How dement-model/dementpy works
Main components
a. Core Simulation Engine (src/)
• dementpy.py (entry point / CLI)
• initialization.py (loads parameters & grid setup)
• grid.py (spatial grid management)
• microbe.py, enzyme.py, monomer.py, substrate.py (domain models)
• output.py, utility.py (I/O and helper functions)
b. Data Inputs (input/, grassland/)
• CSV files driving climate, enzyme traits, substrates, parameters, runtime settings, substrate-monitor inputs
c. Data Outputs (output/ or grassland/... runtime.txt / .pickle)
• Pickled simulation results, runtime logs
d. Post-processing & Analysis (SPA/)
• Shell scripts for batch submission (SGE/slurm)
• Python scripts for extracting / plotting results (data_extraction.py, output.py, drought_trait.py, etc.)
e. Infrastructure & Deployment
• pyproject.toml + uv for dependency management
• HPC integration: SGE job scripts and Slurm .sub files
Architectural patterns & design principles
• Modular layering: clear separation of domain models (microbe, enzyme, substrate), simulation orchestration (dementpy.py + initialization), and I/O utilities
• Pipeline pattern: data flows sequentially from input → simulation engine → output → post-processing
• Plug-in style extensibility: domain modules can be swapped or extended (e.g., add new microbial processes)
• HPC compatibility: batch job abstraction via shell scripts