Official Resources
- Homepage: https://hande.readthedocs.io/
- Documentation: https://hande.readthedocs.io/
- Source Repository: https://github.com/hande-qmc/hande
- License: LGPL v2.1
Overview
HANDE is a modern, efficient implementation of FCIQMC (Full Configuration Interaction Quantum Monte Carlo) and related stochastic quantum chemistry methods. Developed as a community code with contributions from multiple institutions, HANDE provides production-quality implementations of FCIQMC, CCMC (Coupled Cluster Monte Carlo), and DMQMC (Density Matrix QMC) with emphasis on code quality, documentation, and ease of use. The code is designed for both research applications and method development.
Scientific domain: FCIQMC, stochastic quantum chemistry, many-body methods
Target user community: Quantum chemists, method developers, FCIQMC users
Theoretical Methods
- Full Configuration Interaction QMC (FCIQMC)
- Coupled Cluster Monte Carlo (CCMC)
- Density Matrix QMC (DMQMC)
- Initiator approximation
- Semi-stochastic adaptations
- Finite-temperature methods
- Excited states
- Exact diagonalization
Capabilities (CRITICAL)
Category: Open-source FCIQMC/stochastic quantum chemistry
- FCIQMC implementation
- CCMC variants
- DMQMC (finite-T)
- Semi-stochastic methods
- Initiator FCIQMC
- Molecules and lattices
- Ground and excited states
- Real and complex integrals
- Symmetry exploitation
- MPI + OpenMP parallelization
- HDF5 output
- Python interface (pyHANDE)
- Production quality
Sources: Official documentation, GitHub, publications
Key Strengths
Modern Implementation:
- Clean Fortran code
- Well-documented
- Unit tested
- Community-driven
- Best practices
Comprehensive Methods:
- FCIQMC, CCMC, DMQMC
- Multiple approaches
- Flexibility
- Research capabilities
- Production quality
Python Integration:
- pyHANDE interface
- Workflow automation
- Analysis tools
- Visualization
- User-friendly
Code Quality:
- Extensive testing
- Clear documentation
- Active development
- Issue tracking
- Community support
Inputs & Outputs
-
Input formats:
- Lua input scripts
- FCIDUMP integrals
- Model Hamiltonians
- Configuration files
-
Output data types:
- Energies and observables
- HDF5 archives
- Sampling data
- Analysis-ready formats
- pyHANDE objects
Interfaces & Ecosystem
Quantum Chemistry:
- FCIDUMP format
- Molpro
- PySCF
- GAMESS
- Dalton
Python Tools:
- pyHANDE analysis
- Jupyter workflows
- Visualization
- Data management
Workflow and Usage
Installation:
# Clone repository
git clone https://github.com/hande-qmc/hande.git
cd hande
mkdir build && cd build
cmake ..
make -j8
Lua Input (fciqmc.lua):
sys = hubbard_k {
electrons = 8,
lattice = { {4} },
ms = 0,
U = 1.3,
t = 1.0,
}
fciqmc {
sys = sys,
qmc = {
tau = 0.01,
rng_seed = 7,
init_pop = 10,
mc_cycles = 10,
nreports = 100,
target_population = 1e6,
state_size = -500,
spawned_state_size = -100,
},
}
Run HANDE:
# MPI + OpenMP
export OMP_NUM_THREADS=4
mpirun -n 4 hande.x fciqmc.lua > output.out
Python Analysis:
import pyhande
# Load HANDE output
data = pyhande.extract.extract_data('output.out')
# Analyze results
(data_len, reblock_data, covariance) = pyhande.analysis.reblock(data[0])
# Plot
pyhande.analysis.plot_reblocking(reblock_data)
Advanced Features
Semi-Stochastic:
- Deterministic space
- Stochastic remainder
- Improved efficiency
- Reduced noise
- Larger systems
Symmetry:
- Point group symmetries
- Momentum conservation
- Spin symmetries
- Computational efficiency
Finite Temperature:
- DMQMC method
- Thermal properties
- Temperature-dependent
- Phase transitions
- Statistical mechanics
Real-Time:
- Time-dependent methods
- Dynamics (development)
- Excited states
- Response properties
Performance Characteristics
- Speed: Efficient MPI+OpenMP
- Accuracy: Numerically exact (converged)
- Scalability: Good parallel scaling
- System size: Moderate (CI space)
- Purpose: Strongly correlated, benchmarks
Computational Cost
- Walker population dependent
- CI space scaling
- Expensive but exact
- HPC suitable
- Production capable
Limitations & Known Constraints
- CI space: Basis set limited
- Computational cost: Expensive
- Sign problem: Fermions
- System size: Moderate
- HPC recommended: Production calculations
Comparison with Other FCIQMC Codes
- vs NECI: HANDE similar methods, modern codebase
- vs Traditional FCI: HANDE stochastic, larger systems
- Unique strength: Modern implementation, Python integration, community code, documentation, code quality
Application Areas
Strongly Correlated Chemistry:
- Molecules
- Transition metals
- Multi-reference
- Bond breaking
- Strong correlation
Benchmarks:
- Exact energies
- Method validation
- Reference results
- Correlation energies
- Chemical accuracy
Lattice Models:
- Hubbard model
- t-J model
- Quantum magnetism
- Model systems
Method Development:
- FCIQMC research
- Algorithm innovation
- Stochastic methods
- Semi-stochastic
Best Practices
Input Scripts:
- Lua flexibility
- Clear syntax
- Modular approach
- Documentation
Analysis:
- Use pyHANDE
- Reblocking analysis
- Error estimation
- Statistical rigor
Production:
- Sufficient walkers
- Convergence testing
- Multiple runs
- HPC resources
Community and Support
- Open-source (LGPL v2.1)
- Multi-institutional
- GitHub repository
- Active development
- Comprehensive docs
- Community-driven
- Issue tracking
Educational Resources
- Excellent documentation
- Tutorials
- Example inputs
- pyHANDE guides
- FCIQMC literature
- API reference
Development
- Community collaboration
- Multiple institutions
- Active development
- Modern practices
- Regular releases
- User feedback
Research Impact
HANDE provides accessible, well-documented FCIQMC implementations, enabling researchers to apply cutting-edge stochastic quantum chemistry methods with confidence in code quality and results.
Verification & Sources
Primary sources:
- Homepage: https://hande.readthedocs.io/
- GitHub: https://github.com/hande-qmc/hande
- Publications: J. Chem. Theory Comput. 15, 3 (2019)
Secondary sources:
- FCIQMC literature
- User publications
- Quantum chemistry papers
Confidence: VERIFIED - Modern FCIQMC code
Verification status: ✅ VERIFIED
- Website: ACTIVE
- GitHub: ACCESSIBLE
- License: LGPL v2.1 (open-source)
- Category: Open-source FCIQMC code
- Status: Actively developed
- Community: Multi-institutional
- Specialized strength: Modern FCIQMC/CCMC/DMQMC implementation, excellent documentation, Python integration (pyHANDE), code quality emphasis, community-driven, production quality, semi-stochastic methods, Lua input, HDF5 output, comprehensive testing, user-friendly