Official Resources
- Repository: https://github.com/K4ys4r/WannierPy
- Related Project (wannierpy): https://github.com/henriquemiranda/wannierpy
- License: MIT License
- Developers: Community-driven (various variants exist).
Overview
WannierPy represents a collection of Python-based tools and scripts designed to facilitate the post-processing and analysis of Wannier90 output. While multiple forks and versions exist (e.g., K4ys4r/WannierPy, henriquemiranda/wannierpy), they share a common goal: providing a lightweight, scriptable interface to handle Hamiltonian matrices, plot band structures, and compute transport properties without the overhead of heavy compiled codes.
Scientific domain: Condensed matter physics, tight-binding analysis.
Target user community: Students and researchers needing quick, custom Python analysis of Wannier90 data.
Theoretical Methods
- Tight-Binding Interpolation:
- Fourier transform of real-space Hamiltonian $H(R)$ to reciprocal space $H(k)$.
- Band structure diagonalization.
- Transport:
- Calculation of basic conductivities (Boltzmann transport in relaxation time approximation).
- Carrier lifetime interpolation.
- Berry Phase:
- Evaluation of Berry curvature and anomalous Hall conductivity (in some variants).
Capabilities
- I/O Operations: Reads standard
_hr.dat, _centres.xyz, and .eig files.
- Band Structure:
- Fast interpolation along high-symmetry paths.
- Orbital-projected bands (fatbands).
- Visualization:
- Matplotlib-based plotting of bands and DOS.
- Visualization of Fermi surfaces.
- Topological Analysis: (Available in advanced forks) Calculation of Chern numbers and Berry curvature.
Key Strengths
- Lightweight: Pure Python/NumPy implementation; easy to install and modify.
- Educational: excellent resource for learning how Wannier interpolation works "under the hood".
- Flexibility: Users can easily add custom observables or analysis routines directly in Python.
Inputs & Outputs
- Inputs:
wannier90_hr.dat (Real-space Hamiltonian).wannier90.win (Input file for k-path parsing).
- Outputs:
- Matplotlib figures (PDF/PNG).
- Text files with interpolated eigenvalues.
Interfaces & Ecosystem
- Wannier90: The primary source of input data.
- NumPy/SciPy: Heavy reliance for matrix operations.
- Matplotlib: Used for all plotting functionalities.
Computational Cost
- Low: Extremely efficient for standard unit cells; cost scales with the number of Wannier functions ($N_W^3$) and k-points.
Comparison with Other Codes
- vs [WannierBerri](file:///home/niel/git/Indranil2020.github.io/scientific_tools_consolidated/TightBinding/4.1_Wannier_Ecosystem/WannierBerri.md): WannierBerri is a highly optimized, feature-rich production code for high-performance transport/optical calculations; WannierPy is a simpler, lightweight toolkit for basic analysis and quick plotting.
- vs [PyWannier90](file:///home/niel/git/Indranil2020.github.io/scientific_tools_consolidated/TightBinding/4.1_Wannier_Ecosystem/PyWannier90.md): PyWannier90 often refers to the official interface; WannierPy usually denotes community scripts for post-processing.
Application Areas
- Quick Plotting: Rapidly checking band structures after a Wannier90 run.
- Teaching: Demonstrating TB interpolation concepts.
- Custom Analysis: Prototyping new observables before implementing them in Fortran/C++.
Verification & Sources