library(igraph)
library(ADAPTSNA)10 Two Mode Networks - Edgelists
This is a continuation of the previous section working with two mode networks. The same learning elements apply here as applied in Chapter 9.
There is a slightly different approach to bringing in Two mode network data from an Edgelist than from an Adjacency matrix. Instead of a two mode matrix, you may have edgelist data in two mode format.
hp_tm_edgelist <- load_data("Harry Potter_Two_Mode_Edge.csv")
head(hp_tm_edgelist) character group
1 Albus Dumbledore Phoenix
2 Albus Dumbledore Prefect
3 Albus Dumbledore Gryffindor
4 Remus Lupin Phoenix
5 Remus Lupin Prefect
6 Remus Lupin GryffindorDifferent from the adjacency matrix, this edge list has one type of node in one column and the other type in the second column.
This is the same process as any other network created from an edgelist. Directed is set to FALSE because there is not really a direction between individuals and groups. Rather, the ties are a marker of affiliation.
hp_tm_net <- graph_from_data_frame(hp_tm_edgelist, directed = FALSE)Let’s check to see if this is actually a two mode network using bipartite_mapping. This function goes through the edgelist an ensures that the columns have distinct nodes in them (i.e. it is truly a bipartite or two mode network).
bipartite_mapping(hp_tm_net)$res
[1] TRUE
$type
Albus Dumbledore Remus Lupin Molly Weasley
FALSE FALSE FALSE
Siruis Black Severus Snape Alastor Moody
FALSE FALSE FALSE
Minerva McGonagall Rubeus Hagrid Kingsley Shacklebolt
FALSE FALSE FALSE
Nymphadora Tonks Mundungus Fletcher Dedalus Diggle
FALSE FALSE FALSE
Elphias Dode Aberforth Dumbledore Arabella Figg
FALSE FALSE FALSE
Emmeline Vance Sturgis Podmore Hestia Jones
FALSE FALSE FALSE
Aurthur Weasley Bill Weasley Charlie Weasley
FALSE FALSE FALSE
Hermione Granger Harry Potter Cho Chang
FALSE FALSE FALSE
Ron Weasley Lavendar Brown George Weasley
FALSE FALSE FALSE
Fred Weasley Neville Longbottom Colin Creevey
FALSE FALSE FALSE
Luna Lovegood Dean Thomas Katie Bell
FALSE FALSE FALSE
Angelina Johnson Hannah Abbott Lee Jordon
FALSE FALSE FALSE
Anthony Goldstein Ernie Macmilan Justin Finch-Fletchley
FALSE FALSE FALSE
Padma Patil Seamus Finnigan Susan Bones
FALSE FALSE FALSE
Marietta Edgecombe Alicia Spinnet Dennis Creevey
FALSE FALSE FALSE
Ginny Weasley Parvati Patil Nigel Wolpert
FALSE FALSE FALSE
Cormac McLaggen Romilda Vane Michael Corner
FALSE FALSE FALSE
Terry Boot Maisy Reynolds Leanne
FALSE FALSE FALSE
Zacharias Smith Luca Caruso Alice Toplin
FALSE FALSE FALSE
James Potter Lilly Potter Peter Petigrew
FALSE FALSE FALSE
Fabian Prewett Gideon Prewett Frank Longbottom
FALSE FALSE FALSE
Alice Longbottom Edgar Bones Benjy Fenwick
FALSE FALSE FALSE
Caradoc Dearborn Dorcas Meadowes Marlene McKinnon
FALSE FALSE FALSE
Fleur Delacour Bellatrix Lestrange Lucius Malfoy
FALSE FALSE FALSE
Igor Karkaroff Regulus Black Barty Crouch Jr
FALSE FALSE FALSE
Antonin Dolohov Thorfinn Rowle Augustus Rookwood
FALSE FALSE FALSE
Evan Rosier Walden Macnair Alecto Carrow
FALSE FALSE FALSE
Amycus Carrow Avery Jnr Corban Yaxley
FALSE FALSE FALSE
Crabbe Snr Draco Malfoy Gibbon
FALSE FALSE FALSE
Goyle Snr Jugson Mulciber Snr
FALSE FALSE FALSE
Mulciber Jnr Nott Snr Rabastan Lestrange
FALSE FALSE FALSE
Rodolphus Lestrange Tavers Pansy Parkinson
FALSE FALSE FALSE
Millicent Bulstrode Vincent Crabbe Gregory Goyle
FALSE FALSE FALSE
Graham Montague Cassius Warrington Argus Filch
FALSE FALSE FALSE
Dolores Umbridge Jane Court Gabriel Truman
FALSE FALSE FALSE
Cedric Diggory Constance Pickering Natalie Kathryn
FALSE FALSE FALSE
Tom Riddle Felix Rosier Gemma Farley
FALSE FALSE FALSE
Penelope Padgett Rodrick Lyme Annalena Murk
FALSE FALSE FALSE
Angelica Cole Percy Weasley Freddie Clemmons
FALSE FALSE FALSE
Dani Caroll Marcus Turner Penelope Clearwater
FALSE FALSE FALSE
Robert Hillard Chester Davies Phoenix
FALSE FALSE TRUE
Prefect Gryffindor Death.Eaters
TRUE TRUE TRUE
Slytherin Hufflepuff Ravenclaw
TRUE TRUE TRUE
Dumbeldore.s.Army Inquisitorial.Squad
TRUE TRUE It recognises that there are two types of node in this object, so we can set that as a vertex characteristic. In turn, this changes the network from a one mode to a two mode network.
V(hp_tm_net)$type <- bipartite_mapping(hp_tm_net)$typeNow we have changed it into a two mode network and added the characteristic “type” that we are familiar with from working with an adjacency matrix like we did in Chapter 8.
V(hp_tm_net)$type [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[13] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[25] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[37] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[49] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[61] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[73] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[85] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[97] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[109] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[121] FALSE FALSE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUEYou see the true and false statements, as we expect to see. Since this is the case, we will need to use the same +1 alteration to the arguments in our visualisation.
When we plot it, it looks how we expect!
shapes <- c("circle", "square")
colors <-c("green", "orange")
plot(hp_tm_net, vertex.color=colors[V(hp_tm_net)$type+1],
vertex.shape=shapes[V(hp_tm_net)$type+1], vertex.label = NA)
10.1 Summary
Final things to remember here about two mode networks are that, just like one mode network data, they can be stored as either a matrix or an edgelist. When bringing in a two mode network from an adjacency matrix, you can convert the data into a network object using the graph_from_biadjacency_matrix() function so long as you have set it as matrix. When bringing in a two mode network from an edgelist, you have to set the bipartite characteristic.