如果条件在 R 中匹配，则将一列拆分为两个现有列

Question

我有一个这样的 df:

df <- data.frame(id=c("j1", "j2", "j3/j9", "j5", "j2/j8", "j3/j4"), dad=c("j10", "j11", "", "j13", "", ""), mom=c("k2", "k4", "", "k6", "", ""))

我试图只拆分那些在“id”列中包含斜杠“/”的单元格。我想在现有列“爸爸”和“妈妈”中获取拆分字符串。期望的输出是这样的：

df2 <- data.frame(id=c("j1", "j2", "j3/j9", "j5", "j2/j8", "j3/j4"), dad=c("j10", "j11", "j3", "j13", "j2", "j3"), mom=c("k2", "k4", "j9", "k6", "j8", "j4"))

我正在尝试这段代码：

df3 <- tidyr::separate(data = df, col = "id", into = c("dad", "mom"), sep = "/")

但是将整个列“id”拆分为两个新列。知道如何解决这个问题吗？

Answer 1

这是在 separateing 之后使用 coalesce 的一种方法 - 将空格 ("") 转换为 NA (na_if), separate 将 'id' 放入 'dad2'、'mom2' 列，循环 across 'dad'、'mom' 列和 coalesce 与相应的 'dad2'、'mom2' 列

library(dplyr)
library(tidyr)
library(stringr)
df %>% 
  na_if("") %>% 
  separate(id, into = c("dad2", "mom2"), sep = "/", fill = "right", 
       remove = FALSE) %>% 
  mutate(across(dad:mom, ~ coalesce(.x, get(str_c(cur_column(), 
          2)))), .keep = "unused")

-输出

     id dad mom
1    j1 j10  k2
2    j2 j11  k4
3 j3/j9  j3  j9
4    j5 j13  k6
5 j2/j8  j2  j8
6 j3/j4  j3  j4

或者 across2 来自 dplyover

稍微有用一点

library(dplyover)
df %>%
   na_if("") %>% 
   separate(id, into = c("dad2", "mom2"), sep = "/", fill = "right", 
       remove = FALSE) %>% 
   mutate(across2(dad:mom, dad2:mom2, coalesce, .names = "{xcol}")) %>%
   select(names(df))

Answer 2

您可以使用：

library(dplyr)
library(stringr)

df %>% 
  mutate(dad = if_else(str_detect(id, "/"), str_extract(id, ".*(?=/)"), dad),
         mom = if_else(str_detect(id, "/"), str_extract(id, "(?<=/).*"), mom))

这个returns

     id dad mom
1    j1 j10  k2
2    j2 j11  k4
3 j3/j9  j3  j9
4    j5 j13  k6
5 j2/j8  j2  j8
6 j3/j4  j3  j4

Answer 3

您可以使用 grep 获取包含 / 的行，而不是使用 strsplit 并将结果插入 df。

i <- grep("/", df$id)
df[i, c("dad", "mom")] <- do.call(rbind, strsplit(df$id[i], "/"))
#df[i, -1] <- do.call(rbind, strsplit(df$id[i], "/")) #Alternative
df
##     id dad mom
#1    j1 j10  k2
#2    j2 j11  k4
#3 j3/j9  j3  j9
#4    j5 j13  k6
#5 j2/j8  j2  j8
#6 j3/j4  j3  j4

或使用sub.

i <- grep("/", df$id)
df$dad[i] <- sub("/.*", "", df$id[i])
df$mom[i] <- sub(".*/", "", df$id[i])

如果条件在 R 中匹配，则将一列拆分为两个现有列

split a column in two existing columns if a condition is matched in R

split

r