Generate 1D functional data (current or legacy) — data_generator_po

Provides the current 1D generator while keeping access to the previous implementation via version = "legacy" for reproducibility.

Usage

data_generator_po_1d(
  n = 100,
  grid_points = 100,
  noise_sd = 0.25,
  center = TRUE,
  rsq = 0.95,
  mu = 0.1,
  univariate = TRUE,
  response_type = c("gaussian", "binomial"),
  linear_predictor = c("rectangular", "trapezoidal", "linear"),
  n_missing = 1,
  min_distance = NULL,
  version = c("current", "legacy"),
  ...
)

Arguments

n: Number of samples to generate.
grid_points: Number of points in the grid.
noise_sd: Standard deviation of measurement noise.
center: Whether to mean-center each curve.
rsq: Desired R-squared value for the response.
mu: Intercept term added to the linear predictor.
univariate: If TRUE, generate a single functional predictor; otherwise generate two.
response_type: Response distribution ("gaussian" or "binomial").
linear_predictor: Integration approach for the linear predictor ("rectangular", "trapezoidal", or "linear").
n_missing: Number of missing segments per curve.
min_distance: Minimum length of missing segments (defaults to one fifth of the grid length).
version: Choose "current" (default) or "legacy" implementation.
...: Additional arguments forwarded to the legacy implementation.

Value

A list containing simulated curves (noisy and noiseless), missing point indices, the coefficient functions, and the generated response.

Examples

data <- data_generator_po_1d(n = 10)
data_legacy <- data_generator_po_1d(n = 10, version = "legacy", beta_type = "trig")
#> [1] 0.0008242013