遍历多个字符列以派生一个新变量

Question

你能帮我写一段 R 中的代码吗？我想派生一个新变量，其值为 Yes 或 No，但是它需要检查字符变量 a1、a2、a3、a4 以及任何变量值为 'yellow' 那么可用变量应该是 yes else no.

a1 <- c('orange','red') 
a2 <- c('red','yellow') 
a3 <- c('black','orange') 
a4 <- c('red','brown')

testa <- data.frame(a1,a2,a3,a4)

Answer 1

dplyr

您可以使用 rowwise 和 c_across:

testa %>% 
  rowwise() %>% 
  mutate(available = ifelse(any(c_across(a1:a4) == "yellow"), "yes", "no"))

  a1     a2     a3     a4    available
  <chr>  <chr>  <chr>  <chr> <chr>    
1 orange red    black  red   no       
2 red    yellow orange brown yes

基础 R

使用apply:

testa$available <- apply(testa, 1, function(x) ifelse(any(x == "yellow"), "yes", "no"))

Answer 2

a1 <- c('orange','red') 
a2 <- c('red','yellow') 
a3 <- c('black','orange') 
a4 <- c('red','brown')

df <- data.frame(a1,a2,a3,a4)

library(tidyverse)
df %>% 
  mutate(available = ifelse(rowSums(. == "yellow") > 0, "yes", "no"))
#>       a1     a2     a3    a4 available
#> 1 orange    red  black   red        no
#> 2    red yellow orange brown       yes

library(data.table)

setDT(df)[, available := ifelse(rowSums(.SD == "yellow") > 0, "yes", "no")][]
#>        a1     a2     a3    a4 available
#> 1: orange    red  black   red        no
#> 2:    red yellow orange brown       yes

^{由 reprex package (v2.0.1)}

创建于 2022-01-19

Answer 3

您可以只连接每行中的单元格并在结果字符串中查找字符 "yellow"：

> grepl("yellow",paste(as.data.frame(t(testa))))
[1] FALSE  TRUE

然后您可以使用此生成的逻辑向量将标签放入新列中：

testa$available = c("no","yes")[grepl("yellow",paste(as.data.frame(t(testa))))+1]

这将导致 data.frame:

> testa
      a1     a2     a3    a4 available
1 orange    red  black   red        no
2    red yellow orange brown       yes

(如果你想要"yes"不包含黄色的线，只需在向量c("no","yes")中翻转它们)

遍历多个字符列以派生一个新变量

Loop over multiple character columns to derive a new variable

r

conditional-statements

dataframe

dplyr

基础 R