r - 在 r 中按组标记非连续值

Question

我有一个由多个组组成的数据集，这些组具有连续的编号容器(每个组中的容器数量不一定相同)以及 bool 存在/不存在值。我希望能够生成一些输出，指示其中存在不连续的“当前”值的组。

最小的表示如下:

x <- NULL
x$group <- c(rep("A",4),rep("B", 5), rep("C",4))
x$bin <- c(1,2,3,4,1,2,3,4,5,1,2,3,4)
x$status <- c("absent", "present", "absent", "present", "absent", "present", "present", "absent", "absent", "absent", "absent", "present", "present")

as.data.frame(x)

   group bin  status
1      A   1  absent
2      A   2 present
3      A   3  absent
4      A   4 present
5      B   1  absent
6      B   2 present
7      B   3 present
8      B   4  absent
9      B   5  absent
10     C   1  absent
11     C   2  absent
12     C   3 present
13     C   4 present

输出可能是同一数据框中带有标志的另一列，

   group bin  status flag
1      A   1  absent    1
2      A   2 present    1
3      A   3  absent    1
4      A   4 present    1
5      B   1  absent    0
6      B   2 present    0
7      B   3 present    0
8      B   4  absent    0
9      B   5  absent    0
10     C   1  absent    0
11     C   2  absent    0
12     C   3 present    0
13     C   4 present    0

单独的数据框或矩阵，例如:

  group  flag
1     A  TRUE
2     B FALSE
3     C FALSE

或列表:

> flagged_groups
[1] "A"

我觉得通过写这篇文章，我已经整理出了实现这一目标所需要做的一些事情，但我很想听听您的想法，以简洁(整洁)的方式提炼我的数据。

Answer 1

你可以这样做:

library(dplyr)

df %>%
  group_by(group) %>%
  mutate(flag = +any(diff(row_number()[status == "present"]) != 1))

# A tibble: 14 x 4
# Groups:   group [4]
   group   bin status   flag
   <chr> <dbl> <chr>   <int>
 1 A         1 absent      1
 2 A         2 present     1
 3 A         3 absent      1
 4 A         4 present     1
 5 B         1 absent      0
 6 B         2 present     0
 7 B         3 present     0
 8 B         4 absent      0
 9 B         5 absent      0
10 C         1 absent      0
11 C         2 absent      0
12 C         3 present     0
13 C         4 present     0