Transforms

coco_pipe.transforms contains reusable scikit-learn-compatible data transformers for subject structure in vectors, token representations, and temporal trajectories.

• Subject alignment
  • Direct use
  • Decoding integration
  • Fit scope and leakage
  • LEACE erasure
  • Euclidean alignment
  • Riemannian alignment
  • Choosing a method
  • Auditing the result
  • API summary
• Temporal trajectory alignment
  • Decoding integration
  • The transductive assumption
  • Direct use

The public registry exposes four methods:

Registry name	Class	Input	Main operation
leace	LeaceEraser	(observation, feature)	Remove the linearly predictable subject subspace
ea_mean	EuclideanAlign	(observation, feature)	Center every subject
ea_coral	EuclideanAlign	(observation, feature)	Center and whiten every subject
ra	RiemannAlign	(window, token, feature)	Recenter subject covariance matrices and project to tangent space

Temporal trajectories also support participant-specific PCA plus orthogonal Procrustes alignment:

Class	Input	Main operation
TemporalProcrustesAlignment	(observation, feature, time)	Rotate each participant’s PCA trajectories toward a training template

See Subject alignment for the vector/token methods and Temporal trajectory alignment for temporal decoding.