R BNlearn＆SHD功能如何运作？

Question

让我们从文档中获取示例和说明

shd(...)

e1图是：

e1

e2图是：

e2

函数> shd(e2, e1, debug = TRUE) * arcs between A and F do not match. > the learned network contains A - F. > the true network contains A -> F. * arcs between E and F do not match. > the learned network contains F -> E. > the true network contains E -> F. [1] 2 应返回这些图之间的差异计数（方向，存在或不存在某些弧）。但它返回2并表示e2包含A-F弧，但它仅包含A-> F弧。为什么呢？

db.game.aggregate([
 {$unwind: "$array"}, 
 {$group:{"_id" : "$array.ToId", "pointCount": {$sum:"$array.point"}}}, 
 {$project: {_id: 0, "ToId":"$_id", "pointCount": 1}}
 ])