ropensci/targets

Better layouts for tar_visnetwork

mem48 opened this issue · 2 comments

mem48 commented

Prework

  • [ x] I understand and agree to help guide.
  • [ x] I understand and agree to contributing guide.
  • [ x] New features take time and effort to create, and they take even more effort to maintain. So if the purpose of the feature is to resolve a struggle you are encountering personally, please consider first posting a "trouble" or "other" issue so we can discuss your use case and search for existing solutions first.

Proposal

For complex target tar_visnetwork produces hard-to-understand networks. For example

image

In this case, the positioning of the targets creates a lot of crossing lines and makes the logic of the workflow hard to see. Using tar_mermaid on the same workflow produces

image

Which is a lot easier to understand.

I'm guessing there is not an easy single fix to this but could tar_visnetwork expose some of the underlying options in visNetwork so that users could adjust the layout of the network?

wlandau commented

tar_network() returns a visNetwork object that you can modify directly. And tar_network() returns the node and edge information so custom graphs can be created.

I haven't seen a way to achieve a hierarchical layout in visNetwork which is as nice as the mermaid.js graph in terms of the positioning of the edges. I will consider it if a really good set of options becomes available, but as far as I know visNetwork does not support this.

mem48 commented

I case anybody else has this problem I used

net = tar_network(TRUE)
nodes = net$vertices
edges = net$edges

names(nodes)[1] = "id"
nodes$label = nodes$id

visNetwork(nodes, edges) %>%
  visEdges(arrows = "to") %>% 
  visIgraphLayout(layout = "layout_with_sugiyama", smooth =TRUE)

Which gives access to all the layouts described in https://igraph.org/r/doc/layout_.html