merico34 · September 18, 2016 05:57
diff --git a/00_keywords.txt b/00_keywords.txt
 name	keywords
 Felicitas	Development, Corruption, Housing, Transportation, Religion
 Lucile	Inequality, Migration, Race and Gender, Community/Neighborhood
 Corentin	Inequality, Religion, Migration, Poverty, Gender
 Charlotte	Energy, Environment, Smart Cities, Conflict, Gender
 Alice	Environment, Gender, Migration, Urban Studies
 Miranda	Politics, Gender, Human Rights
 Margaux	Inequality, Urban Studies, Migration, Human Rights
 Marina	Critical Theory, Social Inequality, Race and Gender
 Cosima	Race and Gender and Sexuality, Spatial Inequality, Migration 
 Isabelle	Slums, Poverty, Housing, Inequality
 Elena	Race and Gender, Urban Inequality, Immigrant Integration, Collective Action
 Gabriel	Environment, Collective Action, Urban Studies
 Alexander	Transportation, Environmental Justice, Community Empowerment
 Akhil	Tourism, Transportation, Politics
 Gabriella	Multicultural Cohabitation, Environment, Radical Politics and Space Appropriation
 Mohamed	Transport, Inequality, Poverty, Spatial Inequality
 Lucien	Inequality, Poverty, Religious Segregation, Migration
diff --git a/01_packages.r b/01_packages.r
 library(dplyr)
 library(ggnetwork)
 library(ggplot2)
 library(igraph) # keep after ggnetwork
 library(intergraph)
 library(readr)
 library(stringr)
 library(tnet)
diff --git a/02_load.r b/02_load.r
 # students data
 d <- read_tsv("keywords.txt") %>%
  arrange(name)
diff --git a/03_students.r b/03_students.r
 # student names
 s <- d$name
diff --git a/04_keywords.r b/04_keywords.r
 # student keywords
 k <- d$keywords %>%
  str_split(", | and ") %>%
  unlist %>%
  unique %>%
  sort
diff --git a/05_incidence_matrix.r b/05_incidence_matrix.r
 # incidence matrix
 m <- sapply(d$keywords, function(x) {
  str_split(x, ", | and ") %>% unlist
 }) %>%
  lapply(function(x) { as.integer(k %in% x) }) %>%
  unlist %>%
  matrix(ncol = length(k), byrow = TRUE)
diff --git a/06_name_dimensions.r b/06_name_dimensions.r
 rownames(m) <- s # student names
 colnames(m) <- k # student keywords
diff --git a/07_plot.r b/07_plot.r
 # flatten out the network to a data frame
 n <- ggnetwork(m)

 # single out the primary mode (students)
 n$student <- n$vertex.names %in% s

 # plot two-mode network
 ggplot(n, aes(x, y, xend = xend, yend = yend)) + 
  geom_edges(color = "grey50") + 
  geom_nodelabel(data = n[ !n$student, ],
                 aes(label = vertex.names),
                 color = "grey50", label.size = NA) +
  geom_nodelabel(data = n[ n$student, ],
                 aes(label = vertex.names),
                 color = "steelblue", fontface = "bold") +
  theme_blank()
diff --git a/08_collapse.r b/08_collapse.r
 # collapsed one-mode network, using student names
 n <- m %*% t(m) %>%
  graph_from_adjacency_matrix(mode = "undirected", diag = FALSE, weighted = TRUE)
diff --git a/09_degree.r b/09_degree.r
 # weighted degree
 V(n)degree <- as_adjacency_matrix(n, attr = "weight", sparse = FALSE) %>%
  degree_w %>%
  .[, 3]
diff --git a/10_clusters.r b/10_clusters.r
 V(n)$group <- cluster_louvain(n) %>%
  membership %>%
  as.character
diff --git a/11_plot.r b/11_plot.r
 ggplot(ggnetwork(n, layout = "kamadakawai"),
       aes(x, y, xend = xend, yend = yend)) +
  geom_edges(aes(alpha = weight)) +
  geom_nodelabel(aes(label = vertex.names, size = degree, color = group)) +
  scale_alpha_continuous(guide = FALSE) +
  scale_color_brewer(palette = "Set1", guide = FALSE) +
  scale_size_continuous(range = c(3, 6), guide = FALSE) +
  theme_blank()
	name keywords
	Felicitas Development, Corruption, Housing, Transportation, Religion
	Lucile Inequality, Migration, Race and Gender, Community/Neighborhood
	Corentin Inequality, Religion, Migration, Poverty, Gender
	Charlotte Energy, Environment, Smart Cities, Conflict, Gender
	Alice Environment, Gender, Migration, Urban Studies
	Miranda Politics, Gender, Human Rights
	Margaux Inequality, Urban Studies, Migration, Human Rights
	Marina Critical Theory, Social Inequality, Race and Gender
	Cosima Race and Gender and Sexuality, Spatial Inequality, Migration
	Isabelle Slums, Poverty, Housing, Inequality
	Elena Race and Gender, Urban Inequality, Immigrant Integration, Collective Action
	Gabriel Environment, Collective Action, Urban Studies
	Alexander Transportation, Environmental Justice, Community Empowerment
	Akhil Tourism, Transportation, Politics
	Gabriella Multicultural Cohabitation, Environment, Radical Politics and Space Appropriation
	Mohamed Transport, Inequality, Poverty, Spatial Inequality
	Lucien Inequality, Poverty, Religious Segregation, Migration
	library(dplyr)
	library(ggnetwork)
	library(ggplot2)
	library(igraph) # keep after ggnetwork
	library(intergraph)
	library(readr)
	library(stringr)
	library(tnet)
	# students data
	d <- read_tsv("keywords.txt") %>%
	arrange(name)
	# student keywords
	k <- d$keywords %>%
	str_split(", \| and ") %>%
	unlist %>%
	unique %>%
	sort
	# incidence matrix
	m <- sapply(d$keywords, function(x) {
	str_split(x, ", \| and ") %>% unlist
	}) %>%
	lapply(function(x) { as.integer(k %in% x) }) %>%
	unlist %>%
	matrix(ncol = length(k), byrow = TRUE)
	rownames(m) <- s # student names
	colnames(m) <- k # student keywords
	# flatten out the network to a data frame
	n <- ggnetwork(m)

	# single out the primary mode (students)
	n$student <- n$vertex.names %in% s

	# plot two-mode network
	ggplot(n, aes(x, y, xend = xend, yend = yend)) +
	geom_edges(color = "grey50") +
	geom_nodelabel(data = n[ !n$student, ],
	aes(label = vertex.names),
	color = "grey50", label.size = NA) +
	geom_nodelabel(data = n[ n$student, ],
	aes(label = vertex.names),
	color = "steelblue", fontface = "bold") +
	theme_blank()
	# collapsed one-mode network, using student names
	n <- m %*% t(m) %>%
	graph_from_adjacency_matrix(mode = "undirected", diag = FALSE, weighted = TRUE)
	# weighted degree
	V(n)degree <- as_adjacency_matrix(n, attr = "weight", sparse = FALSE) %>%
	degree_w %>%
	.[, 3]
	V(n)$group <- cluster_louvain(n) %>%
	membership %>%
	as.character