How perform rbind fucntion by prefix using R-CodePudding

suppose i have dataset

dat=structure(list(X0_B02 = c(108L, 280L, 362L, 918L, 141L), X0_B03 = c(398L, 
733L, 725L, 1116L, 421L), X0_B04 = c(214L, 464L, 692L, 1394L, 
238L), X1_B02 = c(480L, 484L, 544L, 911L, 380L), X1_B03 = c(760L, 
856L, 849L, 1110L, 611L), X1_B04 = c(546L, 599L, 737L, 1348L, 
431L)), class = "data.frame", row.names = c(NA, -5L))

0_, 1_ these are prefixes to variables. 0_ means zero day, 1_ means the first day, 2_ will mean the second day of observations for variables bo2, bo3, bo4. I need to make that the variables for days will not in columns but in rows, with indicating the day. Like this.

day B02 B03 B04
0   108 398 214
0   280 733 464
0   362 725 692
0   918 1116    1394
0   141 421 238
1   480 760 546
1   484 856 599
1   544 849 737
1   911 1110    1348
1   380 611 431

is the way using rbind fucntion perform it or to get desired output we need another way? Thank you.

CodePudding user response：

Using tidyverse, you can use pivot_longer with a names_pattern. This will allow the number after "X" to be used for the new day column, and the word after underscore for the other column names.

library(tidyr)

pivot_longer(
  dat,
  cols = everything(),
  names_to = c("day", ".value"),
  names_pattern = "X(\\d )_(\\w )"
)

Output

   day     B02   B03   B04
   <chr> <int> <int> <int>
 1 0       108   398   214
 2 1       480   760   546
 3 0       280   733   464
 4 1       484   856   599
 5 0       362   725   692
 6 1       544   849   737
 7 0       918  1116  1394
 8 1       911  1110  1348
 9 0       141   421   238
10 1       380   611   431