BerryEasy

Official Resources

• arXiv: https://arxiv.org/abs/2312.13051
• Publication: A. C. Tyner, arXiv:2312.13051 (2023)
• License: Check repository

Overview

BerryEasy is a GPU-enabled Python package for diagnosis of nth-order and spin-resolved topology in the presence of fields and effects. It provides efficient computation of nested Wilson loops, spin-resolved Wilson loops, and various topological invariants using GPU acceleration for performance.

Scientific domain: Topological band theory, higher-order topology, spin-resolved invariants Target user community: Researchers studying advanced topological phases and spin-resolved topology

Theoretical Methods

• Wilson loop calculations
• Nested Wilson loops for higher-order topology
• Spin-resolved Wilson loops
• Berry phase and Berry curvature
• Wannier charge center evolution
• Symmetry-resolved invariants

Capabilities (CRITICAL)

• GPU Acceleration: CUDA-enabled computations
• Nested Wilson Loops: Higher-order topology diagnosis
• Spin-Resolved: Spin-resolved topological invariants
• nth-Order Topology: Beyond first-order invariants
• Field Effects: External field incorporation
• Multiple Models: Tight-binding and Wannier support

Sources: arXiv preprint, code repository

Key Strengths

GPU Performance:

• CUDA acceleration
• Parallel computations
• Large system handling
• Fast invariant calculation

Advanced Topology:

• Higher-order topological phases
• Spin-resolved invariants
• Nested Wilson loop spectra
• Partial polarization

Comprehensive:

• Multiple invariant types
• Field effect support
• Model flexibility
• Modern implementation

Inputs & Outputs

• Input formats:

  • Tight-binding models
  • Wannier Hamiltonians
  • PythTB-compatible models

• Output data types:

  • Wilson loop spectra
  • Topological invariants
  • Berry curvature
  • Wannier centers

Installation

pip install berryeasy
# Or from source
git clone [repository]
pip install -e .

Usage Examples

import berryeasy as be

# Load tight-binding model
model = be.load_model("wannier90_hr.dat")

# Calculate nested Wilson loop
nwl = be.nested_wilson_loop(model, direction=[0, 1])

# Spin-resolved Wilson loop
swl = be.spin_resolved_wilson(model, spin_operator="Sz")

# Get invariants
z2 = be.calculate_z2(model)

Performance Characteristics

• Speed: GPU-accelerated, significant speedup
• Memory: GPU memory dependent
• Scalability: Handles large models efficiently

Limitations & Known Constraints

• GPU required: Best performance with CUDA GPU
• Preprint stage: Not yet peer-reviewed publication
• Dependencies: Requires GPU libraries

Comparison with Other Tools

• vs PythTB: BerryEasy GPU-accelerated, more invariants
• vs Z2Pack: BerryEasy includes spin-resolved
• vs nested_wloop: BerryEasy has GPU support
• Unique strength: GPU acceleration, spin-resolved nth-order topology

Application Areas

• Higher-order topological insulators
• Spin-Hall insulators
• Axion insulators
• Magnetic topological phases
• Fragile topology

Best Practices

• Use GPU for large calculations
• Verify convergence with k-point density
• Check spin operator definitions
• Compare with known results

Community and Support

• arXiv preprint available
• Academic development

Verification & Sources

Primary sources:

1. arXiv: https://arxiv.org/abs/2312.13051
2. A. C. Tyner, arXiv:2312.13051 (2023)

Confidence: VERIFIED - arXiv preprint

Verification status: ✅ VERIFIED

• arXiv preprint: ACCESSIBLE
• Method: GPU-enabled topological analysis
• Developer: Academic research