].set_xticks(x2)
axs[1].set_xticklabels(x2)
axs[1].set_title("Avg Occupancy per Spatial Orbital")
axs[1].set_xlabel("Orbital Index", fontdict={"fontsize": 12})
axs[1].set_ylabel("Avg Occupancy", fontdict={"fontsize": 12})

plt.tight_layout()
plt.show()

# ------------------------------ Step 1 ------------------------------
# Build N2 molecule
mol = pyscf.gto.Mole()
mol.build(
    atom=[["N", (0, 0, -0.5)], ["N", (0, 0, 0.5)]],
    basis="cc-pvdz",
    symmetry="Dooh",
)

# Define active space
n_frozen = 2
active_space = range(n_frozen, mol.nao_nr())

# Get molecular integrals
scf = pyscf.scf.RHF(mol).run()
norb = len(active_space)
n_electrons = int(sum(scf.mo_occ[active_space]))
n_alpha = (n_electrons + mol.spin) // 2
n_beta = (n_electrons - mol.spin) // 2
nelec = (n_alpha, n_beta)
cas = pyscf.mcscf.CASCI(scf, norb, nelec)
mo = cas.sort_mo(active_space, base=0)
hcore, nuclear_repulsion_energy = cas.get_h1cas(mo)
eri = pyscf.ao2mo.restore(1, cas.get_h2cas(mo), norb)

# Store reference energy from SCI calculation performed separately
reference_energy = -109.22802921665716

print(f"norb = {norb}")
print(f"nelec = {nelec}")

# Get CCSD t2 amplitudes for initializing the ansatz
ccsd = pyscf.cc.CCSD(
    scf, frozen=[i for i in range(mol.nao_nr()) if i not in active_space]
).run()
t1 = ccsd.t1
t2 = ccsd.t2

# Set ansatz properties
n_reps = 1
pairs_aa = [(p, p + 1) for p in range(norb - 1)]

# Let generate_lucj_pass_manager determine the alpha-beta interactions
pairs_ab = None

# Initialize backend
service = QiskitRuntimeService()
backend = service.least_busy(operational=True, simulator=False)
print(f"Using backend {backend.name}")

# Create pass manager
connectivity_map = {"Heron": "heavy-hex", "Nighthawk": "square"}
pass_manager, pairs_ab = ffsim.qiskit.generate_lucj_pass_manager(
    backend=backend,
    norb=norb,
    connectivity=connectivity_map[
        backend.configuration().processor_type["family"]
    ],
    interaction_pairs=(pairs_aa, pairs_ab),
    optimization_level=3,
)

# Create the LUCJ ansatz operator
ucj_op = ffsim.UCJOpSpinBalanced.from_t_amplitudes(
    t2=t2,
    t1=t1,
    n_reps=n_reps,
    interaction_pairs=(pairs_aa, pairs_ab),
    # Setting optimize=True enables the "compressed" factorization
    optimize=True,
)

# create an empty quantum circuit
qubits = QuantumRegister(2 * norb, name="q")
circuit = QuantumCircuit(qubits)

# prepare Hartree-Fock state as the reference state and append it
# to the quantum circuit
circuit.append(ffsim.qiskit.PrepareHartreeFockJW(norb, nelec), qubits)

# apply the UCJ operator to the reference state
circuit.append(ffsim.qiskit.UCJOpSpinBalancedJW(ucj_op), qubits)
circuit.measure_all()


# ------------------------------ Step 2 ------------------------------

isa_circuit = pass_manager.run(circuit)
print(f"Gate counts: {isa_circuit.count_ops()}")


# ------------------------------ Step 3 ------------------------------
sampler = Sampler(mode=backend)
sampler.options.environment.job_tags = ["TUT_SQD"]
job = sampler.run([isa_circuit], shots=100_000)
primitive_result = job.result()
pub_result = primitive_result[0]


# ------------------------------ Step 4 ------------------------------

bit_array = pub_result.data.meas
num_valid = sum(
    is_valid_bitstring(b, norb, nelec) for b in bit_array.get_bitstrings()
)
valid_fraction = num_valid / bit_array.num_shots
print(f"Fraction of sampled configurations that are valid: {valid_fraction}")
expected_fraction_random = (
    math.comb(norb, n_alpha) * math.comb(norb, n_beta) / 2 ** (2 * norb)
)
print(
    f"Expected fraction of valid configurations from uniformly random bitstrings: "
    f"{expected_fraction_random}"
)
# SQD options
energy_tol = 1e-3
occupancies_tol = 1e-3
max_iterations = 5

# Eigenstate solver options
num_batches = 3
samples_per_batch = 1000
max_dim = 1000
symmetrize_spin = True
carryover_threshold = 1e-4
max_cycle = 200

# Use the Hartree-Fock configuration as an initial guess for the
# orbital occupancies
initial_occupancies = (
    np.array([1] * n_alpha + [0] * (norb - n_alpha)),
    np.array([1] * n_beta + [0] * (norb - n_beta)),
)

# Pass options to the built-in eigensolver. If you just want to use the defaults,
# you can omit this step, in which case you would not specify the
# sci_solver argument in the call to diagonalize_fermionic_hamiltonian below.
sci_solver = partial(solve_sci_batch, spin_sq=0.0, max_cycle=max_cycle)

# List to capture intermediate results
result_history = []


result = diagonalize_fermionic_hamiltonian(
    hcore,
    eri,
    bit_array,
    samples_per_batch=samples_per_batch,
    norb=norb,
    nelec=nelec,
    num_batches=num_batches,
    energy_tol=energy_tol,
    occupancies_tol=occupancies_tol,
    max_iterations=max_iterations,
    sci_solver=sci_solver,
    symmetrize_spin=symmetrize_spin,
    max_dim=max_dim,
    initial_occupancies=initial_occupancies,
    carryover_threshold=carryover_threshold,
    callback=callback,
    seed=rng,
)

final_energy = result.energy + nuclear_repulsion_energy
energy_error = final_energy - reference_energy
print(f"Final energy: {final_energy}")
print(f"Final energy error: {energy_error}")

# Data for energies plot
x1 = range(len(result_history))
min_e = [
    min(result, key=lambda res: res.energy).energy + nuclear_repulsion_energy
    for result in result_history
]
e_diff = [abs(e - reference_energy) for e in min_e]

# Chemical accuracy (+/- 1.6 milli-Hartree)
chem_accuracy = 0.0016

# Data for avg spatial orbital occupancy
y2 = np.sum(result.orbital_occupancies, axis=0)
x2 = range(len(y2))

fig, axs = plt.subplots(1, 2, figsize=(12, 6))

# Plot energies
axs[0].plot(x1, e_diff, label="energy error", marker="o")
axs[0].set_xticks(x1)
axs[0].set_xticklabels(x1)
axs[0].set_yscale("log")
axs[0].set_ylim(1e-3, 1e-1)
axs[0].axhline(
    y=chem_accuracy,
    color="#BF5700",
    linestyle="--",
    label="chemical accuracy",
)
axs[0].set_title("Approximated Ground State Energy Error vs SQD Iterations")
axs[0].set_xlabel("Iteration Index", fontdict={"fontsize": 12})
axs[0].set_ylabel("Energy Error (Ha)", fontdict={"fontsize": 12})
axs[0].legend()

# Plot orbital occupancy
axs[1].bar(x2, y2, width=0.8)
axs[1].set_xticks(x2)
axs[1].set_xticklabels(x2)
axs[1].set_title("Avg Occupancy per Spatial Orbital")
axs[1].set_xlabel("Orbital Index", fontdict={"fontsize": 12})
axs[1].set_ylabel("Avg Occupancy", fontdict={"fontsize": 12})

plt.tight_layout()
plt.show()