Diagonal Schedule Syndrome Extraction - Surface Code Experiments

This repository contains implementations comparing standard fixed-bulk conventions vs diagonal schedules for surface code syndrome extraction in two scenarios:

Files

Core Implementation

• diagonal_plaquettes.py: Custom DiagonalPlaquetteGenerator that provides diagonal schedule plaquettes (schedule [6,4,3,5] for Z and [1,4,3,2] for X). Handles both memory experiments and X-junction spatial cubes.

Comparison Scripts

1. Memory Experiment Comparison (complete_comparison.py)

Compares standard vs diagonal schedule memory experiments:

# Basic comparison (k=1,2,3, default noise levels)
./venv/bin/python complete_comparison.py

# Custom k values
./venv/bin/python complete_comparison.py --k-values 2 3 4

# Custom shots and noise levels
./venv/bin/python complete_comparison.py --shots 100000 --noise-levels 0.001 0.002 0.003

# Skip certain analyses
./venv/bin/python complete_comparison.py --skip-distance --skip-logical-error

2. X-Junction Comparison (compare_x_junction.py)

Compares standard vs diagonal schedule for spatial X-junctions:

# Basic comparison (distance, detector count, etc.)
./venv/bin/python compare_x_junction.py

# Include logical error rate analysis (requires pymatching)
./venv/bin/python compare_x_junction.py --error-rates

# Custom shots and noise levels
./venv/bin/python compare_x_junction.py --error-rates --shots 100000 --noise-levels 0.001 0.002 0.003

Outputs include:

• Graph-like distance for both circuits
• Number of qubits, detectors, and observables
• Crumble URLs for visualization
• Logical error rates and plot (when using --error-rates)

Key Features

• Same Detector Count: Both diagonal and standard circuits produce the same number of detectors, ensuring each ancilla connects to the same number of data qubits
• Corner Handling: Properly handles 3-body corner plaquettes when two adjacent arms are missing in spatial cubes
• Boundary Plaquettes: 2-body boundary plaquettes derived from diagonal bulk schedules with appropriate qubit omissions

Requirements

• Python 3.13+
• Dependencies in venv/ (activate with source venv/bin/activate)
• tqec library (installed in venv)