pktron 3.2.3

PKTron - Pakistan's 1st Quantum AI Powered Simulation Framework

PKTron v3.2.3

Pakistan's First Quantum Computing Framework — Top #1 in South Asia, Top #1 Asian, and Top #5 Globally by Features & Modules

pip install pktron
pip install pktron==3.2.3

What's New in v3.2.3 (8 Major Fixes)

1. Real GPU support — CuPy integration in StatevectorSimulator (silent fallback to NumPy)
2. Real Bravyi-Kitaev — Full Fenwick tree implementation, any n ≤ 16, no JW fallback
3. Real SABRE routing — Proper heuristic look-ahead, APSP distance matrix, δ=0.5 decay
4. Quantum BoltzmannMachine — Real quantum circuits (RY+RZZ), parameter-shift training
5. BeH2 Hamiltonian — Real STO-3G integrals, 64×64 Hermitian matrix, E₀ ≈ −15.58 Ha
6. QuantumGAN — Real adversarial gradient exchange via parameter-shift rule
7. HardwareBackend — mode='simulate'|'hardware', SDK stubs per backend, clear docstrings
8. Version 3.2.3 — __all__ with 72 exports, docstrings, consistent metadata

Full Feature Set (72 public classes/functions)

Core Infrastructure

• Gate — Dataclass for quantum gates (name, qubits, params, matrix)
• QuantumCircuit — 30+ gate builder with method chaining (H, X, Y, Z, S, T, RX/RY/RZ, U3, CX, CY, CZ, SWAP, TOFFOLI, FREDKIN, CRX/CRY/CRZ, RZZ, custom gates)
• StatevectorSimulator — GPU-accelerated (CuPy) up to ~25 qubits, silent NumPy fallback
• execute() — Convenience wrapper for the default simulator
• Pauli constants — I, X, Y, Z, H_gate, S, Sdg, T, Tdg, RX, RY, RZ, U3, CNOT, CZ, SWAP, TOFFOLI, FREDKIN

Simulators (9)

Simulator	Description
StatevectorSimulator	Exact statevector, GPU-ready via CuPy
DensityMatrixSimulator	Lindblad master equation with depolarizing, amplitude & phase damping
MPSSimulator	Matrix Product States via SVD, handles 100+ qubits
CliffordSimulator	Aaronson-Gottesman tableau, 1000+ qubits in polynomial time
QuantumTrajectorySimulator	Monte Carlo Wave Function (MCWF) for open systems
PEPSSimulator	2D lattice Projected Entangled Pair States
MERASimulator	Multi-Scale Entanglement Renormalization Ansatz
TensorNetworkSimulator	Greedy contraction ordering
PulseLevelSimulator	Trotter-Suzuki pulse-level simulation

Algorithms (15)

Algorithm	Description
VQE	Variational Quantum Eigensolver, L-BFGS-B optimizer
QAOA + qaoa_max_cut	Quantum Approximate Optimization, Max-Cut
GroverSearch	O(√N/M) oracle calls
Shor	Integer factorization
QuantumFourierTransform	Exact QFT with circuit builder
QuantumPhaseEstimation	n-ancilla phase estimation
HHLAlgorithm	Linear systems Ax=b
SimonsAlgorithm	Hidden period finding
DeutschJozsa	Constant vs balanced oracle
QuantumWalk	Discrete (coin-based) and continuous-time
QuantumAnnealing	Transverse-field Ising QUBO solver
AmplitudeAmplification	Generalized Grover
QuantumCounting	QPE-based solution counting
QSVM	Quantum kernel SVM with SLSQP

Quantum Error Correction (5)

Code	Description
Steane7QEC	[[7,1,3]] code, X/Z syndrome measurement
SurfaceCode	Distance-3, 17-qubit patch, MWPM decoder
BaconShorCode	[[9,1,3]] subsystem code, gauge operators
ColorCode	[[7,1,3]] triangular lattice, lookup-table correction
RepetitionCode	Bit-flip or phase-flip, majority vote

Error Mitigation (5)

Method	Description
ZeroNoiseExtrapolation	Gate folding + Richardson extrapolation
ProbabilisticErrorCancellation	Quasi-probability decomposition
CliffordDataRegression	Near-Clifford training + linear regression
DynamicalDecoupling	XY4, XY8, and UDD sequences
ReadoutErrorMitigation	Calibration matrix inversion

Quantum Chemistry (10 methods)

• Hamiltonians: H2, LiH, H2O, BeH2 (real STO-3G integrals, 64x64, E0 = -15.58 Ha)
• Transforms: Jordan-Wigner, Bravyi-Kitaev (Fenwick tree, n <= 16)
• Ansatze: UCC singles + doubles (Trotterized)
• Solvers: Full CI (exact diagonalization), qEOM (excitation energies), SSVQE (k excited states)

Quantum Machine Learning (8)

Class	Description
QuantumNeuralNetwork	Parameter-shift gradients, RY+CX architecture
QuantumGAN	Real adversarial gradient exchange via parameter-shift
QuantumAutoencoder	Encoder/decoder QNNs for state compression
QuantumCNN	Sliding-window parametrized filters
QuantumBoltzmannMachine	Real quantum circuits (RY+RZZ), parameter-shift training
QuantumFederatedLearning	FedAvg aggregation across QNN clients
QuantumReinforcementLearning	Quantum policy network, REINFORCE-style
QuantumTransferLearning	Classical feature extractor + variational quantum layer

Cryptography

• BB84Protocol — BB84, E91, B92 QKD protocols + QRNG
• PostQuantumCrypto — LWE lattice-based keygen, WOTS hash-based signatures

Hardware & Pulse Control

• DRAGPulse — Gaussian DRAG envelope, leakage suppression
• CrossResonancePulse — ZX interaction, superconducting CNOT
• HardwareBackend — IBM (Perth/Simulator), Google Weber, IonQ Harmony, AWS SV1
• SABRERouter — APSP distance matrix + heuristic look-ahead (delta=0.5), real SWAP insertion

Benchmarking

• QuantumBenchmarking — Quantum Volume, Randomized Benchmarking, XEB, CLOPS

Installation

pip install pktron                 # Core (NumPy + SciPy only)
pip install pktron[gpu]            # + CuPy GPU acceleration
pip install pktron[chemistry]      # + PySCF
pip install pktron[ml]             # + PyTorch + TensorFlow
pip install pktron[full]           # Everything

Quick Start

from pktron import QuantumCircuit, execute, VQE, QuantumChemistry

# Bell state
qc = QuantumCircuit(2).h(0).cx(0, 1)
result = execute(qc, shots=1000)
print(result['counts'])  # {'00': ~500, '11': ~500}

# VQE ground state of H2
H = QuantumChemistry.h2_hamiltonian(0.735)
energy = VQE(H).run(ansatz_depth=2)['energy']
print(f"H2 ground state: {energy:.4f} Ha")

License

MIT — Inquiries: [e-mail] GitHub: https://github.com/thecetqap/pktron