I am using LINGAM in python to build a DAG. The output is with the weight and edges. How to interpret these edges?
I am assuming the sign says proportionality of the causality. i.e if it is + it is closely related and - is lesser related. An example of DAG
A positive edge from variable x1 to variable x2 implies a positive correlation between changes in variable x1 and changes in variable x2. To put it differently, when x1 goes up so does x2. The strength of the positive causal connection is revealed by the weight of the positive edge.
On the other side, the negative edge between variable x1 and variable x4 shows that changes in x1 have a negative correlation with changes in x4. If x1 increases, x4 decreases or vice versa. Again, the weight of the negative edge corresponds to the intensity of the negative causal association.
