r - 准备一个单列中包含多个节点的数据框以在 igraph 中使用

Question

我有一个数据框，格式大致如下:

df1 <- data.frame(
    x = c(
        "Ben, Ellen", 
        "BV, Ellen, Trev", 
        "Ben, Trev, Pat", 
        "Ellen, Ben", 
        "Ev, Ryan"))

其中每个观察都是一组节点( friend )。

我想对数据进行igraph处理，但我假设它首先需要看起来像这样:

df2 <- data.frame(
    x = c(
        "Ben","BV","BV","Ellen","Ben",
        "Ben","Trev","Ellen","Ev"),
    y = c(
        "Ellen","Ellen","Trev","Trev","Trev",
        "Pat","Pat","Ben","Ryan"))

如何对每个观察进行字符串分割并确保每个节点( friend )与观察中的每个其他节点链接？基本上，如何将 df1 转换为 df2？

Answer 1

使用基本函数可能更容易

ss <- strsplit(as.character(df1$x),", ")
el <- do.call("rbind", Map(function(x) t(combn(x,2)), ss))

返回的数据

      [,1]    [,2]   
 [1,] "Ben"   "Ellen"
 [2,] "BV"    "Ellen"
 [3,] "BV"    "Trev" 
 [4,] "Ellen" "Trev" 
 [5,] "Ben"   "Trev" 
 [6,] "Ben"   "Pat"  
 [7,] "Trev"  "Pat"  
 [8,] "Ellen" "Ben"  
 [9,] "Ev"    "Ryan" 

你可以用 igraph 绘制它

plot(graph_from_edgelist(el))