CANM_data: Synthetic data for CANM.

In DMIRLAB-Group/CANM: Causal Discovery with Cascade Nonlinear Additive Noise Models

Description

Amore general version for Generating data from transitive ANM.

Usage

CANM_data(depth = 1, sample_size = 5000, ratio = 1, k = 3,
  changepoints = 6, sd = 1, noisedist = "rnorm", ...)

Arguments

depth	The depth of the CANM (the number of the latent intermediated variables).
sample_size	The number of samples.
ratio	The ratio of noise.
k	The number of components in the mixture Gaussian distribution of X.
changepoints	The mapping function at each layer are generated from a cubic spline interpolation using a 6-dimensional grid with respect to 6 random generated points as knots for the interpolation; Here the changepoint control the number of knots as controlling the nonlinearity of the mapping function.
sd	The standard error of the random generative knots.
noisedist	The noise distribution at each layer.
...	Additional parameters for the specify noise distribution.

Value

Data

Examples

set.seed(0)
data=CANM_data(depth=2,ratio=0.2)
plot(data[,1],data[,2])

DMIRLAB-Group/CANM documentation built on May 31, 2019, 12:40 a.m.