A simple wrapper for GGMpathStats.

GGMpathStats.fused(sparsePlist, ...)

Arguments

sparsePlist

A list of sparsified precision matrices.

...

Arguments passed to GGMpathStats.

Value

A list of path stats.

Note

The function currently fails if no paths are present in one of the groups.

See also

Author

Anders E. Bilgrau, Carel F.W. Peeters <carel.peeters@wur.nl>, Wessel N. van Wieringen

Examples

## Obtain some (high-dimensional) data
set.seed(1)
ns <- c(10, 11)
Slist <- createS(ns, p = 7, topology = "banded")
Tlist <- default.target.fused(Slist, ns)

## Obtain regularized precision and sparsify
Plist <- ridgeP.fused(Slist, ns, Tlist, lambda = c(1, 1.6))
#> i = 1   | max diff = 6.0206645311e-01
#> i = 2   | max diff = 1.6272241569e-01
#> i = 3   | max diff = 4.8213930894e-03
#> i = 4   | max diff = 2.9467129378e-04
#> i = 5   | max diff = 1.9347840638e-05
#> i = 6   | max diff = 1.4191856301e-06
#> i = 7   | max diff = 1.1537054419e-07
#> i = 8   | max diff = 1.0121394755e-08
#> Converged in 8 iterations, max diff < 1.49e-08.
sparsePlist <- sparsify.fused(Plist, threshold = "absValue", absValueCut = 0.20)
#> - Retained elements:  15 
#> - Corresponding to 71.43 % of possible edges 
#>  
#> - Retained elements:  10 
#> - Corresponding to 47.62 % of possible edges 
#>  
SPlist <- lapply(sparsePlist, "[[", "sparsePrecision")

## Obtain information on mediating and moderating paths between nodes 14 and 23
res <- GGMpathStats.fused(SPlist, node1 = 3, node2 = 4, graph = FALSE)
#> 
#> 
#> ========================================
#> Class: class1
#> Covariance between node pair : -0.79168 
#> ----------------------------------------
#>             path length contribution
#> 1           3--4      2      1.07541
#> 2     3--1--2--4      1      -1.0261
#> 3  3--1--2--6--4      2      0.78106
#> 4  3--1--2--7--4      2     -0.77765
#> 5        3--2--4      3     -0.76944
#> 6  3--2--6--5--4      3     -0.68468
#> 7     3--2--6--4      3      -0.6788
#> 8     3--2--7--4      3     -0.64249
#> 9  3--5--6--2--4      2      0.63593
#> 10       3--5--4      3      0.53368
#> 11    3--5--6--4      3      0.44948
#> 12 3--5--6--7--4      3      -0.4117
#> 13    3--6--2--4      3      0.37015
#> 14    3--6--5--4      3     -0.31502
#> 15       3--6--4      4     -0.27736
#> 16    3--6--7--4      4     -0.27498
#> 17    3--7--2--4      4     -0.25402
#> 18 3--7--6--5--4      4      0.24596
#> 19    3--7--6--4      4     -0.21394
#> 20       3--7--4      4     -0.17207
#> ----------------------------------------
#> Sum path contributions       : -2.40659 
#> 
#> 
#> ========================================
#> Class: class2
#> Covariance between node pair : 0.19206 
#> ----------------------------------------
#>         path length contribution
#> 1       3--4      2      0.32147
#> 2    3--2--4      1     -0.30642
#> 3 3--2--5--4      2      0.22803
#> 4 3--5--2--4      2     -0.18215
#> 5    3--5--4      3      0.06634
#> 6    3--6--4      3      0.06479
#> ----------------------------------------
#> Sum path contributions       : 0.19206