Feature Descriptions

This chapter contains an overview table of the catch22 features as well as detailed descriptions and examples of each feature, organized into 9 distinct conceptual categories.

The catch22 feature set provides open access to a powerful set of time-series analysis features that implemented efficiently for fast native feature computation in multiple coding languages. But the features (taken from the larger hctsa time-series feature set) can be hard to interpret 😥

The material in this chapter should help you to understand what each of these 22 features is doing, and point you to the relevant time-series theory that underpins it.

catch22 Feature Descriptions

Select the feature overview table for a high level description of each feature, or delve into detailed explanations and examples by exploring each of the 9 categories of time-series features included in catch22:

Feature overview table

Distribution shape

Extreme event timing

Linear autocorrelation structure

Nonlinear autocorrelation

Symbolic

Incremental differences

Simple forecasting

Self-affine scaling

Other

---