pppca

Point Process Principal Component Analysis (PPPCA) provides kernel PCA tools for multivariate point processes defined on $[0,1]^d$.

Key features

• Fit PPPCA models and evaluate eigenfunctions on demand.

• Persist model state for reuse and projection of new samples.

• Visualize low-dimensional eigenfunctions.

Contents

• Usage
  • Quickstart
  • Save and reload models
  • Project new samples
  • Visualize eigenfunctions
• API Reference
  • load_pppca_features()
  • plot_eigenfunctions()
  • pppca()
  • project_pppca()
  • save_pppca_features()