Split columns by delimitator and place in new columns in the same data frame R-CodePudding

I'm trying to split columns in a dataframe and place them next to each other based on a single factor column. I would like to turn this:

   TOD  Value
1  Day   135
2  Day   513
3  Day   567
4  Day   848
5  Day   578
6  Night 145
7  Night 267
8  Night 589
9  Night 258
10 Night 278

and turn it into this:

   TOD  Value TOD_2 Value_2
1  Day   135  Night  145
2  Day   513  Night  267
3  Day   567  Night  589
4  Day   848  Night  258
5  Day   578  Night  278

How do I do this?

CodePudding user response：

Assuming your data frame is called df, you could do:

do.call(cbind, split(df, df$TOD)) |> 
 setNames(c(names(df), paste0(names(df), "_2")))

#>   TOD Value TOD_2 Value_2
#> 1 Day   135 Night     145
#> 2 Day   513 Night     267
#> 3 Day   567 Night     589
#> 4 Day   848 Night     258
#> 5 Day   578 Night     278

Data from question in reproducible format

df <- structure(list(TOD = c("Day", "Day", "Day", "Day", "Day", "Night", 
"Night", "Night", "Night", "Night"), Value = c(135L, 513L, 567L, 
848L, 578L, 145L, 267L, 589L, 258L, 278L)), 
class = "data.frame", row.names = c("1", 
"2", "3", "4", "5", "6", "7", "8", "9", "10"))

CodePudding user response：

A tidy solution using group_split and bind_cols:

library(dplyr)

df %>%
  group_split(TOD) %>%
  Map(function(x, y) {names(x) <- paste(names(x), y, sep = "_"); x}, ., 1:length(.)) %>%
  bind_cols()

# # A tibble: 5 × 4
#   TOD_1 Value_1 TOD_2 Value_2
#   <chr>   <int> <chr>   <int>
# 1 Day       135 Night     145
# 2 Day       513 Night     267
# 3 Day       567 Night     589
# 4 Day       848 Night     258
# 5 Day       578 Night     278