R:使用列中的逗号分隔值制作网络图

R: Making a network graph using comma separated values in a column

有什么方法可以制作这样的网络图,例如:

Colour               IDs
Red                  12,14,15
Green                17,20,25
Blue                 20,39,46

其中 (12,14,15) 是网络,(17,20,25) 是网络,(20,39,46) 也是网络,但是 20 来自 GreenBlue 将连接。

是否有此功能,或者我是否必须手动拆分逗号分隔值?

谢谢!

library(data.table)
library(igraph)

DT <- fread("Colour               IDs
Red                  12,14,15
Green                17,20,25
Blue                 20,39,46")


L <- lapply(strsplit(DT$IDs, ","), combn, 2, simplify = FALSE)
names(L) <- DT$Colour
edges <- rbindlist(lapply(L, transpose), idcol = "color")

g <- graph_from_data_frame(edges[, 2:3], directed = FALSE)
E(g)$color <- tolower(edges$color)
plot(g)

您可以试试下面的代码

library(data.table)
library(igraph)

setDT(df)[
    ,
    as.data.frame(t(combn(strsplit(IDs, ",")[[1]], 2))),
    Colour
][
    ,
    .(V1, V2, color = Colour)
] %>%
    graph_from_data_frame(directed = FALSE) %>%
    plot()

其中 combn 生成边列表

使用 tidyverseigraph 库:

library(tidyverse)
library(igraph)

dat %>% 
  separate_rows(IDs) %>% 
  group_by(Colour) %>% 
  do(data.frame(t(combn(.$IDs, 2)))) %>% 
  select(X1, X2, color = Colour) %>% 
  graph_from_data_frame(directed = F) %>% 
  plot()


当某些组只有一个节点时的替代方案:

dat %>% 
  separate_rows(IDs) %>% 
  group_by(Colour) %>% 
  summarise(new = ifelse(n() > 1, paste(combn(IDs, 2), collapse = "-"), as.character(IDs))) %>% 
  separate_rows(new, sep = "(?:[^-]*(?:-[^-]*){1})\K-") %>% 
  separate(new, into = c("X1", "X2")) %>% 
  select(X1, X2, color = Colour) %>% 
  graph_from_data_frame(directed = F) %>% 
  plot()