在R中置换一列[重复]

Question

我有一个数据集，df:

Action                        Date

Begin                         3/16/2020 12:35:47 PM
End                           3/16/2020 12:35:49 PM
Begin                         3/16/2020 01:35:47 PM
End                           3/16/2020 01:35:49 PM

期望的结果：

Begin                          End

3/16/2020 12:35:47 PM           3/16/2020 12:35:49 PM
3/16/2020 01:35:47 PM           3/16/2020 01:35:49 PM

输入：

 structure(list(Action = structure(c(1L, 2L, 1L, 2L), .Label =       c("Begin", 
 "End"), class = "factor"), Date = structure(c(3L, 4L, 1L, 2L),      .Label = c("3/16/2020 1:35:47 PM", 
 "3/16/2020 1:35:49 PM", "3/16/2020 12:35:47 PM", "3/16/2020     12:35:49 PM"
  ), class = "factor")), class = "data.frame", row.names = c(NA, 
 -4L))

我尝试过的：

I think this is a permute type problem, but not exactly sure.
Permute(df, invertdf=FALSE)

任何建议都有帮助。

Answer 1

我们可以在创建序列列后转为“宽”格式

library(dplyr)
library(data.table)
library(tidyr)
df %>% 
    mutate(rn = rowid(Action)) %>% 
    pivot_wider(names_from = Action, values_from = Date) %>%
    select(-rn)
# A tibble: 2 x 2
#  Begin                 End                  
#  <fct>                 <fct>                
#1 3/16/2020 12:35:47 PM 3/16/2020 12:35:49 PM
#2 3/16/2020 1:35:47 PM  3/16/2020 1:35:49 PM 

---

或data.table

library(data.table)
dcast(setDT(df), rowid(Action) ~ Action, value.var = 'Date')[, Action := NULL][]
#              Begin                   End
#1: 3/16/2020 12:35:47 PM 3/16/2020 12:35:49 PM
#2:  3/16/2020 1:35:47 PM  3/16/2020 1:35:49 PM

---

或者不使用带有unstack 的外部包（假设'Begin'、'End' 的长度相同）

unstack(df, Date ~ Action)
#              Begin                   End
#1 3/16/2020 12:35:47 PM 3/16/2020 12:35:49 PM
#2  3/16/2020 1:35:47 PM  3/16/2020 1:35:49 PM