Rmd analysis

.gitignore

@@ -1,4 +1,4 @@
-.Rproj.user
-.Rhistory
-.RData
-.Ruserdata
+.Rproj.user
+.Rhistory
+.RData
+.Ruserdata

README.md

@@ -1,27 +1,27 @@
-# SEP Graph
-
-A graphical representation and analysis of the Stanford Encyclopedia of Philosophy.
-
-## Project Description
-
-The [Stanford Encyclopedia of Philosophy][SEP] (hereafter, SEP) is a "dynamic reference work" that has a significant influence in philosophy. This project is an attempt to represent the data in the SEP graphically. The methodology of the representation is as follows:
-
-+ Each article in the SEP is a node.
-+ A directed edge connects a source and target node iff the article represented by the source contains a hyperlink to the article represented by the target.
-+ Otherwise, nodes are not directly connected.
-
-## Visualizations
-
-[I'm hosting the visualizations for the SEP here.][1]
-
-
-## Thanks
-
-Thanks to the developers of [threejs][2], [statnet][3], [d3][4], and [igraph][5] (and everyone working on R I forgot to mention). 
-
-[SEP]: https://plato.stanford.edu
-[1]: https://adamdedwards.github.io/sep-graph
-[2]: https://github.com/bwlewis/rthreejs
-[3]: http://statnet.csde.washington.edu/
-[4]: https://github.com/d3
-[5]: https://github.com/igraph
+# SEP Graph
+
+A graphical representation and analysis of the Stanford Encyclopedia of Philosophy.
+
+## Project Description
+
+The [Stanford Encyclopedia of Philosophy][SEP] (hereafter, SEP) is a "dynamic reference work" that has a significant influence in philosophy. This project is an attempt to represent the data in the SEP graphically. The methodology of the representation is as follows:
+
++ Each article in the SEP is a node.
++ A directed edge connects a source and target node iff the article represented by the source contains a hyperlink to the article represented by the target.
++ Otherwise, nodes are not directly connected.
+
+## Visualizations
+
+[I'm hosting the visualizations for the SEP here.][1]
+
+
+## Thanks
+
+Thanks to the developers of [threejs][2], [statnet][3], [d3][4], and [igraph][5] (and everyone working on R I forgot to mention). 
+
+[SEP]: https://plato.stanford.edu
+[1]: https://adamdedwards.github.io/sep-graph
+[2]: https://github.com/bwlewis/rthreejs
+[3]: http://statnet.csde.washington.edu/
+[4]: https://github.com/d3
+[5]: https://github.com/igraph

analysis.Rmd

@@ -0,0 +1,147 @@
+---
+title: "SEP Analysis"
+author: "Adam Edwards"
+date: "November 12, 2018"
+output: html_document
+---
+
+```{r setup, include=FALSE}
+knitr::opts_chunk$set(echo = TRUE)
+
+list.of.packages <- c("igraph","networkD3","visNetwork","RColorBrewer")
+new.packages <- list.of.packages[!(list.of.packages %in% installed.packages()[,"Package"])]
+if(length(new.packages)>0) install.packages(new.packages,repos = "http://cran.us.r-project.org")
+
+#library(shiny)
+library(igraph)
+#library(threejs)
+library(networkD3)
+library(visNetwork)
+library(RColorBrewer)
+#library(extrafont)
+
+```
+
+## R Code
+
+First, we need to load the data into R and create a data frame.
+
+```{r load data}
+data <- read.csv("data/fall1998_edge_list.txt", header = TRUE)
+as.data.frame(data)
+```
+
+Once the data is loaded, we can create an "igraph" object. We also want to create an edge weight of 1 for each link so that we can combine multiple links into linearly weighted edges. The `simplify` function will also eliminate self-loops.
+
+```{r create igraph}
+g <- graph_from_data_frame(as.data.frame(data), directed = TRUE)
+E(g)$weight <- 1
+g <- simplify(g, edge.attr.comb=list(weight="sum"))
+```
+
+The igraph package has several functions that allow us to compute vertex-level and edge-level attributes for the graph. For now, we'll add node labels, degree, betweenness, and group membership using the `cluster_walktrap` function to assign group membership based on short, randow walks around the graph
+
+```{r igraph attributes,echo=FALSE, warning=FALSE}
+V(g)$label <- V(g)$name
+V(g)$degree <- degree(g)
+V(g)$betweenness <- betweenness(g,V(g),directed=TRUE)
+#V(g)$closeness <- closeness(g,mode="all")
+#V(g)$eigenvector <- eigen_centrality(g, directed = TRUE, weights=E(g)$weight)[[1]]
+V(g)$group <- membership(cluster_walktrap(g))
+
+pc <- cluster_walktrap(g)
+colors <- colorRampPalette(brewer.pal(11, "Spectral"))(max(V(g)$group))
+```
+
+We can see the graph object with these new properties and the communities that the clustering algorithm has identified. One interesting feature of this data set is how the clustering algorithm identifies what I take to be the major areas that professional philosophers identify as subfields. There's no expert knowledge applied to these graphs; it's an emergent feature of how the SEP is organized.
+
+```{r igraph}
+g
+
+communities(cluster_walktrap(g))
+```
+
+## Some Charts
+
+We can do some pre-analysis of the graph by looking at the distributions of the different properties:
+
+```{r pre-analysis, echo=FALSE}
+hist <- hist(V(g)$degree,xlab = "Degree",main = "Degree Centrality",col=colors[3])
+V(g)$name[match(tail(sort(V(g)$degree),10),V(g)$degree)]
+
+hist <- hist(V(g)$betweenness,xlab = "Betweenness",main = "Betweenness Centrality",col=colors[9])
+V(g)$name[match(tail(sort(V(g)$betweenness),10),V(g)$betweenness)]
+
+#hist <- hist(V(g)$eigenvector,xlab = "Eigenvector",main = "Eigenvector Centrality")
+#V(g)$name[match(tail(sort(V(g)$eigenvector),10),V(g)$eigenvector)]
+```
+
+
+## Visualizations
+
+With igraph
+
+```{r igraph viz, echo=FALSE}
+
+V(g)$size <- 10
+V(g)$color <- colors[V(g)$group]
+E(g)$arrow.size <- .2
+E(g)$edge.color <- "gray80"
+
+
+plot(g,vertex.label=NA)
+
+plot(pc,g,vertex.label=NA)
+
+```
+
+```{r threejs viz, echo=FALSE}
+
+#l <- layout_with_kk(g,dim=3)
+#l <- norm_coords(l, ymin=-1, ymax=1, xmin=-1, xmax=1)
+
+#gjs <- graphjs(g,layout=l*0.2,vertex.label=V(g)$label,
+#                 minx = NULL, maxx = 300, miny = NULL, maxy = 300, minz = NULL, maxz = 300)
+#gjs
+```
+
+With NetworkD3
+
+```{r networkd3, echo=FALSE}
+
+gd3 <- igraph_to_networkD3(g,V(g)$group)
+gd3$nodes$deg <- as.character(degree(g, v = V(g)))
+gd3$nodes$btw <- as.character(betweenness(g, v = V(g), directed = TRUE))
+#gd3$nodes$clo <-as.character(closeness(g, v = V(g),mode="all"))
+#gd3$nodes$eig <-as.character(eigen_centrality(g, directed = TRUE)[[1]])
+gd3$nodes$size <- as.character("1")
+gd3$nodes$id <- row.names(gd3$nodes)
+
+
+
+nd3 <- forceNetwork(Links       = gd3$links,
+                    Nodes       = gd3$nodes,
+                    Source      = "source",
+                    Target      = "target",
+                    Group       = "group",
+                    NodeID      = "name", 
+                    Nodesize    = "deg",
+                    arrows      = FALSE,
+                    legend      = FALSE,
+                    bounded     = TRUE,
+                    opacity     = 1,
+                    opacityNoHover = .4)
+nd3
+```
+
+With visNetwork
+
+```{r visNetwork, echo=FALSE}
+
+visIgraph(g)
+
+```
+
+## References/Acknowledgements
+
+Ognyanova, K. (2018) Network visualization with R. Retrieved from www.kateto.net/network-visualization.

assets/sep.js

@@ -1,17 +1,17 @@
-
-var community = 5;
-
-window.onload = function(){document.getElementById('comm_frame').src = "pages/graphs/community_5.html";};
-
-function prevcomm() {
-    if(community > 1) { community = community - 1; }
-    else { community = 18; }
-
-    document.getElementById('comm_frame').src = "pages/graphs/community_"+community+".html";
-}
-
-function nextcomm() {
-  if(community <= 19) { community = community + 1; }
-  else { community = 1; }
-    document.getElementById('comm_frame').src = "pages/graphs/community_"+community+".html";
-}
+
+var community = 5;
+
+window.onload = function(){document.getElementById('comm_frame').src = "pages/graphs/community_5.html";};
+
+function prevcomm() {
+    if(community > 1) { community = community - 1; }
+    else { community = 18; }
+
+    document.getElementById('comm_frame').src = "pages/graphs/community_"+community+".html";
+}
+
+function nextcomm() {
+  if(community <= 19) { community = community + 1; }
+  else { community = 1; }
+    document.getElementById('comm_frame').src = "pages/graphs/community_"+community+".html";
+}

assets/styles.css

@@ -1,40 +1,41 @@
-h1 { font-family: 'Inconsolata', monospace; font-size: 36px; font-style: normal; font-variant: normal; font-weight: 700; line-height: 41px; }
-h2 { font-family: 'Inconsolata', monospace; font-size: 30px; font-style: normal; font-variant: normal; font-weight: 700; line-height: 35px; }
-h3 { font-family: 'Inconsolata', monospace; font-size: 27px; font-style: normal; font-variant: normal; font-weight: 700; line-height: 32px; }
-
-p { font-family: 'Inconsolata', monospace; font-size: 18px; font-style: normal; font-variant: normal; font-weight: 400; line-height: 1.2em; }
-
-iframe {
-  border: solid 2px black;
-  width: 80%;
-  padding: 20px;
-  display:block;
-  margin:auto;
-}
-
-button {
-    margin: 0px 15px;
-    border: 0;
-    border-radius: 8px;
-    padding: 10px 20px;
-
-    text-align: center;
-    text-decoration: none;
-    display: inline-block;
-    font-size: 16px;
-
-    -webkit-transition-duration: 0.4s;
-    transition-duration: 0.4s;
-}
-
-button:hover {
-    background-color: #3A3A3A;
-    color: white;
-}
-
-#button-row {
-    width: 500px ;
-    margin-left: auto ;
-    margin-right: auto ;
-    padding: 20px;
-}
+h1 { font-family: 'Inconsolata', monospace; font-size: 36px; font-style: normal; font-variant: normal; font-weight: 700; line-height: 41px; }
+h2 { font-family: 'Inconsolata', monospace; font-size: 30px; font-style: normal; font-variant: normal; font-weight: 700; line-height: 35px; }
+h3 { font-family: 'Inconsolata', monospace; font-size: 27px; font-style: normal; font-variant: normal; font-weight: 700; line-height: 32px; }
+
+p { font-family: 'Inconsolata', monospace; font-size: 18px; font-style: normal; font-variant: normal; font-weight: 400; line-height: 1.2em; }
+
+iframe {
+  border: solid 2px black;
+  width: 80%;
+  padding: 20px;
+  display:block;
+  margin:auto;
+}
+
+button {
+    margin: 0px 15px;
+    border: 0;
+    border-radius: 8px;
+    padding: 10px 20px;
+
+    text-align: center;
+    text-decoration: none;
+    display: inline-block;
+    font-family: 'Inconsolata', monospace;
+    font-size: 16px;
+
+    -webkit-transition-duration: 0.4s;
+    transition-duration: 0.4s;
+}
+
+button:hover {
+    background-color: #3A3A3A;
+    color: white;
+}
+
+#button-row {
+    width: 500px ;
+    margin-left: auto ;
+    margin-right: auto ;
+    padding: 20px;
+}