CODEWITHSUNDEEP
r
Banjo
Banjo

Reputation: 1251

Change the name of multiple columns

I have a data set with column names like below: 

colnames(samp) 

[1] "RESPID"             "SAMPLE"             "Weight"             "Q1"                 "Q19A_1"            
 [6] "Q19B_1"             "Q19C_1"             "Q19E_1"             "Q19F_1"             "RECORDERLOOP_Q20_1"
[11] "RECORDERLOOP_Q20_2" "RECORDERLOOP_Q20_3" "RECORDERLOOP_Q20_4" "Q20_1_1"            "Q20_2_1"           
[16] "Q20_3_1"

For the column names that start with "Q19" or "Q20" (i.e. a certain pattern), I want to remove _1 (i.e. _ and the number).

I know how it works for one column (e.g. Q19). It would be something like this: 

library(dplyr)

samp_subset = samp %>%
  select(dplyr::contains("Q19")) 

colnames(samp_subset) = sub('.{02}$', '', colnames(samp_subset))

However, I don't know how to define the expression of certain columns (e.g. for Q19 and Q20 but not for RESPID or Sample etc.).

Upvotes: 0

Views: 32

Answers (2)

akrun
akrun

Reputation: 887118

We can use

library(dplyr)
library(stringr)
df %>%
    rename_at(vars(matches("^Q(19|20)")), ~ str_remove(., "_\\d+$"))

Upvotes: 1

Ronak Shah
Ronak Shah

Reputation: 388982

Using dplyr, you can try rename_at

library(dplyr)
df %>%  rename_at(vars(matches("^Q19|^Q20")), ~sub("_\\d+$", "", .))

Using base R, I think would involve two steps identify the columns and replace the values. 

vals <- grep("^Q19|^Q20", names(df))
names(df)[vals] <- sub("_\\d+$", "", names(df)[vals])

Upvotes: 2

Related Questions