additive_pupil_sim: Simulate pupil data to test papss

In JoKra1/papss: What the package does (short line)

Simulate pupil data to test papss

Description

Uses the pupil response function described by Hoeks & Levelt (1993) with the refinements proposed by Wierda e t al. (2012) to simulate pupil dilation time-course data with three sources of deviation:

1. Simulates systematic per-subject deviation from a 'shared' population trend in demand

2. Simulates per-trial deviation from each subject's individual 'true demand'

3. Assumes random noise (N(0, sigma), with constant sigma) for each trial

Usage

additive_pupil_sim(
  nk = 20,
  n_sub = 10,
  n_trials = 250,
  pulse_loc_diff = 1,
  n_diff = 4,
  spars_deg = 0.5,
  sub_dev = 0.15,
  slope_dev = 1.5,
  trial_dev = 0.05,
  residual_dev = 15,
  n = 10.1,
  t_max = 930,
  f = 1/(10^24),
  should_plot = T,
  seed = 124
)

Arguments

nk	Number of knots for the B-spline basis.
n_sub	How many subjects to simulate.
n_trials	How many trials to simulate.
pulse_loc_diff	Assume a pulse every 'pulse_loc_diff' samples (one sample = 20 ms)
n_diff	Maximum number of spline basis coefficients with systematic per-subject variation
spars_deg	The degree of sparsity enforced in spline basis coefficient vector
sub_dev	Standard deviation of normal distribution used to sample by-subject variation
slope_dev	Standard deviation of normal distribution used to sample by-subject slope variation
trial_dev	Standard deviation of normal distribution used to sample by-trial variation (in spike weights and coefficients)
residual_dev	Standard deviation of normal distribution used to sample residuals per trial
n	Parameter defined by Hoeks & Levelt (number of laters)
t_max	Parameter defined by Hoeks & Levelt (response maximum in ms)
f	Parameter defined by Wierda et al. (scaling factor)
should_plot	Whether the generated data should be visualized.
seed	For replicability

Details

Demand is modelled according to a simple B-spline (see Eilers & Marx, 2010) with equally spaced knots with associated coefficients of which only a small percentage will be different from zero (to ensure that the simulated demand trajectory is sparse).

Examples

pupil_sim <- additive_pupil_sim(n_sub=5)
sim_dat <- pupil_sim$data

JoKra1/papss documentation built on June 15, 2022, 8:57 a.m.