CODEWITHSUNDEEP
rdplyr
niklai
niklai

Reputation: 388

Rename selected columns with a vector of strings

I need to rename several columns which names have a string pattern. Let's use this dataframe as example.

library(tidyverse, tibble)

df = as.tibble(matrix(0, nrow = 3, ncol = 30))

colnames(df) = c("p1", "BNT2", "BNT3", "BNT4","BNT5","BNT6","BNT7","BNT8","BNT9","BNT10",
                 "BNT11","BNT12","BNT13","BNT14" ,"BNT15", "groupTime186", "groupTime187", "groupTime188", "groupTime189", "groupTime190", "groupTime191", 
                 "groupTime192", "groupTime193", "groupTime194", "groupTime195" ,"groupTime196", "groupTime197", 
                 "groupTime198", "groupTime199", "groupTime200")

# A tibble: 3 x 30
     p1  BNT2  BNT3  BNT4  BNT5  BNT6  BNT7  BNT8  BNT9 BNT10 BNT11 BNT12 BNT13 BNT14 BNT15 groupTime186 groupTime187 groupTime188
  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>        <dbl>        <dbl>        <dbl>
1     0     0     0     0     0     0     0     0     0     0     0     0     0     0     0            0            0            0
2     0     0     0     0     0     0     0     0     0     0     0     0     0     0     0            0            0            0
3     0     0     0     0     0     0     0     0     0     0     0     0     0     0     0            0            0            0
# ... with 12 more variables: groupTime189 <dbl>, groupTime190 <dbl>, groupTime191 <dbl>, groupTime192 <dbl>, groupTime193 <dbl>,
#   groupTime194 <dbl>, groupTime195 <dbl>, groupTime196 <dbl>, groupTime197 <dbl>, groupTime198 <dbl>, groupTime199 <dbl>,
#   groupTime200 <dbl>

Normally I would use gsub and set_names to capture the item number and to construct the new name. Like this:

df %>% 
  set_names(gsub("p([0-9]{1,2})|BNT([0-9]{1,2})", "BOS_\\1\\2_cod", names(.)))

With this I can re-use the correlative numbers from the original names. The problem is that, because of the software we use to export responses, time-columns usually have a numeration that does not start from 01, so I can't re-use the numeration. Instead, I have to select only the time-columns and use colnames and paste0 to construct the names, and then rejoin the time-columns. Like this:

colnames(df) = paste0("BOS_", sprintf("%02d", 1:15), "_time")

I don't believe this is a good way to approach this task because requires more steps and it is not embedded in the original piped code that renames the answer-columns.

My question is: How can I select the columns to be renamed and feed them with a vector that contains the new names? Or alternatively, can I use a sequence, like sprintf("%02d", 1:15), so gsub replace the first column with the first term of the sequence? Ideally, I want a solution that can be embedded in a piped code (dplyr).

UPDATE: The expected output is the same dataframe but named in this way:

 [1] "BOS_01_raw"  "BOS_02_raw"  "BOS_03_raw"  "BOS_04_raw"  "BOS_05_raw"  "BOS_06_raw"  "BOS_07_raw"  "BOS_08_raw"  "BOS_09_raw"  "BOS_10_raw" 
[11] "BOS_11_raw"  "BOS_12_raw"  "BOS_13_raw"  "BOS_14_raw"  "BOS_15_raw"  "BOS_01_time" "BOS_02_time" "BOS_03_time" "BOS_04_time" "BOS_05_time"
[21] "BOS_06_time" "BOS_07_time" "BOS_08_time" "BOS_09_time" "BOS_10_time" "BOS_11_time" "BOS_12_time" "BOS_13_time" "BOS_14_time" "BOS_15_time"

As I said before, I can rename the BNT items because they already are numerated, but the groupTime columns are the problem.

Upvotes: 2

Views: 2955

Answers (1)

niklai
niklai

Reputation: 388

I managed to solve the problem thanks to @mt1022 comment. According to How to rename multiple columns given character vectors of column names and replacement in dplyr 0.6.0?:

First a vector with the new names have to be created.

names_boston =  c(paste0("BOS_", sprintf("%02d", 1:31), "_time"))

Then the columns can be selected using grep, and the new names can be feed to rename_at. 

df %>%
rename_at(vars(grep("Time", names(.))), ~names_boston)

And to avoid creating new vectors you can actually feed the vector to the previous line of code:

df %>%
    rename_at(vars(grep("Time", names(.))), ~c(paste0("BOS_", sprintf("%02d", 1:31), "_time")))

Upvotes: 4

Related Questions