Official Resources
- Source Repository: https://github.com/CorentinB78/NCA
- License: MIT License
- Language: Python
Overview
This is a standalone implementation of the Non-Crossing Approximation (NCA) for solving quantum impurity problems, written in Python. Unlike the TRIQS-based version, this is a self-contained code, suitable for learning or specific lightweight applications without the full TRIQS dependency stack.
Scientific domain: Quantum Impurity Solvers, Many-body Physics
Target user community: Students, Researchers needing a light, pure-Python NCA solver
Theoretical Methods
- Non-Crossing Approximation (NCA)
- Resolvent Operator formalism
- Pseudo-particle Green's functions
- Integral equations for self-consistency
Capabilities
- Impurity Solver: Solves the Single Impurity Anderson Model (SIAM).
- Spectral Functions: Calculates impurity spectral functions $A(\omega)$.
- Finite Parameters: Handles finite interaction $U$, impurity level $\epsilon_d$, and finite temperatures.
- Hybridization: Accepts arbitrary hybridization functions $\Delta(\omega)$.
Key Strengths
Simplicity:
- Pure Python implementation effectively lowers the barrier to entry.
- Minimal dependencies (NumPy/SciPy).
Education:
- Clear code structure for understanding the integral equations of NCA.
Inputs & Outputs
- Input formats:
- Python script configuration:
U, ed, beta, Gamma (hybridization width/function).
- Output data types:
- Spectral densities (text files or arrays).
- Occupancies.
Interfaces & Ecosystem
- Standalone: Independent of heavy frameworks like TRIQS.
- Integration: Can be imported as a Python module.
Performance Characteristics
- Efficiency: NCA is computationally efficient compared to QMC, but slower than simple approximations like Hubbard-I.
- Cost: Low execution time for typical single-impurity problems.
Limitations & Known Constraints
- Approximation: NCA validity is limited (good for large N, high T, strong coupling; fails at low T Fermi liquid regime).
- Features: Lacks multi-orbital support and advanced features of production codes.
Comparison with Other Codes
- vs TRIQS-NCA: Lighter, standalone, easier to install.
- vs ED: Continuous bath support, but approximate.
Verification & Sources
Primary sources:
- GitHub: https://github.com/CorentinB78/NCA
Verification status: ✅ VERIFIED
- Source code: OPEN (Python)
- Method: NCA