Modeling Guides

Guides to modeling tasks that go beyond a single functor.

  • Manipulating Tables and Lookup Tables — functors and techniques for reading, iterating, restructuring, and computing statistics from tables and lookup tables, from a table with a known shape through to tables with an arbitrary number of columns or keys.
  • Working with Subregions Using Region Manager — when and why to split a model’s computation by region; defining regions from a categorical map with Region Manager; reading region metadata; visiting one region or looping over all of them; splitting a global map into regional pieces and merging them back; and the two supporting patterns — per-region functor parameters and per-region accumulation — a region-based model typically needs.
  • Weights of Evidence Calibration — calibrating Weights of Evidence from a pair of landscape maps and a set of predictive variables through to a finished set of ranges and coefficients: what Weights of Evidence is and how to read a weight’s sign, preparing categorical and continuous variables, the Skeleton and its parameters, determining ranges and why the underlying algorithm works, determining coefficients and reading the report, checking variable independence, saving and reusing the results, using the calibrated weights downstream, and a basic worked example.
  • Additional LUCC Resources — extensions to the land-use and land-cover change simulation model of the Guidebook: varying transition matrices, Weights of Evidence coefficients and Patcher and Expander parameters during a simulation, using subregions, sojourn time and deterministic transitions, and local saturation.
  • Map Metadata — writing a metadata file next to each saved map: template keywords, a template example, and an expanded metadata file.
  • Remote Execution — running parts of a model on other machines: setting up the coordinator and agents, configuring Dinamica EGO, the Run Remotely container, using several coordinators, and the limitations.
  • Handling Errors — capturing errors with Skip On Error and Skip All On Error, falling back to a matched set of values with junctions, testing for success only, and keeping partial results with conditional containers.
  • Useful Tips for Creating Models — practical tips that don’t fit neatly under a single functor or language feature.

Table of contents


This documentation was generated with the help of large language models (LLMs) and may contain errors. Please verify critical information.