CsnowyLstar
commented
3 days ago Thank you for your attention to our work!
It should be noted that if you're not considering network dynamical systems (NDSs, as shown in Eq. 14 of the main text), you won't need to consider the structure of 'edges_ex.csv'. Otherwise, since the interaction structure in NDSs are the same (with different weights), we can record the common interaction conditions between subsystems using 'edges_ex.csv' (corresponding to the interaction dynamics G).
Here, we use a triplet ([u],[v],[w]) to record this coupling interaction structure, where the first element represents the node of the subsystem $i$ being driven, the second element represents the neighbor in the coupling subsystem $j$ that influences u, and the third element represents the neighbor within subsystem $i$ that influences u. Moreover, both v and w are located in the interaction dynamics G.
Of course, this is just a simple way of recording, which can clearly document the several systems considered in our work. For more complex situations, a more detailed method of recording may be required.
github-lyj
Dear author, I am deeply appreciative of the code you have shared, and have gained tremendous insights from your paper. However, I find myself puzzled about the methodology for inferring interactions between subsystems. My current understanding is that the files 'edges_ex.csv' and 'edges.csv' are instrumental in this process. What perplexes me is the rationale behind the numerical configurations in 'edges_ex', and how it synergistically works with 'edges'. I am eagerly looking forward to your clarification on this matter. Many thanks in advance for your time and assistance!