qutip-sampler 0.1.0

Quantum annealing sampler using QuTiP sesolve

qutip-sampler

A dimod-compatible quantum annealing sampler backed by QuTiP sesolve.

It maps an Ising / QUBO problem onto a transverse-field Ising Hamiltonian, evolves the ground state via a linear annealing schedule, and draws bitstring samples from the final quantum state.

Installation

pip install qutip-sampler

Quick start

import dimod
from qutip_sampler import QuTipSampler

# SPIN (Ising) problem
bqm = dimod.BinaryQuadraticModel({'a': -1.0, 'b': -1.0}, {('a', 'b'): 0.5}, 0.0, 'SPIN')
result = QuTipSampler().sample(bqm, num_reads=100, seed=42)
print(result.first.sample)   # e.g. {'a': 1, 'b': 1}

# Ising directly
result = QuTipSampler().sample_ising({'a': -1.0}, {('a', 'b'): 0.5}, num_reads=50)

# QUBO
Q = {('x', 'x'): -1.0, ('y', 'y'): -1.0, ('x', 'y'): 0.5}
result = QuTipSampler().sample_qubo(Q, num_reads=50, seed=0)
print(result.first.sample)   # e.g. {'x': 1, 'y': 1}

Parameters

Parameter	Default	Description
anneal_time	10.0	Total annealing time T passed to sesolve
n_steps	200	Number of time steps in the annealing schedule
num_reads	100	Number of bitstring samples drawn from the final state
seed	None	RNG seed for reproducible sampling

anneal_time and n_steps can be set at construction time or overridden per call.

How it works

1. Collect all variables from the problem and sort them.
2. Build the Ising Hamiltonian H_ising = Σ hᵢ σᵢᶻ + Σ Jᵢⱼ σᵢᶻ σⱼᶻ.
3. Build the transverse-field Hamiltonian H_T = −Σ σᵢˣ, whose ground state is |+⟩⊗ⁿ.
4. Evolve H(t) = (1 − t/T) H_T + (t/T) H_ising from t=0 to t=T using QuTiP sesolve.
5. Sample bitstrings from the probability distribution |ψ(T)|².

License

MIT