Mixins (overview)

Much of SimBA’s analysis logic lives in mixin classes — reusable libraries of static/instance methods that the pipelines, GUI and CLI all draw on. Rather than call the pipelines, you can import a mixin directly and use its methods on your own arrays and DataFrames.

Each mixin is documented in the topical section that matches what it does. Use this page as the map; follow a link for the full method reference.

Mixin	Documented in
Geometry — convert body-parts into polygons/circles/lines and compute spatial relationships	Geometry transformations
Statistics — sliding/static-window statistics, distances, drift and distribution tests	Statistics transformations
Circular statistics — wraparound-aware angle/heading statistics	Circular transformations
Feature extraction — low-level feature primitives used by the extraction pipelines	Feature extraction mixins
Time-series — time-series complexity and windowed descriptors	Time-series transformations
Image — frame slicing and visual-feature extraction from tracking data	Image transformations
Network — build and analyse graphs from pose time-series	Network transformations
Plotting — shared plotting/visualization helpers	Plotting and visualization tools
Model / training — train, grid-search and run inference with classifiers	Model tools
Config reader — parse SimBA project config and metadata	Config reader