dataSimu_group_sparse: Simulate multiple data sets with group and element-wise...

In YipengUva/RpESCA: R implementaion of pESCA models with various concave penalties

Description

This function can generate multiple data sets of mixed data types. Details of the simultion process can be found in the paper https://arxiv.org/abs/1902.06241. In the current version, both group and element-wise sparsity are included into the loading matrix.

Usage

dataSimu_group_sparse(n, ds, dataTypes = "GGG", noises = rep(1, 3),
  margProb = 0.1, sparse_ratio = 0, SNRgc = 1, SNRlc = rep(1, 3),
  SNRd = rep(1, 3))

Arguments

n	the number of objects
ds	a vector for the number of variables in each data set
dataTypes	a string indicates the data type of each data set, possible options include 'G': Gaussian, 'B': Bernoulli.
noises	noise levels of simulated data sets
margProb	desired marginal probability for binary data simulation, used to simulate imbalanced binary data.
sparse_ratio	controls the sparse level of element-wise sparsity
SNRgc	SNR for global common structure
SNRlc	SNRs of the local common structures
SNRd	SNRs of the distinct structures

Value

A list contains the simulated data sets and the simulated parameters.

• X: a list contains the simulated multiple data sets;

• Theta_simu: simulated natural parameter matrix Theta;

• mu_simu: simulated offset term mu;

• U_simu: simulated U;

• D_simu: simulated D;

• V_simu: simulated V;

• E_simu: simulated noise term E;

• dataTypes: vector form simulated data types;

• S_simu: desired group sparse pattern;

• SNRs: used SNRs in simulating common (global, local) and distinct structures;

• varExpTotals_simu: variation explained ratios for each data set computed using the simulated parameters;

• varExpPCs_simu: variation explained raitos for each PC for each data set computed using the simulated parameters;

Examples

## Not run: 
dataSimulation <- dataSimu_group_sparse(n=200,
                                       ds=c(400,200,100),
                                       dataTypes='GGB')

## End(Not run)

YipengUva/RpESCA documentation built on July 2, 2019, 6:41 p.m.