如果你假设当逗号出现时，姓氏是第一个单词，并且这个人只有一个姓氏，并且它出现在最后，那么这样做：

library(stringr)
df$last = ifelse(str_detect(df$name, ','), 
                 str_replace(df$name, ',.*',''), 
                 str_extract(df$name, '\\S*$'))

您将获得：

id         name    last
1  1   Jack Smith   Smith
2  2   May, Flora     May
3  3      Jackson Jackson
4  4 Baker, Gavin   Baker
5  5        Walls   Walls

如果有逗号，从逗号开始用空字符串替换。如果没有逗号，则抓取最后一个单词作为姓氏。

使用gsub生成统一格式的名称的基本R选项

> transform(df,last_name = gsub("\\w+\\s", "", gsub("(\\w+), (\\w+)", "\\2 \\1", name)))
  id         name last_name
1  1   Jack Smith     Smith
2  2   May, Flora       May
3  3      Jackson   Jackson
4  4 Baker, Gavin     Baker
5  5        Walls     Walls

你可以尝试使用正则表达式的组合。第一个删除逗号右边的所有字符。第二个删除大写字母后面的空格前面的所有字母字符。

library(dplyr)
 
data.frame(id = c(1,2,3,4,5),
           name = c("Jack Smith", "May, Flora", "Jackson","Baker, Gavin","Walls")) |> 
    mutate(last_name = gsub(pattern = "[,].*$",  replacement = "", x = name) |> 
                       gsub(pattern = "^[A-z]+ (?=[A-Z])", replacement = "", x = _ , perl = TRUE))

  id         name last_name
1  1   Jack Smith     Smith
2  2   May, Flora       May
3  3      Jackson   Jackson
4  4 Baker, Gavin     Baker
5  5        Walls     Walls

基地R路：

df$last_name <- mapply(\(n, comma) 'if'(comma, n[1], tail(n, 1)),
                       strsplit(df$name, ',? '),
                       grepl(',', df$name))
df
#   id         name last_name
# 1  1   Jack Smith     Smith
# 2  2   May, Flora       May
# 3  3      Jackson   Jackson
# 4  4 Baker, Gavin     Baker
# 5  5        Walls     Walls

我们可以使用stringr::str_extract和if_else：

library(dplyr)
library(stringr)

df %>%
    mutate(last_name = if_else(str_detect(name, ","),
                               str_extract_all(name, "^[A-Za-z]+"),
                               str_extract_all(name, "[A-Za-z]+$")))

  id         name last_name
1  1   Jack Smith     Smith
2  2   May, Flora       May
3  3      Jackson   Jackson
4  4 Baker, Gavin     Baker
5  5        Walls     Walls

这里有一个tidyverse解决方案，它做了一些一般性的假设--真实的数据可能比示例数据更混乱：

library(dplyr)
library(stringr)

df <- data.frame(
  id = c(1, 2, 3, 4, 5),
  name = c("Jack Smith", "May, Flora", "Jackson", "Baker, Gavin", "Walls")
)

reformat_name <- function(name) {
  split_name <- stringr::str_split(name, ", ") %>%
    unlist() %>%
    rev() %>%
    paste(collapse = " ")
  
  return(split_name)
  
}

df %>%
  rowwise() %>%
  mutate(name = ifelse(stringr::str_detect(name, ", "), reformat_name(name), name),
         last_name = ifelse(stringr::str_detect(name, " "), trimws(stringr::str_extract(name, "\\s[^ ]+$")), name))
         
#> # A tibble: 5 × 3
#> # Rowwise: 
#>      id name        last_name
#>   <dbl> <chr>       <chr>    
#> 1     1 Jack Smith  Smith    
#> 2     2 Flora May   May      
#> 3     3 Jackson     Jackson  
#> 4     4 Gavin Baker Baker    
#> 5     5 Walls       Walls

试试看

library(dplyr)
library(stringr)
df %>% 
   mutate(last = str_extract(name, "(\\w+)(?=,|$)"))

• 输出

id         name    last
1  1   Jack Smith   Smith
2  2   May, Flora     May
3  3      Jackson Jackson
4  4 Baker, Gavin   Baker
5  5        Walls   Walls