Skip to content

Why DisSModel?

Pythonic Modeling

Before DisSModel, many spatial modeling tools required learning domain-specific languages (DSL) or older stacks. We believe Spatial Modeling is Data Science. By building on Python, we offer:

  • Full Stack Integration: Use Scikit-Learn or PyTorch inside your model rules.
  • Active Community: Leverage the best geospatial libraries (GeoPandas, libpysal) directly.
  • Academic Rigor: Designed at UFMA to support high-level research in coastal dynamics and LUCC.

Key Advantages

1. Reproducibility First

Unlike many modeling tools where results are "volatile", DisSModel treats every run as an Experiment. With built-in SHA256 hashing and TOML snapshots, you can reproduce a simulation exactly as it was run years ago.

2. The Dual-Substrate Hybrid

You don't have to choose between GIS precision and speed. Use Vector for administrative boundaries and Raster for physical processes (like floods or fires) within the same environment.

3. Decoupled Architecture

Your scientific code (the Model) doesn't need to know if the data is on your SSD or in a MinIO bucket in the cloud. The io and executor modules handle the plumbing, so you can focus on the equations.

From Research to Production

DisSModel started as a thesis project and evolved into a production-ready framework. It is currently being used to model mangrove succession and coastal flooding, proving its capability in real-world, high-impact scenarios.