Principal Machines for a Sufficient Dimension Reduction

A repository for a unified principal sufficient dimension reduction via principal machines. It includes codes for '(Linear / Nonlinear) principal sufficient dimension reduction (psdr and npsdr)' and 'realtime (rtpsdr)' and their generic methods.

Copy Link

Version

Install

install.packages('psvmSDR')

Monthly Downloads

152

Version

3.0.1

License

GPL-2

Maintainer

Jungmin Shin

Last Published

February 16th, 2026

Functions in psvmSDR (3.0.1)

psdr_bic

Structural dimension selection for principal SDR

npsdr

A unified Principal sufficient dimension reduction method via kernel trick

plot.psdr

Plot sufficient predictors from a psdr object

npsdr_x

Reconstruct estimated sufficient predictors for new data

plot.npsdr

Plot sufficient predictors from an npsdr object

rtpsdr

Real time sufficient dimension reduction through principal least squares SVM

psdr

Unified linear principal sufficient dimension reduction methods