Compute all the P-values of the d-separation tests implied by the missing edges of a given acyclic graph (DAG). The conditioning set Z is represented, in a DAG, by the union of the parent sets of X and Y (Shipley, 2000). The results of every test, in a DAG, is then combined using the Fisher’s statistic in an overall test of the fitted model C = -2*sum(log(P-value(k))), where C is distributed as a chi-squared variate with df = 2k, as suggested by Shipley (2000).
Shipley.test(
graph,
data,
MCX2 = FALSE,
cmax = Inf,
limit = 100,
verbose = TRUE,
...
)A directed graph as an igraph object.
A data matrix with subjects as rows and variables as columns.
If TRUE, a Monte Carlo P-value of the combined C test is enabled using the R code of Shipley extracted from <https://github.com/BillShipley/CauseAndCorrelation>.
Maximum number of parents set, C. This parameter can be used to perform only those tests where the number of conditioning variables does not exceed the given value. High-dimensional conditional independence tests can be very unreliable. By default, cmax = Inf.
An integer value corresponding to the graph size (vcount)
tolerance. Beyond this limit, multicore computation is enabled to
reduce the computational burden. By default, limit = 100.
If TRUE, Shipley's test results will be showed to screen (default = TRUE).
Currently ignored.
A list of three objects: (i) "dag": the DAG used to perform the Shipley test (ii) "dsep": the data.frame of all d-separation tests over missing edges in the DAG and (iii) "ctest": the overall Shipley's' P-value.
Shipley B (2000). A new inferential test for path models based on DAGs. Structural Equation Modeling, 7(2): 206-218. <https://doi.org/10.1207/S15328007SEM0702_4>
#\donttest{
# Nonparanormal(npn) transformation
als.npn <- transformData(alsData$exprs)$data
#> Conducting the nonparanormal transformation via shrunkun ECDF...done.
sem <- SEMrun(alsData$graph, als.npn)
#> NLMINB solver ended normally after 1 iterations
#>
#> deviance/df: 10.92504 srmr: 0.2858859
#>
C_test <- Shipley.test(sem$graph, als.npn, MCX2 = FALSE)
#> d-separation test (basis set) of 420 edges...
#>
/ 0 % elapsed=00s
- 1 % elapsed=00s, remaining~00s
\ 2 % elapsed=00s, remaining~00s
| 4 % elapsed=00s, remaining~00s
/ 5 % elapsed=00s, remaining~00s
- 6 % elapsed=00s, remaining~00s
\ 7 % elapsed=00s, remaining~00s
| 8 % elapsed=00s, remaining~00s
/ 10% elapsed=00s, remaining~00s
- 11% elapsed=00s, remaining~00s
\ 12% elapsed=00s, remaining~00s
| 13% elapsed=00s, remaining~00s
/ 14% elapsed=00s, remaining~00s
- 15% elapsed=00s, remaining~00s
\ 17% elapsed=00s, remaining~00s
| 18% elapsed=00s, remaining~00s
/ 19% elapsed=00s, remaining~00s
- 20% elapsed=00s, remaining~00s
\ 21% elapsed=00s, remaining~00s
| 23% elapsed=00s, remaining~00s
/ 24% elapsed=00s, remaining~00s
- 25% elapsed=00s, remaining~00s
\ 26% elapsed=00s, remaining~00s
| 27% elapsed=00s, remaining~00s
/ 29% elapsed=00s, remaining~00s
- 30% elapsed=00s, remaining~00s
\ 31% elapsed=00s, remaining~00s
| 32% elapsed=00s, remaining~00s
/ 33% elapsed=00s, remaining~00s
- 35% elapsed=00s, remaining~00s
\ 36% elapsed=00s, remaining~00s
| 37% elapsed=00s, remaining~00s
/ 38% elapsed=00s, remaining~00s
- 39% elapsed=00s, remaining~00s
\ 40% elapsed=00s, remaining~00s
| 42% elapsed=00s, remaining~00s
/ 43% elapsed=00s, remaining~00s
- 44% elapsed=00s, remaining~00s
\ 45% elapsed=00s, remaining~00s
| 46% elapsed=00s, remaining~00s
/ 48% elapsed=00s, remaining~00s
- 49% elapsed=00s, remaining~00s
\ 50% elapsed=00s, remaining~00s
| 51% elapsed=00s, remaining~00s
/ 52% elapsed=00s, remaining~00s
- 54% elapsed=00s, remaining~00s
\ 55% elapsed=00s, remaining~00s
| 56% elapsed=00s, remaining~00s
/ 57% elapsed=00s, remaining~00s
- 58% elapsed=00s, remaining~00s
\ 60% elapsed=00s, remaining~00s
| 61% elapsed=00s, remaining~00s
/ 62% elapsed=00s, remaining~00s
- 63% elapsed=00s, remaining~00s
\ 64% elapsed=00s, remaining~00s
| 65% elapsed=00s, remaining~00s
/ 67% elapsed=00s, remaining~00s
- 68% elapsed=00s, remaining~00s
\ 69% elapsed=00s, remaining~00s
| 70% elapsed=00s, remaining~00s
/ 71% elapsed=00s, remaining~00s
- 73% elapsed=00s, remaining~00s
\ 74% elapsed=00s, remaining~00s
| 75% elapsed=00s, remaining~00s
/ 76% elapsed=00s, remaining~00s
- 77% elapsed=00s, remaining~00s
\ 79% elapsed=00s, remaining~00s
| 80% elapsed=00s, remaining~00s
/ 81% elapsed=00s, remaining~00s
- 82% elapsed=00s, remaining~00s
\ 83% elapsed=00s, remaining~00s
| 85% elapsed=00s, remaining~00s
/ 86% elapsed=00s, remaining~00s
- 87% elapsed=00s, remaining~00s
\ 88% elapsed=00s, remaining~00s
| 89% elapsed=00s, remaining~00s
/ 90% elapsed=00s, remaining~00s
- 92% elapsed=00s, remaining~00s
\ 93% elapsed=00s, remaining~00s
| 94% elapsed=00s, remaining~00s
/ 95% elapsed=00s, remaining~00s
- 96% elapsed=00s, remaining~00s
\ 98% elapsed=00s, remaining~00s
| 99% elapsed=00s, remaining~00s
/ 100% elapsed=00s, remaining~00s
#> C_test df pvalue
#> 1 7038.822 840 0
#MC_test <- Shipley.test(sem$graph, als.npn, MCX2 = TRUE)
#}