simulate_unregistered_curves: Simulate unregistered curves

In julia-wrobel/registr: Curve Registration for Exponential Family Functional Data

Description

This function simulates unregistered curves, providing the time values for both the unregistered curves (t_star) and the registered curves (t). Curves all have one peak, the location of which is shifted on the unregistered domain, meant to mimic accelerometer data.

Usage

simulate_unregistered_curves(
  I = 50,
  D = 100,
  lambda = 15,
  seed = 1988,
  period = 2 * pi,
  spline_based = FALSE,
  phase_variation = TRUE
)

Arguments

I	Number of subjects. Defaults is 50.
D	Number of grid points per subject. Default is 100.
lambda	Standard deviation for subject-specific amplitudes.
seed	Seed for reproducibility. Default is 1988.
period	Controls the period of the mean curve
spline_based	If FALSE curve is constructed using sine and cosine functions, if TRUE, curve is constructed using B-spline basis.
phase_variation	If TRUE, creates phase variation (registered curves are observed on uneven grid). If FALSE, no phase variation.

Value

A simulated dataframe with variables id, value, index, latent_mean, and t. Index is the domain on which curves are unregistered and t is the domain on which curves are registered.

Author(s)

Julia Wrobel [e-mail], Jeff Goldsmith [e-mail]

julia-wrobel/registr documentation built on Jan. 16, 2022, 2:51 a.m.