Mark.groups in Igraph Error: unknown vertex.names selected - r

Question

I am trying to create a network plot in igraph where
communities are marked by a color overlay as created by mark.groups
nodes are colored by a node attribute: deu
nodes are shaped a node attribute: topic_type
For this, I created an igraph object
And now, I try the following code:
set.seed(2)
plot(graph_deu,
mark.groups=list(c(33,1,34,2,36,53,54,56,42,43,55,57,18), c(35,48,50,27), c(38,45,46,47,49,28,25)),
mark.col=c("lemonchiffon", "slategray1", "thistle1"),
mark.border = NA,
edge.width =E(graph_deu)$weight,
vertex.size = deu_deg,
vertex.color = deu,
vertex.shape = topic_type,
vertex.label = node_labels,
vertex.label.cex=1.5
)
And I get the error:
Fehler in simple_vs_index(x, ii, na_ok) : Unknown vertex selected.
This seems to be igraph not finding the vertices as specified in mark.groups, but I have actually no idea why it would not find these vertices, as they are all correctly numbered.
Then, to avoid mark.groups, I tried another option - I directly plot the community object (mod2), however, in this case, nodes get colored according to community and not according to attribute deu:
plot(mod2, graph_deu,
edge.width =E(graph_deu)$weight,
vertex.size = deu_deg,
vertex.color = deu,
vertex.shape = topic_type,
vertex.label = node_labels,
vertex.label.cex=1.5)
This produces a network where vertices are colored by community, not by deu-attribute. What I would like it to look: I would like the communities to be circled by the semi-transparent overlay, but the nodes should be individually colored by deu-attribute.
Your help will be very appreciated. this is my first post on stackoverflow so if I should provide more pieces of code to reproduce I am happy to share it, I hope though that my igraph object is enough for the problem at hand.

Answer 1

Your graph has only 24 nodes, but you are referring to nodes using number higher than that, e.g. 36,53,54. If you use a number, igraph assumes that is the number of the node, so these number don't make sense for this graph. What you mean is the nodes with the names "36","53","54". The names are strings, not numbers. What you need to do is find the node numbers that correspond to these names. I show one way to do that below. Also, your plot statement refers to a number of variables that you did not provide so I commented them out here.
graph_deu = upgrade_graph(graph_deu)
plot(graph_deu)
Group1 = as.numeric(V(graph_deu)[as.character(c(33,1,34,2,36,53,54,56,42,43,55,57,18))])
Group2 = as.numeric(V(graph_deu)[as.character(c(35,48,50,27))])
Group3 = as.numeric(V(graph_deu)[as.character(c(38,45,46,47,49,28,25))])
set.seed(2)
plot(graph_deu,
mark.groups=list(Group1, Group2, Group3),
mark.col=c("lemonchiffon", "slategray1", "thistle1"),
mark.border = NA,
edge.width =E(graph_deu)$weight,
# vertex.size = deu_deg,
# vertex.color = deu,
# vertex.shape = topic_type,
# vertex.label = node_labels,
vertex.label.cex=1.5
)