344 ML model and CLI improvements - Githubissues

GispoCoding / eis_toolkit

Python library for mineral prospectivity mapping

https://eis-he.eu/

European Union Public License 1.2

26 stars 8 forks source link

344 ML model and CLI improvements #345

Closed nmaarnio closed 8 months ago

nmaarnio commented 8 months ago

This PR concerns both CLI and ML tools.

CLI

Added enum getter/unpacker (was required for list enums to be interpreted correctly but applied to all to be sure)
All CLI functions with enum parameters use the above getter
Harmonized how enum parameters are defined for CLI functions (now always Annotated[EnumParameterType, typer.Option(case_sensitive=False) = EnumParameterType.some_default_value)

ML modeling

Added links to Sklearn and Keras docs where relevant
Added criterion parameter to Random Forest models
Added raster grid check to ML model data preparation and renamed training rasters -> feature rasters for clarity (since some call labels as training data)