Skip to content

Nicole 0.1.0

Choose a tag to compare

View all tags
@Phy-David-Zhang Phy-David-Zhang released this 26 Jan 00:08
· 605 commits to stable since this release
v0.1.0
6267e01

Nicole 0.1.0 - Initial Stable Release

Release Date: January 26, 2026

Nicole is a Python library for block-sparse tensor computations with Abelian symmetries, designed for tensor network algorithms in quantum many-body physics.

🎯 Core Features

Symmetry-Aware Tensor Framework

  • Automatic block-sparse structure for Abelian symmetries
  • Support for U(1) and Z(2) symmetry groups with automatic charge conservation
  • ProductGroup implementation for direct products of multiple Abelian symmetries
  • Index class with directional quantum number flow (IN/OUT)
  • Flexible index tagging system (itags) for intuitive tensor operations

Tensor Operations

  • Contraction: np.tensordot-style interface with automatic index pairing
    • Flexible axes specification with exclusion support
    • Automatic ambiguity detection and validation
    • Full support for scalar tensors (0D)
  • Trace: Automatic pairing with exclusion options
  • Arithmetic: Element-wise addition, subtraction with sector union
  • Manipulation: permute, transpose, conjugate, retag, merge_axes, flip
  • Block access: Efficient block extraction with getsub/subsector

Tensor Decomposition

  • SVD with symmetry preservation and truncation support
    • Bond dimension (chi) and singular value (tol) truncation
    • Automatic index direction handling
  • Eigenvalue decomposition for symmetric tensors
  • High-level decomp function with customizable truncation
  • Multi-axis decomposition support

Tensor Construction

  • identity: Identity tensors with automatic sector matching
  • isometry/isometry_n: Isometry for single and multi-index fusion
  • random: Random tensor generation with symmetry constraints
  • zeros: Zero tensors for initialization
  • oplus: Direct sum operation for selective axis merging

Specialized Operators

  • diag: Diagonal matrix tensor construction
  • inv: Tensor inversion with automatic index flipping
  • Operator arithmetic for quantum systems

Quantum Many-Body Systems (load_space)

Preset-based Hilbert space construction with physical operators:

  • Ferm: Spinless fermions with U(1) charge conservation
  • Band: Hardcore bosons with U(1)⊗U(1) or Z(2)⊗U(1) symmetries
  • Spin: Spin-1/2 chains with U(1) or Z(2) symmetries
  • Automatic operator generation (creation, annihilation, number, spin)
  • Vacuum sector support for open boundary conditions

📊 Statistics

  • 662 comprehensive tests covering all functionality
  • ~18,000 lines of code
  • 119 commits across 8 feature branches
  • Pure Python implementation with NumPy backend
  • Full type hints for IDE support

🎓 Target Users

Researchers and students in quantum many-body physics, condensed matter theory, and quantum information who work with tensor network methods (DMRG, TEBD, PEPS) and require efficient handling of Abelian symmetries.

📚 API Surface

Core: Tensor, Index, Direction
Symmetry: U1Group, Z2Group, ProductGroup, AbelianGroup
Operations: contract, trace, decomp, svd, eig
Manipulation: permute, transpose, conjugate, retag, merge_axes, flip
Construction: identity, isometry, isometry_n, random, zeros
Operators: diag, inv, oplus
Utilities: load_space, blocks, tensor_summary, index_summary, subsector

Assets 2
Loading