def test_initialization(self):
a = TensorNetwork()
a.add_node(1, [0], numpy.array([1, 2]))
a.add_node(2, [0], numpy.array([2, 3]))
a.open_edge(0)
a.open_edge(1)
a.open_edge(0)
self.assertEqual(a.shape, (2, None, 2))
a.add_node(3, [0, 1], numpy.array([[1, 2], [3, 4]]))
self.assertEqual(a.shape, (2, 2, 2))
def setUp(self):
self.a = TensorNetwork()
for i in range(5):
self.a.open_edge(i)
for i in range(5):
for j in range(5, 10):
self.a.add_node((i, j), [i, j], numpy.random.rand(2, 2))
self.b = TensorNetwork()
for i in range(5):
self.b.open_edge(i)
for i in range(5):
for j in range(5, 10):
self.b.add_node((i, j), [i, j], 1j * numpy.random.rand(2, 3))
self.c = Tensor(numpy.random.rand(2, 2, 2, 2, 2))
def test_copy(self):
a = TensorNetwork()
a.add_node(0, [0, 1], numpy.random.rand(10, 8))
a.add_node(1, [1, 2], numpy.random.rand(8, 9))
a.add_node(2, [2, 0], numpy.random.rand(9, 10))
b = a.copy()
c = deepcopy(a)
self.assertNotEqual(b.identifier, a.identifier)
self.assertNotEqual(c.identifier, a.identifier)
node = list(b.nodes_by_name)[0]
self.assertTrue(a.nodes[(0, node)]['tensor']._data is b.nodes[(0, node)]['tensor']._data)
self.assertFalse(a.nodes[(0, node)]['tensor']._data is c.nodes[(0, node)]['tensor']._data)
b.update_node(0, numpy.random.rand(10, 8))
self.assertNotEqual(a.contract(), b.contract())
def test_transpose(self):
a = TensorNetwork()
b = numpy.random.rand(3, 4, 5)
a.add_node(0, [0, 1, 2], b)
a.open_edge(0)
a.open_edge(1)
a.open_edge(2)
perm = (2, 0, 1)
self.assertTrue(numpy.allclose((a % perm).contract(),
numpy.transpose(b, perm)))
def setUp(self):
numpy.random.seed(517100123)
self.a = TensorNetwork()
self.open_edges = []
self.closed_edges = []
for i in range(10):
self.a.add_edge(i, bond_dim=2)
if numpy.random.randint(2):
self.a.open_edge(i)
self.open_edges.append(i)
else:
self.closed_edges.append(i)
self.bipart = numpy.random.randint(2, size=(10, 10))
for i in range(10):
edges = numpy.where(self.bipart[i])[0]
self.a.add_node(i,
edges,
numpy.random.rand(*([2] * len(edges))))
def test_shape(self):
a = TensorNetwork()
a.add_node(0, [0, 1], numpy.zeros((2, 3)))
self.assertEqual((), a.shape)
sp = []
for _ in range(1000):
o = numpy.random.randint(2)
a.open_edge(o)
sp.append(o + 2)
self.assertEqual(tuple(sp), a.shape)
def test_update_nodes(self):
peps = TensorNetwork()
for i in range(3):
for j in range(3):
peps.add_node((i, j), [(i, j, 'h'), ((i + 1) % 3, j, 'h'), (i, j, 'v'), (i, (j + 1) % 3, 'v'), (i, j, 'o')])
peps.open_edge((i, j, 'o'))
self.assertEqual(peps.shape, tuple([None] * 9))
diagonal = numpy.zeros((2, 2, 2, 2, 2))
diagonal[0, 0, 0, 0, 0] = 1
diagonal[1, 1, 1, 1, 1] = 1
ready_nodes = []
for node in peps.nodes_by_name:
self.assertTrue(peps.is_valid)
self.assertFalse(peps.is_ready)
ready_nodes.append(node)
peps.update_node(node, diagonal)
self.assertTrue(peps.is_ready)
self.assertTrue(peps.shape == tuple([2] * 9))
'T_1_inv': 1 / 30000.0,
'T_phi_inv': 1 / 60000.0,
'p_axis': 1e-4,
'p_plane': 5e-4,
'delta_phi': 0.01,
'T_g_1Q': 20.0,
'T_g_2Q': 40.0,
'tau_m': 300.0,
'tau_d': 300.0,
'gamma': 0,
'alpha0': 4,
'kappa': 1 / 250,
'chi': 1.3 * 1e-3
}
Z = TensorNetwork([0, 0])
Z.add_node('PH', [0], np.ones(2))
PlaceHolder = Measurement(1, Z, name='PH_nz')
TraceState = State(1, TensorNetwork([0, 0], bond_dim=2))
qubit_group_name = {
(0, 0): 'D1',
(0, 2): 'D2',
(2, 0): 'D3',
(2, 2): 'D4',
(1, 1): 'X1',
(1, 3): 'Z1',
(3, 1): 'Z2',
(3, 3): 'X2',
(-1, -1): 'dummy'
}
def test_fix_index(self):
a = TensorNetwork()
a.add_edge(1, bond_dim=2)
a.add_edge(2, bond_dim=2)
a.add_edge(3, bond_dim=2)
for i in range(10):
a.open_edge(1)
a.open_edge(2)
a.open_edge(3)
for i in range(5):
a.fix_index(0)
a.fix_index(10 - i)
a.fix_index(20 - 2 * i)
self.assertEqual(a.shape, tuple([2] * 15))
c = numpy.zeros((2, 2, 2, 2, 2))
c[0, 0, 0, 0, 0] = 1
data = numpy.einsum('ABCDE, FGHIJ, KLMNO -> KAFLBCGMDHINEJO', c, c, c)
self.assertTrue(numpy.allclose(a.contract(), data))
import numpy as np
import time
import math
from acqdp import circuit
from acqdp.tensor_network import ContractionScheme, TensorNetwork
from datetime import timedelta
import json
import argparse
import os
t = TensorNetwork(open_edges=[0, 1, 1, 0])
t.add_node(0, [0, 1], np.array([[1, 1j], [1j, np.exp(np.pi * 1j / 6)]]))
ISWAP_CZ = circuit.Unitary(2, t, "FSim", True)
def GRCS(f, in_state=0, simplify=False):
with open(f) as fin:
n = int(fin.readline())
c = circuit.Circuit()
if in_state is not None:
for qubit in range(n):
c.append(circuit.CompState[(in_state >> (n - qubit - 1)) & 1],
[qubit], -1)
gate_table = {
'h':
circuit.HGate,
'x_1_2':
circuit.Unitary(1,
np.array([[1 / np.sqrt(2), -1j / np.sqrt(2)],
[-1j / np.sqrt(2), 1 / np.sqrt(2)]]),
name='X_1_2'),
def test_basic_contraction(self):
a = TensorNetwork()
a.add_node(0, [0, 1], numpy.ones((2, 2)))
self.assertTrue(numpy.isclose(a.contract(), 4 + 0j))
a.add_node(1, [0, 1], numpy.ones((2, 2)))
self.assertTrue(numpy.isclose(a.contract(), 4 + 0j))
a.update_node(1, numpy.array([[70168, 52 * 1j], [65.77, -1e-3]]))
self.assertTrue(numpy.isclose(a.contract(), 70233.769 + 52 * 1j))
class TensorSumTestCase(unittest.TestCase):
def setUp(self):
self.a = TensorNetwork()
for i in range(5):
self.a.open_edge(i)
for i in range(5):
for j in range(5, 10):
self.a.add_node((i, j), [i, j], numpy.random.rand(2, 2))
self.b = TensorNetwork()
for i in range(5):
self.b.open_edge(i)
for i in range(5):
for j in range(5, 10):
self.b.add_node((i, j), [i, j], 1j * numpy.random.rand(2, 3))
self.c = Tensor(numpy.random.rand(2, 2, 2, 2, 2))
def test_addition(self):
c = self.a + ~(self.b) + 8 * self.c
self.assertEqual(type(c), TensorSum)
self.assertTrue(numpy.allclose(c.contract(),
self.a.contract()
+ numpy.conj(self.b.contract())
+ 8 * self.c.contract()))
def test_associativity(self):
c = (self.a + self.b) + self.c
d = self.a + (self.b + self.c)
e = (self.a + (self.c - self.b)) + (self.a + (2 * self.b - self.a))
self.assertTrue(numpy.allclose(c.contract(), d.contract()))
self.assertTrue(numpy.allclose(c.contract(), e.contract()))
def test_transpose(self):
axesA = (2, 4, 0, 1, 3)
axesB = (3, 4, 1, 0, 2)
axesC = (3, 0, 1, 2, 4)
axes_ = (1, 3, 0, 4, 2)
c = (self.a % axesA
+ self.b % axesB
+ self.c % axesC) % axes_
data = numpy.transpose(numpy.transpose(self.a.contract(), axesA)
+ numpy.transpose(self.b.contract(), axesB)
+ numpy.transpose(self.c.contract(), axesC), axes_)
self.assertTrue(numpy.allclose(c.contract(), data))
def test_add_and_remove_term(self):
c = TensorSum()
for i in range(100):
c.add_term(i, self.a % numpy.random.permutation(5))
choice = numpy.random.choice(100, 30, replace=False)
for j in choice:
c.remove_term(j)
with self.assertRaises(KeyError):
c.remove_term(choice[0])
with self.assertRaises(KeyError):
c.add_term(0, None)
c.add_term(0, None)
def test_shape_cache(self):
c = TensorSum()
misshaped = numpy.random.rand(2, 2, 3, 2, 2)
c.add_term(0, self.a)
c.add_term(1, self.a)
c.remove_term(0)
with self.assertRaises(ValueError):
c.add_term(2, misshaped)
c.add_term(2, self.a)
c.remove_term(1)
c.remove_term(2)
# Now that C is empty we can add any shape to it
c.add_term(0, misshaped)
c.add_term(1, misshaped)
with self.assertRaises(ValueError):
c.add_term(2, self.a)
# The function update_term does not fail early, so the user can update all tensors one-by-one
c.update_term(0, self.a)
c.update_term(1, self.a)
c.add_term(2, self.a)
def test_copy(self):
c = TensorSum()
a = Tensor(numpy.random.rand(2, 2, 2, 2, 2))
for i in range(2):
c.add_term(i, (a % numpy.random.permutation(5)).expand())
d = c.copy()
testing.assert_allclose(c.contract(), d.contract())
e = deepcopy(c)
testing.assert_allclose(c.contract(), e.contract())
d.update_term(0, numpy.random.rand(2, 2, 2, 2, 2))
self.assertFalse(numpy.allclose(c.contract(), d.contract()))
def test_fix_nested_edge_from_inside(self):
a = TensorNetwork()
a.add_edge(0, bond_dim=2)
a.open_edge(0)
a.open_edge(0)
b = TensorNetwork()
b.add_node(0, [0, 1], a)
b.open_edge(0)
b.open_edge(1)
c = TensorNetwork()
c.add_node(0, [0, 1], b)
c.open_edge(0)
c.open_edge(1)
a.fix_edge(0)
self.assertTrue(numpy.allclose(c.contract(), numpy.array([[1, 0], [0, 0]])))
class TensorNetworkGraphTestCase(unittest.TestCase):
def setUp(self):
numpy.random.seed(517100123)
self.a = TensorNetwork()
self.open_edges = []
self.closed_edges = []
for i in range(10):
self.a.add_edge(i, bond_dim=2)
if numpy.random.randint(2):
self.a.open_edge(i)
self.open_edges.append(i)
else:
self.closed_edges.append(i)
self.bipart = numpy.random.randint(2, size=(10, 10))
for i in range(10):
edges = numpy.where(self.bipart[i])[0]
self.a.add_node(i,
edges,
numpy.random.rand(*([2] * len(edges))))
def test_graph_properties(self):
self.assertTrue(set(self.a.open_edges) == set(self.open_edges))
self.assertTrue(self.a.closed_edges_by_name == set(self.closed_edges))
nodes_loc = list(numpy.where(numpy.random.randint(2, size=10))[0])
edges_from_nodes = set()
for node in nodes_loc:
edges_from_nodes |= set(numpy.where(self.bipart[node])[0])
self.assertTrue(self.a.edges_by_name_from_nodes_by_name(nodes_loc) == edges_from_nodes)
closed_edges_from_nodes = set()
for edge in edges_from_nodes:
if set(numpy.where(self.bipart[:, edge])[0]).issubset(nodes_loc):
if edge not in self.open_edges:
closed_edges_from_nodes.add(edge)
self.assertTrue(self.a.closed_edges_by_name_from_nodes_by_name(nodes_loc) == closed_edges_from_nodes)
edges_loc = list(numpy.where(numpy.random.randint(2, size=10))[0])
nodes_from_edges = set()
for edge in edges_loc:
nodes_from_edges |= set(numpy.where(self.bipart[:, edge])[0])
self.assertTrue(self.a.nodes_by_name_from_edges_by_name(edges_loc) == nodes_from_edges)
def test_graph_manipulation(self):
data = self.a.contract()
closed_edges = list(self.a.closed_edges_by_name)
self.a.merge_edges(closed_edges, "merge")
for edge in closed_edges:
self.assertTrue((edge not in self.a.edges_by_name) or (len(self.a.network.nodes[(1, edge)]) == 0))
for node in range(10):
edge_list = numpy.where(self.bipart[node])[0]
for i in range(len(edge_list)):
if self.a.network.nodes[(0, node)]['edges'][i] == 'merge':
self.a.rewire(node, i, edge_list[i])
self.assertTrue(numpy.allclose(self.a.contract(), data))
node = numpy.random.randint(10)
pop = self.a.pop_node(node)
self.a.add_node(node, pop['edges'], pop['tensor'])
self.assertTrue(numpy.allclose(self.a.contract(), data))
def test_khp(self):
data = self.a.contract()
data_khp = self.a.contract('khp')
self.assertTrue(numpy.allclose(data, data_khp))
def test_single_tensor(self):
a = TensorNetwork()
a.add_node('X', [0], numpy.array([1, 0]), is_open=True)
self.assertTrue(numpy.allclose(a.contract('khp'), [1, 0]))
"""
for q1 in high_freq_group:
if angles is not None:
if isinstance(angles, dict):
gate = Diagonal(2, np.array([1, 1, np.exp(1j * angles[q1]), -np.exp(-1j * angles[q1])]), name='NoisyCZ')
else:
gate = Diagonal(2, np.array([1, 1, 1, np.exp(1j * angles)]), name='NoisyCZ')
else:
gate = CZGate
for q2 in low_freq_group:
if abs(q1[0] - q2[0]) == 1 and abs(q1[1] - q2[1]) == 1:
circuit.append(gate, [q1, q2], time_step=time_step)
C = TensorNetwork([0, 0, 0, 0], bond_dim=2)
C.add_node('PH', [0], np.ones(2))
NDCompMeas = Channel(1, C, name='ND')
def add_noisy_surface_code(circuit, qubit_coords=None, connections=None, time=None, params=default_params):
if time is None:
time = max(circuit.max_time, 0)
qubit_group_name = {
(0, 0): 'D1',
(0, 2): 'D2',
(2, 0): 'D3',
(2, 2): 'D4',
(1, 1): 'X1',
(1, 3): 'Z1',
def test_mps(self):
c = TensorNetwork()
for i in range(10):
c.add_node(i, [i, (i, 'o'), (i + 1) % 10], numpy.ones((2, 3, 2)))
c.open_edge((i, 'o'))
self.assertEqual(c.shape, tuple([3] * 10))
d = TensorNetwork()
d.add_node('C', tensor=c)
d.add_node('~C', tensor=~c)
norm_sq = c.norm_squared
self.assertAlmostEqual(norm_sq, d.contract())
d.expand(['C'])
# d.raw_data = None
self.assertAlmostEqual(norm_sq, d.contract())
d.expand(['~C'])
# d.raw_data = None
self.assertAlmostEqual(norm_sq, d.contract())
d.encapsulate([('C', i) for i in range(10)])
# d.raw_data = None
self.assertAlmostEqual(norm_sq, d.contract())
d.encapsulate([('~C', i) for i in range(10)])
# d.raw_data = None
self.assertAlmostEqual(norm_sq, d.contract())
def lightcone(self, qubits, csp, num_layers):
"""Construct simplified tensor network corresponding to all QAOA circuit element acting non-trivially on a set
of qubits."""
qubits_set = set(qubits)
turns = {}
tn = TensorNetwork(dtype=complex)
for i in range(num_layers):
for qubit in qubits_set:
tn.add_node((2 * i + 1, qubit), [(i, qubit), (i + 1, qubit)], None)
tn.add_node((-2 * i - 1, qubit), [(-i, qubit), (-i - 1, qubit)], None)
clauses = []
new_set = set()
for clause in csp:
if set(clause).intersection(qubits_set):
clauses.append(clause)
new_set |= set(clause)
tn.add_node((2 * i + 2, clause), [(i + 1, q) for q in clause], None)
tn.add_node((-2 * i - 2, clause), [(-i - 1, q) for q in clause], None)
turns.update({qubit: i + 1 for qubit in new_set.difference(qubits_set)})
qubits_set |= new_set
for qubit in turns:
tn.merge_edges([(turns[qubit], qubit), (-turns[qubit], qubit)], merge_to=(0, qubit))
tn.update_dimension({e: 2 for e in tn.edges_by_name})
return tn, qubits_set
def test_matmul(self):
a = TensorNetwork()
res = numpy.eye(2)
a.open_edge(0)
for i in range(100):
random_mat = numpy.random.rand(2, 2)
a.add_node(i, [i, i + 1], random_mat)
if i > 0:
a.close_edge(i)
a.open_edge(i + 1)
res = res.dot(random_mat)
self.assertTrue(numpy.allclose(res, a.contract()))