simulate_data: Simulate data

In DavidKLim/NIMIWAE: Non-Ignorably Missing Importance Weighted Autoencoder

Simulate data

Description

Simulate data

Usage

simulate_data(
  N,
  D,
  P,
  sim_index,
  seed = 9 * sim_index,
  ratio = c(8, 2),
  g_seed = 333,
  beta = c(rep(-1/4, floor(P/2)), rep(1/4, P - floor(P/2))),
  nonlinear = F
)

Arguments

N	Number of observations
D	Dimension of latent Z
P	Number of features
sim_index	Integer index for sim. Can use to set seed
seed	Seed. Either set custom or set some value x sim_index. Default: 9 x sim_index
ratio	Train-valid-test ratio for splitting of observations
g_seed	Seed for train-valid-test dataset splitting of observations
beta	coefficients for each column of X in simulating a binary class response variable

Value

list of objects: data (N x P matrix), classes (subgroups of observations), params (those used for simulating data), and g (partitioning of data into train-valid-test sets)

Author(s)

David K. Lim, [e-mail]

References

https://github.com/DavidKLim/NIMIWAE

Examples

simulate_data(N = 10000, D = 2, P = 8, sim_index = 1)

DavidKLim/NIMIWAE documentation built on Jan. 19, 2024, 11:18 p.m.