用符号分隔数字列

Question

我想知道，是否有更直接的解决以下问题的方法。我想用符号和与分类列相关的条件分隔数字列中的值。

如果这是我的数据帧:

library(tidyverse)
a <- tribble(
~category , ~val ,
'A' , 1 ,
'A' , -1 ,
'B' , -2 ,
'C' , 1 ,
'Z' , 3 ,
'Z' , -4
)

我想创建四个新列，如果类别= 'Z'，则为正负值，如果类别== 'Z，则为正负值

下面的代码产生了期望的结果:

a %>%
mutate(good = if_else(category != 'Z' & val >= 0 , val , 0) ,
bad = if_else(category != 'Z' & val < 0 , val , 0) ,
reserve = if_else(category == 'Z' & val >= 0 , val , 0) , 
risk = if_else(category == 'Z' & val < 0 , val , 0))

但是，由于我有许多类别和值列，以及许多将它们分开的规则，我最终会得到大量的if-else条件。有人知道更直接的解决方法吗?

Answer 1

思路与上一个答案相似。按条件划分类别，然后再扩大范围。下面我将向您展示如何使用将逻辑语句转换为数字索引来索引类别。

library(tidyverse)
a |>
mutate(grp = c("bad", "good", "risk", "reserve")[
(val>0) + (category == "Z")  + (val<0 & category == "Z") + (val>0 & category == "Z") +1
],
id = row_number()) |>
pivot_wider(names_from = grp, values_from = val, values_fill = 0) |>
select(-id)
#> # A tibble: 6 x 5
#>   category  good   bad reserve  risk
#>   <chr>    <dbl> <dbl>   <dbl> <dbl>
#> 1 A            1     0       0     0
#> 2 A            0    -1       0     0
#> 3 B            0    -2       0     0
#> 4 C            1     0       0     0
#> 5 Z            0     0       3     0
#> 6 Z            0     0       0    -4

Answer 2

如果你可以放弃val列，我会这样做:

library(tidyverse)
a <- tribble(
~category , ~val ,
'A' , 1 ,
'A' , -1 ,
'B' , -2 ,
'C' , 1 ,
'Z' , 3 ,
'Z' , -4
)
a2 <- a %>%
mutate(
condition = case_when(
category != 'Z' & val >= 0 ~ 'good',
category != 'Z' & val < 0 ~ 'bad',
category == 'Z' & val >= 0 ~ 'reserve',
category == 'Z' & val < 0 ~'risk'
)
)
a2 %>%
rownames_to_column(".flag.row") %>%
pivot_wider(
names_from = 'condition',
values_from = 'val',
values_fill = 0
) %>%
select(-.flag.row)