data_from_scratch: Simulates data under the PLN model with control on the...

In Rmomal/EMtree: Infers Direct Species Association Networks using Tree Averaging

Simulates data under the PLN model with control on the dependency structure

Description

Simulates data under the PLN model with control on the dependency structure

Usage

data_from_scratch(
  type,
  p = 20,
  n = 50,
  r = 5,
  covariates = NULL,
  dens = log(p)/p,
  k = 3,
  norm = FALSE,
  signed = FALSE,
  v = 0.01,
  draw = FALSE
)

Arguments

type	type of graph, either "tree", "erdos", "cluster" or "scale-free"
p	wanted number of columns (species)
n	wanted number of rows (samples/observations)
r	within/between connectivity ratio for cluster graphs
covariates	a data.frame or matrix containing data covariates.
dens	density of edges for cluster graphs or edge probability for Erdös graphs
k	number of groups for the cluster structure.
norm	should the Gaussian parameters be normalized ?
signed	boolean: should the graph be composed of positive and negative partial correlations ?
v	noise parameter of the precision matrix
draw	boolean, plots the graph if set to TRUE

Value

a list containing

• data: simulated counts

• omega: the precision matrix

Examples

set.seed(1)
p<-30
Y1<-data_from_scratch("tree",p=p,draw=TRUE)
str(Y1)
Y2<-data_from_scratch("cluster",p=p,r=10, dens=10/p, k=3,draw=TRUE)

Rmomal/EMtree documentation built on Dec. 14, 2024, 8:16 a.m.