如何绘制有向网络图，在R的边两边(通过iGraph、网络或其他包)的文本？

Question

如何绘制有向网络图，在R的边两边(通过iGraph、网络或其他包)的文本？

​

​

变量(在向量自回归模型(VAR)中)在节点中。

格兰杰因果关系(F统计量的p值)位于边的两边。我把p值放在每个箭头的开头。我用更多的强调引出了近乎重要的因果关系.

我不知道如何通过iGraph或网络包或其他方式获得这样的网络图。

Answer 1

您可以同时使用edge.label和vertex.label选项。

library(igraph)
el <- data.frame(sender = c("lnbist1f","lnbist1f","lnbist1f",
                            "kur1f","kur1f","kur1f",
                            "lnaltin","lnaltin","lnaltin",
                            "mfaiz1f","mfaiz1f","mfaiz1f"),
                 receiver = c("mfaiz1f","lnaltin","kur1f",
                              "lnbist1f","lnaltin","mfaiz1f",
                              "mfaiz1f","lnbist1f","kur1f",
                              "lnbist1f","lnaltin","kur1f"),
                 pval = c(0.5,0.6,0.1, #I just typed random p-vals here
                          0.45,0.88,0.24,
                          0.12,0.51,0.99,
                          0.001,0.056,0.123)
                 )
arrows = c(2,1,0,0,1,1,1,0,0,2,1,1)
el <- as.matrix(el)
g <- graph_from_edgelist(el[,1:2], directed = T)
coordinates <- matrix(c(4, 4, 1, 1, 4, -2,7,1), nrow = 4, byrow=TRUE)
plot(g, edge.label=el[,3], 
     vertex.shape="crectangle", 
     vertex.size=45,
     edge.arrow.mode=arrows,
     layout = coordinates)

edge.arrow.mode允许您控制箭头。您可以使用edge.label.y和edge.label.x选项移动边缘标签。

​

​

Answer 2

阅读图形绘图函数的文档化，似乎没有太多的选项可以操作边缘标签，但是您可以使用label.dist和label.degree来操作顶点标签。你可以这样做：

g <- graph.empty(n = 3) 
g <- graph(c(1, 2, 3, 2, 1, 3), directed = TRUE)
labels <- c("A", "B", "C", "D", "E", "F")
coords <- matrix(c(1, 1, 2, 2, 3, 3), nrow = 3)
plot(g, layout = coords, vertex.label = labels, vertex.label.dist = 3.5, vertex.label.degree = c(-pi/4, pi/2, pi, -pi/2, 0, 3*pi/4))

图表