Reference guide for ecoClassify SyncroSim Package
In SyncroSim, all inputs and outputs associated with a model are stored in a single file (with the extension .ssim), referred to as a SyncroSim Library. The inputs and outputs contained within a SyncroSim Library are organized into Datafeeds (or tabs in the SyncroSim UI). Each Datafeed can be made up of one or more tables of data, called Datasheets. Each Datasheet is associated with one of three scopes: Library, Project or Scenario. See the SyncroSim documentation for more details.
ecoClassify has one Project scoped Datasheet and multiple Scenario scoped Datasheets. The Scenario Datasheets are used by the package’s four pipeline stages:
- 1-Train Classifier trains a Random Forest, MaxEnt or CNN model from training rasters and user-classified rasters.
- 2a-Prepare Spatial Tiling (optional) splits large predicting rasters into tiles so prediction can run in parallel.
- 2-Predict applies the trained model to new rasters.
- 3-Post-Process filters the binary predictions and applies rule-based restrictions.
Within this Reference guide, you will find details on the function of each field within ecoClassify Datasheets. For details on Datasheets not covered in this reference guide, such as Multiprocessing, see the SyncroSim Core Datasheets Overview.