mobiu-q 1.0.3

Soft Algebra Optimizer for Quantum & Complex Optimization

Mobiu-Q

Mobiu-Q is a next-generation optimizer built on Soft Algebra and Demeasurement theory, enabling stable and efficient optimization in quantum variational algorithms (VQE, QAOA).

✨ Features

• Soft Algebra update rule for stable convergence
• New: SPSA Demeasurement for noisy hardware
• New: Adaptive "Noisy Mode" presets

---

📦 Installation

pip install mobiu-q

🚀 Quick Start (Simulation)

Best for clean simulations or statevector backends.

import numpy as np
from mobiu_q import MobiuQCore, Demeasurement

# Initialize in standard mode (aggressive learning)
opt = MobiuQCore(mode="standard") 

params = np.random.uniform(-np.pi, np.pi, 4)

for step in range(50):
    # 1. Calculate Energy
    E = energy_fn(params)
    
    # 2. Exact Gradient (Finite Difference)
    grad = Demeasurement.finite_difference(energy_fn, params)
    
    # 3. Step
    params = opt.step(params, grad, E)

⚡ Noise Resilience (Real Hardware)

When running on noisy quantum computers (IBM, IonQ) or noisy simulators, use SPSA and noisy mode.

# Initialize in noisy mode (robust learning, lower LR)
opt = MobiuQCore(mode="noisy") 

for step in range(100):
    # Demeasurement.spsa returns BOTH gradient and estimated energy
    # This saves 50% of measurement cost compared to separate calls.
    grad, estimated_energy = Demeasurement.spsa(energy_fn, params, c_shift=0.1)
    
    params = opt.step(params, grad, estimated_energy)

📚 Built-in Problems

from mobiu_q import list_problems, get_energy_function

print(list_problems())
# ['h2_molecule', 'lih_molecule', 'transverse_ising', ...]

energy_fn = get_energy_function("lih_molecule")

⚙️ API Reference

MobiuQCore

MobiuQCore(
    license_key=None,       # Your license key (or use MOBIU_Q_LICENSE_KEY env var)
    mode="standard",        # 'standard' (0.05 LR) or 'noisy' (0.02 LR)
    base_lr=None,           # Custom learning rate (overrides mode default)
    verbose=True            # Print session status messages
)

💡 Choosing the Right Mode

Mode	Learning Rate	Use When
standard	0.05	Simulators (Qiskit Aer, PennyLane, Cirq)
noisy	0.02	Real quantum hardware (IBM, IonQ, Rigetti)

💡 Important Note on Modes

The mode parameter ("standard" / "noisy") only configures the optimizer's internal stability (Learning Rate, Soft Algebra decay).

• It does not automatically calculate gradients.
• For noisy environments, you must explicitly call Demeasurement.spsa(...) in your loop to generate robust gradients.

Demeasurement

• Demeasurement.finite_difference(fn, params): Accurate, requires 2*N measurements.
• Demeasurement.spsa(fn, params): Noisy-resistant, requires only 2 measurements total.

🔥 Full Demo: Clean vs. Noisy

See how Mobiu-Q adapts to different environments. This script compares:

1. Clean Simulation: Using mode="standard" and Finite Difference gradients.
2. Noisy Simulation: Using mode="noisy" and SPSA (robust against shot noise).

import numpy as np
from mobiu_q import MobiuQCore, Demeasurement, get_energy_function

# Load problem
energy_fn = get_energy_function("h2_molecule")

# --- 1. Clean Run ---
opt_clean = MobiuQCore(mode="standard")
params = np.random.uniform(-0.5, 0.5, 4)

# Run standard optimization loop...

# --- 2. Noisy Run (Simulated) ---
def noisy_fn(p):
    return energy_fn(p) + np.random.normal(0, 0.05) # Add noise

opt_noisy = MobiuQCore(mode="noisy") # Auto-tunes LR for noise

for step in range(100):
    # Use SPSA for robust gradient estimation (only 2 measurements!)
    grad, est_energy = Demeasurement.spsa(noisy_fn, params)
    params = opt_noisy.step(params, grad, est_energy)

© Mobiu Technologies, 2025