1

I want to visualize and analize a networkDynamic object using R. After creating it:

NetDyn.dynACCP <- networkDynamic(base.net=NetAccionCP, edge.spells=esUS2)
NetDyn.dynACCP
class(NetDyn.dynACCP)

I proceeded to visualize it: 

render.d3movie(NetDyn.dynACCP, displaylabels=FALSE, bg="#ffffff", 
               vertex.border="#333333", vertex.col=vertex, usearrows = TRUE, 
               vertex.cex = 0.3,
               render.par=list(tween.frames=5, show.time = TRUE), script.type="remoteSrc")

with success, however, at this point there appear all nodes even though they become active on a later time period. To collapse the network depending on the edge activity I used the function reconcile.vertex.activity, by this way the vertices become active only when they have edges activated. Nontheless, the vertex.attribute called color based on a type of the nodes turn to be messed. I mean messed, because attributes of some nodes are shifted to other nodes, as a result, they are turned into false. 

reconGraphACCP <- reconcile.vertex.activity(NetDyn.dynACCP, mode="expand.to.edges")
reconGraphACCP

I tried to correct this by adding TEAS or dynamic attributes to the object:

NetDyn.dynACCP <- networkDynamic(base.net=NetAccionCP, edge.spells=esUS2, 
                   vertex.spells=vsUS2, create.TEAs = TRUE, vertex.TEA.names = 'tipoActor')
NetDyn.dynACCP
class(NetDyn.dynACCP)

for this, I also include vertex.spells with the vertex attribute, now color as color.active,

Dynamic (TEA) attributes, Vertex TEAs: animation.x.active, animation.y.active, color.active

Nevertheless, I still cannot make the vertex attribute color.active to match the correspondent nodes at the visualization. The same happens if I used the function network.extract to analyze a determinate time period:

plot(network.extract(reconGraphACCP, onset=1992, terminus=1999), 
     main="1992-1999", vertex.col=color.active)

I would appreciate any help at this point.

CJ Yetman
  • 8,373
  • 2
  • 24
  • 56
Vlad
  • 125
  • 4
  • 9
  • Just a comment, this may help someone to help finding out a clue to solve my problem, I have found that when applied the function `network.extract`, the `vertex.ids` change, therefore, it ist not posible that the collapsed time periods preserve the same `ids`, so the `vertex.attributes` change over time periods, since the `vertex.ids` change. Someone with an idea? – Vlad Jun 17 '18 at 04:28
  • 1
    I think you will need to provide some example data, it is very hard to help debug otherwise. For example how are the objects `esUS2` and `vsUS2` structured? – skyebend May 16 '19 at 04:17
  • Thanks, about the question I have going so far to improve some problems with the `vertex attributes` of the dynamic object. However, there still are some details that I have not solved, as commented beforehand. The structure is the recommended in the literature, so: `onset`, `terminus`, `head`, `tail`. these for the `esUS2` or relational information, time variables are in years. On the other side the `vsUS2` corresponds to the vertices information, and contents `onset`, `terminus` and `ID`. – Vlad May 30 '19 at 17:27

0 Answers0