Select edges in a graph
Select edges from a graph object of class dgr_graph
.
select_edges( graph, conditions = NULL, set_op = "union", from = NULL, to = NULL, edges = NULL )
graph |
A graph object of class |
conditions |
An option to use filtering conditions for the retrieval of edges. |
set_op |
The set operation to perform upon consecutive selections of
graph nodes. This can either be as a |
from |
An optional vector of node IDs from which the edge is outgoing for filtering the list of edges present in the graph. |
to |
An optional vector of node IDs to which the edge is incoming for filtering the list of edges present in the graph. |
edges |
An optional vector of edge IDs for filtering the list of edges present in the graph. |
A graph object of class dgr_graph
.
# Create a node data frame (ndf) ndf <- create_node_df( n = 4, type = "basic", label = TRUE, value = c(3.5, 2.6, 9.4, 2.7)) # Create an edge data frame (edf) edf <- create_edge_df( from = c(1, 2, 3), to = c(4, 3, 1), rel = c("a", "z", "a"), value = c(6.4, 2.9, 5.0)) # Create a graph with the ndf and edf graph <- create_graph( nodes_df = ndf, edges_df = edf) # Explicitly select the edge `1`->`4` graph <- graph %>% select_edges( from = 1, to = 4) # Verify that an edge selection has been made # using the `get_selection()` function graph %>% get_selection() # Select edges based on the relationship label # being `z` graph <- graph %>% clear_selection() %>% select_edges( conditions = rel == "z") # Verify that an edge selection has been made, and # recall that the `2`->`3` edge uniquely has the # `z` relationship label graph %>% get_selection() # Select edges based on the edge value attribute # being greater than 3.0 (first clearing the current # selection of edges) graph <- graph %>% clear_selection() %>% select_edges( conditions = value > 3.0) # Verify that the correct edge selection has been # made; in this case, edges `1`->`4` and # `3`->`1` have values for `value` > 3.0 graph %>% get_selection()
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.