gmG: Graphical Model 8-Dimensional Gaussian Example Data

In pcalg: Methods for Graphical Models and Causal Inference

Graphical Model 8-Dimensional Gaussian Example Data

Description

These two data sets contain a matrix containing information on eight gaussian variables and the corresonding DAG model.

Usage

data(gmG)

Format

gmG and gmG8 are each a list of two components

x:

a numeric matrix 5000 \times 8.

g:

a graph, i.e., of formal class "graphNEL" from package graph with 6 slots
.. ..@ nodes : chr [1:8] "1" "2" "3" "4" ...
.. ..@ edgeL :List of 8
........

Details

The data was generated as indicated below. First, a random DAG model was generated, then 5000 samples were drawn from “almost” this model, for gmG: In the previous version, the data generation wgtMatrix had the non-zero weights in reversed order for each node. On the other hand, for gmG8, the correct weights were used in all cases

Source

The data set is identical to the one generated by

    ## Used to generate "gmG"
    set.seed(40)
    p <- 8
    n <- 5000
    ## true DAG:
    vars <- c("Author", "Bar", "Ctrl", "Goal", paste0("V",5:8))
    gGtrue <- randomDAG(p, prob = 0.3, V = vars)
    gmG  <- list(x = rmvDAG(n, gGtrue, back.compatible=TRUE), g = gGtrue)
    gmG8 <- list(x = rmvDAG(n, gGtrue),                       g = gGtrue)
  

Examples

data(gmG)
str(gmG, max=3)
stopifnot(identical(gmG $ g, gmG8 $ g))
if(dev.interactive()) { ## to save time in tests
  round(as(gmG $ g, "Matrix"), 2) # weight ("adjacency") matrix
  plot(gmG $ g)
  pairs(gmG$x, gap = 0,
	panel=function(...) smoothScatter(..., add=TRUE))
}

pcalg documentation built on Sept. 26, 2023, 9:06 a.m.