R通過結合兩個變量的公數值來重塑數據

Question

我想通過合並兩個變量來重塑數據框。 例如：

這是一個新數據：

    dat = data.frame(
    var1 = c("a", "a", "a", "Emily", "b", "Bob", "c"),
    var2 = c("Jhon", "Emily", "Julie", "Angela", "Bob", "Paul", "Paul"),
    stringsAsFactors = F
)

排除的輸出：

  #   var1   var2   var3   var4   var5 
  # 1    a   Jhon  Emily  Julie Angela
  # 2    b    Bob   Paul      c   <NA>

Answer 1

使用base R，您可以執行以下操作：

relation=function(dat){

 .relation=function(x){
    k = unique(sort(c(dat[dat[, 1] %in% x, 2], x, dat[dat[, 2] %in% x, 1])))
    if(setequal(x,k)) toString(k) else .relation(k)}

  grp = sapply(unique(dat[,1]), .relation)
  read.table(text = unique(grp), fill=T, sep=",")
}

relation(dat)
  V1      V2     V3    V4     V5
1  a  Angela  Emily  Jhon  Julie
2  b     Bob      c  Paul   

Answer 2

dat = data.frame(var1 = c("a", "a", "a", "Emily", "b", "Bob"), 
                 var2 = c("Jhon", "Emily", "Julie", "Angela", "Bob", "Paul"))

library(igraph)
g <- graph_from_data_frame(dat)
plot(g)
starts <- V(g)[degree(g, mode = "in") == 0] 
finals <- V(g)[degree(g, mode = "out") == 0]
res <- lapply(starts, function(x) unique(names(unlist(all_simple_paths(g, 
                                                            from = x, 
                                                            to = finals, 
                                                            mode = "out")))))
res

# matrix/data frame (?)
max_len <- max(sapply(res, length))
data.frame(do.call(rbind, lapply(res, function(x) c(x, rep(NA, max_len - length(x))))))

Answer 3

我提出了一種解決方案，該解決方案首先“清除”數據結構，然后使用dcast對其進行dcast 。

library(data.table)

dt.dat <- data.table(dat)

# Cleaning the dataset by adding the persons not assigned to a group by the connection over names
dt.dat.complete <- 
rbindlist(list(dt.dat[!(var1 %in% merge(dt.dat, dt.dat, by.x = "var2", by.y = "var1")[,var2]),]
          , 
merge(dt.dat, dt.dat, by.x = "var2", by.y = "var1")[, .(var1, var2.y)]
))

# Add sequence for the column names
dt.dat.complete[,seq := seq_len(.N), 
                    by=var1]

dcast.data.table(dt.dat.complete, var1  ~ paste0("col",seq) + seq,fun.aggregate = NULL, value.var = "var2")

     var1 col1_1 col2_2 col3_3 col4_4
1:    a   Jhon  Emily  Julie Angela
2:    b    Bob   Paul   <NA>   <NA>