稀疏态演化示例

本教程展示 SparseState 如何随算子操作演化，并进一步演示 QFT 和 QRAM 等组合算子的使用。

import pysparq as ps

ps.System.clear()

# Create a 2-bit register
ps.System.add_register("q", ps.UnsignedInteger, 2)

# Initial state
state = ps.SparseState()

print("Initial state:")
ps.pprint(state)
print(f"\nNumber of basis states: {state.size()}")

Hadamard 变换

Hadamard 创建量子叠加态。

# Hadamard_Int: apply Hadamard to a specified number of qubits
ps.Hadamard_Int("q", 1)(state)

print("After Hadamard_Int(q, 1):")
ps.pprint(state)
print(f"\nNumber of basis states: {state.size()}")

# Hadamard_Int_Full: apply a full Hadamard to all qubits
ps.Hadamard_Int_Full("q")(state)

print("After Hadamard_Int_Full(q):")
ps.pprint(state)
print(f"\nNumber of basis states: {state.size()}")

状态打印模式

StatePrint 支持多种显示模式。

print("Default mode:")
print(ps.StatePrint(state, mode=ps.StatePrintDisplay.Default))

print("Binary mode:")
ps.pprint(state, mode=ps.StatePrintDisplay.Binary)

print("Prob mode:")
ps.pprint(state, mode=ps.StatePrintDisplay.Prob)

算术操作演化

ps.System.clear()

# Create registers
ps.System.add_register("a", ps.UnsignedInteger, 2)
ps.System.add_register("b", ps.UnsignedInteger, 2)
ps.System.add_register("sum", ps.UnsignedInteger, 2)

state = ps.SparseState()

# Initialize
ps.Init_Unsafe("a", 1)(state)
ps.Init_Unsafe("b", 2)(state)

print("Initial state:")
ps.pprint(state)

# Addition: sum ^= a + b
ps.Add_UInt_UInt("a", "b", "sum")(state)

print("After Add_UInt_UInt:")
ps.pprint(state)
# sum = 0 ^ (1 + 2) = 3

# Applying it again undoes it (XOR mechanism)
ps.Add_UInt_UInt("a", "b", "sum")(state)

print("After applying Add_UInt_UInt again:")
ps.pprint(state)
# sum = 3 ^ 3 = 0

叠加态上的算术

ps.System.clear()

ps.System.add_register("x", ps.UnsignedInteger, 2)
ps.System.add_register("y", ps.UnsignedInteger, 2)

state = ps.SparseState()

# x is in superposition
ps.Hadamard_Int_Full("x")(state)

# y is initialized to a constant
ps.Init_Unsafe("y", 1)(state)

print("Initial superposition:")
ps.pprint(state)

# In-place addition: y += x (each branch computed independently)
ps.Add_UInt_UInt_InPlace("x", "y")(state)

print("After Add_UInt_UInt_InPlace:")
ps.pprint(state)
# Each branch: y = 1 + x

# Undo
ps.Add_UInt_UInt_InPlace("x", "y").dag(state)

print("After dagger:")
ps.pprint(state)

访问基态数据

# Iterate over all basis states
print("Iterating over basis states:")
for i, system in enumerate(state.basis_states):
    x_id = ps.System.get_id("x")
    y_id = ps.System.get_id("y")

    x_val = system.get(x_id).value
    y_val = system.get(y_id).value
    amp = system.amplitude

    print(f"  basis state {i}: x={x_val}, y={y_val}, amplitude={amp}")

QRAM 数据加载

在叠加态下通过 QRAM 批量加载经典数据：

import numpy as np

ps.System.clear()
n_addr, n_data = 3, 4
ps.System.add_register("addr", ps.UnsignedInteger, n_addr)
ps.System.add_register("data", ps.UnsignedInteger, n_data)

state = ps.SparseState()

# Classical data (8 memory locations)
memory = np.array([1, 3, 5, 7, 2, 4, 6, 8], dtype=np.uint64)

# Superpose the address register → query all addresses at once
ps.Hadamard_Int("addr")(state)

# QRAM load: data = memory[addr]
qram = ps.QRAMCircuit_qutrit(n_addr, n_data, memory)
ps.QRAMLoad(qram, "addr", "data")(state)

# The state holds the amplitudes of all (addr, memory[addr]) pairs
print(ps.StatePrint()(state))
print(f"\nNumber of basis states: {state.size()}")

QFT 变换

ps.System.clear()
ps.System.add_register("reg", ps.UnsignedInteger, 3)
state = ps.SparseState()

ps.Init_Unsafe("reg", 1)(state)
print("Initial:")
print(ps.StatePrint()(state))

ps.QFT("reg")(state)
print("After QFT:")
print(ps.StatePrint()(state))

ps.InverseQFT("reg")(state)
print("After inverse QFT:")
print(ps.StatePrint()(state))  # restored to |1⟩

总结

• SparseState 只存储非零基态，基态数量随叠加增长

• Hadamard 创建叠加态

• 算术操作作用于每个基态（稀疏遍历）

• SelfAdjointOperator 两次应用恢复原值，BaseOperator 用 dag() 撤销

• QRAM 在叠加态下实现经典数据的批量查询

• QFT / InverseQFT 用于相位估计类算法