按 ID 和日期发生的某些事件

Question

我有一个如下所示的数据框：

<ID>     <Event>      <Date>
1         Ate          2021-01-01
1         Drank        2021-01-01
1         Ate          2021-02-23
2         Ate          2021-01-02     
2         Ran          2021-01-02
2         Ate          2021-02-23
3         Drank        2021-01-01
3         Ran          2021-02-23

我希望确定的是在某个日期是否为每组ID发生了事件。在这种情况下，我想确定每组日期中哪个 ID“吃了”，时间段。

预期的结果是 table，看起来像：

<ID>     <Event>      <Date>            <Outcome>
1         Ate          2021-01-01       Yes
1         Drank        2021-01-01       Yes
1         Jumped       2021-02-23       No
2         Ate          2021-01-02       Yes
2         Ran          2021-01-02       Yes
2         Ate          2021-02-23       No
3         Drank        2021-01-01       No
3         Ran          2021-02-23       No

我希望这是有道理的，谢谢你的帮助！

Answer 1

如果我们要检查 'Ate' 是否与每个 'ID'、'Date' 的另一个 'Event' 一起出现，请按 'ID'、'Date'，检查行数（n()）是否大于1并且找到（&）'Ate' %in% 'Event'

library(dplyr)
df1 %>% 
  group_by(ID, Date) %>% 
  mutate(Outcome = c("No", "Yes")[(n() > 1 & 'Ate' %in% Event) + 1]) %>%
  ungroup

-输出

# A tibble: 8 x 4
#     ID Event Date       Outcome
#  <int> <chr> <chr>      <chr>  
#1     1 Ate   2021-01-01 Yes    
#2     1 Drank 2021-01-01 Yes    
#3     1 Ate   2021-02-23 No     
#4     2 Ate   2021-01-02 Yes    
#5     2 Ran   2021-01-02 Yes    
#6     2 Ate   2021-02-23 No     
#7     3 Drank 2021-01-01 No     
#8     3 Ran   2021-02-23 No

为了检查每组有 duplicate 'Ate' 而没有其他值的情况，我们可以使用 n_distinct（而不是 n()）即检查'Event' 的不同元素的数量大于 1

df1 %>% 
  group_by(ID, Date) %>% 
  mutate(Outcome = c("No", "Yes")[n_distinct(Event) > 1 &
           'Ate' %in% Event) + 1]) %>%
  ungroup

数据

df1 <- structure(list(ID = c(1L, 1L, 1L, 2L, 2L, 2L, 3L, 3L), Event = c("Ate", 
"Drank", "Ate", "Ate", "Ran", "Ate", "Drank", "Ran"), Date = c("2021-01-01", 
"2021-01-01", "2021-02-23", "2021-01-02", "2021-01-02", "2021-02-23", 
"2021-01-01", "2021-02-23")), class = "data.frame", row.names = c(NA, 
-8L))

Answer 2

这里有一个data.table选项

setDT(df)[,Outcome := c("No","Yes")[1+isTRUE(Date==Date[Event=="Ate"] & .N>1)],.(ID,Date)]

给予

   ID Event       Date Outcome
1:  1   Ate 2021-01-01     Yes
2:  1 Drank 2021-01-01     Yes
3:  1   Ate 2021-02-23      No
4:  2   Ate 2021-01-02     Yes
5:  2   Ran 2021-01-02     Yes
6:  2   Ate 2021-02-23      No
7:  3 Drank 2021-01-01      No
8:  3   Ran 2021-02-23      No

按 ID 和日期发生的某些事件

Occurrence of some event by ID and Date

r

subset

数据