Skip to contents

Generate sample data from a nonlinear additive noise model (for RESIT)

Source: R/generate_resit_sample.r
generate_resit_sample.Rd

Generates a 4-variable nonlinear SEM whose structure can only be recovered by a nonlinear method such as lingam_resit(); linear LiNGAM variants are not expected to work on this data.

Usage

generate_resit_sample(n = 300L, seed = NULL)

Arguments

n

number of samples (default: 300)

seed

random seed (default: NULL, i.e. do not reset the RNG state)

Value

list with three elements:

  • data: data.frame of the 4 observed variables (x0-x3).

  • adjacency_matrix: the true 4x4 adjacency matrix following the m[to, from] convention. Entries are 0/1 edge indicators (not coefficients), directly comparable to the output of lingam_resit().

  • causal_order: the true causal order (1-based column positions, source first).

Details

The data-generating process (all error terms e() are runif(n, -0.5, 0.5)):


x0 ~ runif(-1, 1)
x1 = 3.0 * x0^2 + e()
x2 = 2.0 * tanh(x1) + 0.8 * x0^3 + e()
x3 = 1.5 * sin(x2) + e()

Examples

nonlinear <- generate_resit_sample(n = 300, seed = 1)
head(nonlinear$data)
#>           x0          x1           x2           x3
#> 1 -0.4689827  0.83354648  1.596482936  1.831404164
#> 2 -0.2557522 -0.20891458  0.003539709  0.272152307
#> 3  0.1457067  0.05628747 -0.237588016 -0.580258311
#> 4  0.8164156  1.96115504  2.607512659  0.451737357
#> 5 -0.5966361  0.94314057  1.779071173  1.193345462
#> 6  0.7967794  2.39567131  2.846533624 -0.001835138
nonlinear$adjacency_matrix
#>    x0 x1 x2 x3
#> x0  0  0  0  0
#> x1  1  0  0  0
#> x2  1  1  0  0
#> x3  0  0  1  0