Question

我在下面有以下df：

 name name..2 IGD
1 yaaA    recF  16
2 recF    yaaB  18
3 yaaD    yaaE  22
4  dck     dgk  -3
5 dnaX    yaaK  24
6 yaaK    recR  15
7  recR    yaaL  18
8  xpaC    yaaN  19
9  yaaO     tmk  -3
10 yaaQ    yaaR  13
11 yaaR    holB  12
12 holB    yaaT   3
13 yaaT    yabA  15
14 yabB    yazA -13
15 yazA    yabC -25

我正在尝试找到一种方法，将name和name..2中的值粘贴在一起，其中name..2与下一行中的name匹配，然后将其放入新的df中，该外观应如下所示：

1 yaaA recF
2 yaaD
3 dck
4 dnaX yaaK recR
5 xpaC
6 yaaO
7 yaaQ yaaR holB yaaT
8 yabB yazA

是否可以使用r函数？我曾尝试搜索SO，但尚未找到解决此问题的解决方案。预先感谢您的帮助。

Answer 1

这里的逻辑与@ Wen-Ben相似，是一种dplyr的实现方式

library(dplyr)

df %>%
  group_by(group = cumsum(name != lag(name2, default = TRUE))) %>%
  summarise(name = toString(name))

#  group name                  
#  <int> <chr>                 
#1     1 yaaA, recF            
#2     2 yaaD                  
#3     3 dck                   
#4     4 dnaX, yaaK, recR      
#5     5 xpaC                  
#6     6 yaaO                  
#7     7 yaaQ, yaaR, holB, yaaT
#8     8 yabB, yazA

主要思想是创建一个每次name != name2都会递增的分组变量。

Answer 2

在Base R中，我们使用tail head和cumsum创建组密钥，然后使用aggregate

df$id=cumsum(c(TRUE, tail(df$name,-1) != head(df$name2,-1)))

output=aggregate(name ~ id, data = df, toString)
output
  id                   name
1  1             yaaA, recF
2  2                   yaaD
3  3                    dck
4  4       dnaX, yaaK, recR
5  5                   xpaC
6  6                   yaaO
7  7 yaaQ, yaaR, holB, yaaT
8  8             yabB, yazA

Answer 3

这是在clusters内标识igraph的另一种选择

library(igraph)
library(tidyverse)
df %>%
    select(-IGD) %>%
    graph_from_data_frame() %>%
    clusters() %>%
    magrittr::extract2(1) %>%
    split(., .) %>%
    map_dfr(~tibble(x = toString(names(.x)[-length(.x)])))
## A tibble: 8 x 1
#  x
#  <chr>
#1 yaaA, recF
#2 yaaD
#3 dck
#4 dnaX, yaaK, recR
#5 xpaC
#6 yaaO
#7 yaaQ, yaaR, holB, yaaT
#8 yabB, yazA

这个想法是从igraph构造一个df[c("name", "name..2")]，然后识别连接节点的集群。簇就是组，我们要做的就是删除最后一个元素（节点）。

样本数据

df <- read.table(text =
    " name name..2 IGD
1 yaaA    recF  16
2 recF    yaaB  18
3 yaaD    yaaE  22
4  dck     dgk  -3
5 dnaX    yaaK  24
6 yaaK    recR  15
7  recR    yaaL  18
8  xpaC    yaaN  19
9  yaaO     tmk  -3
10 yaaQ    yaaR  13
11 yaaR    holB  12
12 holB    yaaT   3
13 yaaT    yabA  15
14 yabB    yazA -13
15 yazA    yabC -25", header = T)

Answer 4

我们也可以在data.table中进行

library(data.table)
setDT(df)[, .(name = toString(name)), 
      .(group = cumsum(name != shift(name2, fill = TRUE)))]
#   group                   name
#1:     1             yaaA, recF
#2:     2                   yaaD
#3:     3                    dck
#4:     4       dnaX, yaaK, recR
#5:     5                   xpaC
#6:     6                   yaaO
#7:     7 yaaQ, yaaR, holB, yaaT
#8:     8             yabB, yazA

数据

df <- structure(list(name = c("yaaA", "recF", "yaaD", "dck", "dnaX", 
"yaaK", "recR", "xpaC", "yaaO", "yaaQ", "yaaR", "holB", "yaaT", 
"yabB", "yazA"), name2 = c("recF", "yaaB", "yaaE", "dgk", "yaaK", 
"recR", "yaaL", "yaaN", "tmk", "yaaR", "holB", "yaaT", "yabA", 
 "yazA", "yabC"), IGD = c(16L, 18L, 22L, -3L, 24L, 15L, 18L, 19L, 
 -3L, 13L, 12L, 3L, 15L, -13L, -25L)), class = "data.frame",
  row.names = c("1", 
  "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", 
 "14", "15"))

将来自不同列和不同行的值粘贴到R

4 个答案:

样本数据

数据