template_gp_warp: Template Gaussian Process Warping

In wzhorton/warptk: Warping Toolkit

Description

Performs hierarchical curve warping by modeling the warping functions with GP.

Usage

template_gp_warp(
  y_list,
  feat_list,
  template_feats,
  niter = 1000,
  nburn = 1000,
  int_p = 20,
  asig = 0.1,
  bsig = 0.1,
  at = 0.1,
  bt = 0.1,
  al = 0.1,
  bl = 0.1,
  aa = 1,
  ba = 30,
  be = 2,
  tune = 2,
  progress = TRUE,
  debug = FALSE
)

Arguments

y_list	list of curves with equal length
feat_list	list of feature vectors with equal length
template_feats	a feature vector with length mathing those in feat_list
niter, nburn	MCMC iterations and burns. Total run time is the sum.
int_p	number of internal knots specified for the data curves
asig, bsig, at, bt, al, bl	hyperparameters for inverse gamma priors on model variances
aa, ba	hyperparameters for the uniform prior on inverse range parameter
tune	tuning parameter for the proposal normal distribution of alpha.
progress	logical; indicates whether or not to print percent progress.
debug	logical; stops function right before return and enters debug mode.

Value

a list of warped curves, the estimated mean curve, and a vector of MCMC acceptance rates. Curve list elements are vectors with length matching the elements of y_list.

wzhorton/warptk documentation built on Nov. 14, 2020, 10:01 p.m.