R 条件变异日期列使用 groupby 和填充

R conditional mutate date column using groupby and fill

那里有类似措辞的问题,但 none 显示了我想要做的事情。我在下面有一个数据框示例。我想 group_by ID 并创建一个 Date2 列,其中 rank=2。我很难弄清楚这一点。

     ID   Rank    Date       Date2
1   5678   1    2000-01-01   2010-05-02    
2   5678   2    2010-05-02   2010-05-02  
3   1234   1    2000-01-01   2015-06-03     
4   1234   2    2015-06-03   2015-06-03   

这是我目前的情况:

df <- df %>% group_by(ID) %>%fill(Date2,.direction='up')

我该怎么做?

试试这个:

library(dplyr)
#Code
df %>% group_by(ID) %>% mutate(Date2=Date[Rank==2])

输出:

# A tibble: 4 x 4
# Groups:   ID [2]
     ID  Rank Date       Date2     
  <int> <int> <chr>      <chr>     
1  5678     1 2000-01-01 2010-05-02
2  5678     2 2010-05-02 2010-05-02
3  1234     1 2000-01-01 2015-06-03
4  1234     2 2015-06-03 2015-06-03

使用了一些数据:

#Data
df <- structure(list(ID = c(5678L, 5678L, 1234L, 1234L), Rank = c(1L, 
2L, 1L, 2L), Date = c("2000-01-01", "2010-05-02", "2000-01-01", 
"2015-06-03")), row.names = c("1", "2", "3", "4"), class = "data.frame")

此外,如果您想使用 fill(),您可以尝试此代码。您必须使用 ifelse() 之类的条件来分配日期,然后填写值:

#Code 2
df %>% group_by(ID) %>%
  mutate(Date2=ifelse(Rank==2,Date,NA)) %>%
  fill(Date2,.direction = 'up')

输出:

# A tibble: 4 x 4
# Groups:   ID [2]
     ID  Rank Date       Date2     
  <int> <int> <chr>      <chr>     
1  5678     1 2000-01-01 2010-05-02
2  5678     2 2010-05-02 2010-05-02
3  1234     1 2000-01-01 2015-06-03
4  1234     2 2015-06-03 2015-06-03

我们可以按'ID'分组,用逻辑表达式得到对应的'Date1'

library(dplyr)
df %>%
    group_by(ID) %>%
    mutate(Date2 = Date[Rank == 2][1])
# A tibble: 4 x 4
# Groups:   ID [2]
#     ID  Rank Date       Date2     
#  <int> <int> <chr>      <chr>     
#1  5678     1 2000-01-01 2010-05-02
#2  5678     2 2010-05-02 2010-05-02
#3  1234     1 2000-01-01 2015-06-03
#4  1234     2 2015-06-03 2015-06-03

或者另一种选择是使用 match

df %>%
  group_by(ID) %>%
  mutate(Date2 = Date[match(2, Rank)])

或使用data.table

library(data.table)
setDT(df)[, Date2 := Date[match(2, Rank)], ID]

base R

df$Date2 <- with(df, rep(Date[Rank == 2], table(ID)))

数据

df <- structure(list(ID = c(5678L, 5678L, 1234L, 1234L), Rank = c(1L, 
2L, 1L, 2L), Date = c("2000-01-01", "2010-05-02", "2000-01-01", 
"2015-06-03")), row.names = c("1", "2", "3", "4"), class = "data.frame")