Bonus Waffle House example -- which models are "DAG endorsed"?

[rpruim]

From the end of chapter 6...

The default output of adjustmentSets() produces only minimal sets of variables to condition on. But there may be other adjustments sets that are not minimal. Setting type = all will list them.

library(dagitty)
dag_6.2 <- 
  dagitty("dag {
  A -> D
  A -> M -> D
  A <- S -> M
  S -> W -> D
           }"
  )

paths(dag_6.2, from = "W", to = "D")
#> $paths
#> [1] "W -> D"                "W <- S -> A -> D"      "W <- S -> A -> M -> D"
#> [4] "W <- S -> M -> D"      "W <- S -> M <- A -> D"
#> 
#> $open
#> [1]  TRUE  TRUE  TRUE  TRUE FALSE
adjustmentSets(dag_6.2, exposure = "W", outcome = "D")
#> { A, M }
#> { S }
adjustmentSets(dag_6.2, exposure = "W", outcome = "D", type = "all")
#> { A, M }
#> { S }
#> { A, S }
#> { M, S }
#> { A, M, S }

^{Created on 2021-03-25 by the reprex package (v1.0.0)}

It might be good to mention this so users are aware that the minimal sets are not the only ones endorsed by the DAG. It might even be worth adding a model the is endorsed but not minimal.

[ASKurz]

Good suggestion. Done.