如何使用stringr从字符变量中提取姓氏？技术求助

Hey there! Let's figure out why your code was throwing an error and fix it properly.

Your original code failed because str_match_all(Min3$Name, Name) tried to pair a 25-element vector (Min3$Name) with a 63-element list (Name). R can't handle this length mismatch, hence the error. Here's how to solve it with stringr:

Step 1: Build a Unified Regex Pattern #

First, we'll turn all surnames in the Name list into a single regex pattern, using | to separate each option. This lets the regex engine match any surname from the list in Min3$Name:

library(stringr)

# Convert the Name list into a regex alternative pattern
# Add str_escape() inside if your surnames have special characters like . or ()
name_pattern <- str_c(Name, collapse = "|")

Step 2: Extract Matches to the New Column #

Use str_extract instead of str_match_all—it returns a character vector with the same length as Min3$Name, which fits perfectly as a data frame column:

Min3$Name2 <- str_extract(Min3$Name, name_pattern)

Handling Edge Cases #

• If each entry in Min3$Name has only one matching surname, the code above works perfectly. For entries with multiple matches (e.g., a filename containing multiple surnames), use str_extract_all and combine results:

  # Combine multiple matches into a comma-separated string
  Min3$Name2 <- str_extract_all(Min3$Name, name_pattern) %>% 
    sapply(str_c, collapse = ", ")
  

• If Min3$Name contains pure surnames (no extra filename characters), you don't even need regex—just use %in% to filter:

  Min3$Name2 <- ifelse(Min3$Name %in% Name, Min3$Name, NA)
  

That should get you the new column you need!

内容的提问来源于stack exchange，提问作者NColl