根据R中的部分匹配重新编码变量

Question

This question probably has a simple answer, so I apologize in advance. I would like to use R to recode the values in v2, df1 and make them look like the data values in v2, df2. Is it possible to do this using a partial match of the values in v2, df1 that contain, say 'Y' and recode those values to 'Yr' as in v2, df2?

> df1
  v1   v2
1  1 Yr01
2  2 Yr02
3  3 Yr03
4  4 Yr04
5  5 Yr05

> df2
  v1 v2
1  1 Yr
2  2 Yr
3  3 Yr
4  4 Yr
5  5 Yr
> 

Answer 1

You can use grepl() to generate a vector of booleans depending on whatever you define as your regex. See here for details on an easy intro to regex: http://www.regular-expressions.info/tutorial.html

df1 <- read.table(text = "
  v1   v2
  1  1 Yr01
  2  2 Yr02
  3  3 Yr03
  4  4 Yr04
  5  5 Yr05", 
  header = TRUE, stringsAsFactors = FALSE)

df1[grepl("Y", df1$v2), "v2"] <- "Yr"
> df1

  v1 v2
1  1 Yr
2  2 Yr
3  3 Yr
4  4 Yr
5  5 Yr

If your data is a factor, you can convert to character first, then use the code above.

Answer 2

我认为这对您有用，但是根据您有多少个组和数据框的大小，可能会有更好的方法：

df1$v2 <- ifelse(grepl("Y", df1$v2), "Yr", df1$v2)

Answer 3

正则表达式的另一种用法。

df1$v2 <- gsub("Y.*","Yr", df1$v2)