当行中的一个特定值匹配时，将一行复制到数据帧多次

Question

我有一个看起来像的数据框

Age    Sex    Height    Weight   Status    obesity  
17      m      5.6      36.3     single    Normal  
30      f      4.5      80.5     married   Abnormal
31      f      5.5      60.7     single    Normal
35      m      4.9      110.9    single    Very Abnormal

我想要类似的东西

Age    Sex    Height    Weight   Status    obesity  
17      m      5.6      36.3     single    Normal  
30      f      4.5      80.5     married   Abnormal
31      f      5.5      60.7     single    Normal
35      m      4.9      110.9    single    Very Abnormal
30      f      4.5      80.5     married   Abnormal
30      f      4.5      80.5     married   Abnormal
30      f      4.5      80.5     married   Abnormal
30      f      4.5      80.5     married   Abnormal
30      f      4.5      80.5     married   Abnormal
35      m      4.9      110.9    single    Very Abnormal
35      m      4.9      110.9    single    Very Abnormal
35      m      4.9      110.9    single    Very Abnormal
35      m      4.9      110.9    single    Very Abnormal
35      m      4.9      110.9    single    Very Abnormal
35      m      4.9      110.9    single    Very Abnormal

要复制数据框中的一行，必须在特定列或所示行的值之一中满足特定条件，例如obesity == Abnormal。然后行的值需要被复制给定的次数。复制可以尽可能多，例如1000。

Answer 1

这是一个通用的解决方案，可能有点过头了。这个想法是建立一个带有条件或规则的表来识别行和重复次数：

library(data.table)
rows_to_repeat <- tibble::tribble(
  ~rule,    ~n_rep,
  "obesity == 'Abnormal'",      5,
  "obesity == 'Very Abnormal'", 6,
  "Age > 30",                   2
  )
setDT(rows_to_repeat)[]

                         rule n_rep
1:      obesity == 'Abnormal'     5
2: obesity == 'Very Abnormal'     6
3:                   Age > 30     2

然后，我们遍历规则，选择满足规则的行并根据要求复制它们。最后将额外的行追加到原始数据集DT：

rbindlist(
  c(list(DT), 
    unlist(
      lapply(seq_len(nrow(rows_to_repeat)), function(.i) 
        replicate(rows_to_repeat[.i, n_rep], 
                  DT[eval(parse(text = rows_to_repeat[.i, rule]))], 
                  simplify = FALSE)),
      recursive = FALSE)))

    Age Sex Height Weight  Status       obesity
 1:  17   m    5.6   36.3  single        Normal
 2:  30   f    4.5   80.5 married      Abnormal
 3:  31   f    5.5   60.7  single        Normal
 4:  35   m    4.9  110.9  single Very Abnormal
 5:  30   f    4.5   80.5 married      Abnormal
 6:  30   f    4.5   80.5 married      Abnormal
 7:  30   f    4.5   80.5 married      Abnormal
 8:  30   f    4.5   80.5 married      Abnormal
 9:  30   f    4.5   80.5 married      Abnormal
10:  35   m    4.9  110.9  single Very Abnormal
11:  35   m    4.9  110.9  single Very Abnormal
12:  35   m    4.9  110.9  single Very Abnormal
13:  35   m    4.9  110.9  single Very Abnormal
14:  35   m    4.9  110.9  single Very Abnormal
15:  35   m    4.9  110.9  single Very Abnormal
16:  31   f    5.5   60.7  single        Normal
17:  35   m    4.9  110.9  single Very Abnormal
18:  31   f    5.5   60.7  single        Normal
19:  35   m    4.9  110.9  single Very Abnormal

数据

library(data.table)
DT <- readr::read_table(
  "Age    Sex    Height    Weight   Status    obesity  
  17      m      5.6      36.3     single    Normal  
  30      f      4.5      80.5     married   Abnormal
  31      f      5.5      60.7     single    Normal
  35      m      4.9      110.9    single    Very Abnormal"
)
setDT(DT)

Answer 2

你可以这样做：

 df %>%
   filter(grepl('Abnormal',obesity)) %>%
   slice(rep(1:n(), each=5)) %>%
   rbind(df)

给出（每行重复 5 次）：

   Age Sex Height  Status       obesity
1   30   f   80.5 married      Abnormal
2   30   f   80.5 married      Abnormal
3   30   f   80.5 married      Abnormal
4   30   f   80.5 married      Abnormal
5   30   f   80.5 married      Abnormal
6   35   m  110.9  single Very Abnormal
7   35   m  110.9  single Very Abnormal
8   35   m  110.9  single Very Abnormal
9   35   m  110.9  single Very Abnormal
10  35   m  110.9  single Very Abnormal
11  17   m   36.3  single        Normal
12  30   f   80.5 married      Abnormal
13  31   f   60.7  single        Normal
14  35   m  110.9  single Very Abnormal