py-why
diff --git a/‎.idea/csv-plugin.xml‎
Lines changed: 0 additions & 114 deletions b/‎.idea/csv-plugin.xml‎
Lines changed: 0 additions & 114 deletions
diff --git a/‎docs/source/search_methods_index/Constraint-based causal discovery methods/FCI.rst‎
Lines changed: 5 additions & 1 deletion b/‎docs/source/search_methods_index/Constraint-based causal discovery methods/FCI.rst‎
Lines changed: 5 additions & 1 deletion
diff --git a/‎docs/source/search_methods_index/Constraint-based causal discovery methods/pag.png‎
46.4 KB b/‎docs/source/search_methods_index/Constraint-based causal discovery methods/pag.png‎
46.4 KB
@@ -55,7 +55,11 @@ For detailed usage, please kindly refer to its `usage example <https://github.co
 
 Returns
 -------------------
-**G**: a CausalGraph object, where G.graph[j,i]=1 and G.graph[i,j]=-1 indicates  i --> j; G.graph[i,j] = G.graph[j,i] = -1 indicates i --- j; G.graph[i,j] = G.graph[j,i] = 1 indicates i <-> j; G.graph[j,i]=1 and G.graph[i,j]=2 indicates  i o-> j.
+**G**: a CausalGraph object, where G is a PAG and the illustration of its end nodes is as follows:
+
+.. image:: pag.png
+  :width: 350
+
 
 **edges**: list. Contains graph's edges properties. If edge.properties have the Property 'dd', then there is no latent confounder. Otherwise, there might be latent confounders. If edge.properties have the Property 'nl', then it is definitely direct. Otherwise, it is possibly direct.