1

A while ago I asked for help on how to do the reverse of what I want to do now, this discussion can be found here. Anyway, I now need to join my data back into the format it once was. That is, to join the separate rows (each containing a word), into one sentence per ID. For example:

Input:

id word
1 Lorem
1 ipsum
1 dolor
1 sit
1 amet
2 consectetur
2 adipiscing
2 elit
3 Donec
...

Output:

id text
1 Lorem ipsum dolor sit amet
2 consectetur adipiscing elit
3 Donec euismod enim quis 
4 nunc fringilla sodales
5 Etiam tempor ligula vitae 
6 pellentesque dictum

At first, I tried to do this with the reshape package and its melt() and cast() functions. I also tried the tidyr package. However, these functions rely on a variable name column specifying the column name for each of the new columns. Not exactly my case (and each sentence can be of different length).

How can I do this task in R?

Community
  • 1
  • 1
Joshua
  • 722
  • 12
  • 27
  • Does this answer your question? [Aggregating by unique identifier and concatenating related values into a string](https://stackoverflow.com/questions/16596515/aggregating-by-unique-identifier-and-concatenating-related-values-into-a-string) – camille Dec 03 '19 at 01:22

1 Answers1

2

We can use data.table. We convert the 'data.frame' to 'data.table' (setDT(df1)), grouped by 'id', we paste the 'word' together.

library(data.table)
setDT(df1)[, list(text= paste(word, collapse=' ')), by = id]
#   id                        text
#1:  1  Lorem ipsum dolor sit amet
#2:  2 consectetur adipiscing elit
#3:  3                       Donec

Or using dplyr, we can similarly group by 'id' and paste the 'word' column.

library(dplyr)
df1 %>%
    group_by(id) %>%
    summarise(text= paste(word, collapse=' '))

Or a base R option is

aggregate(word~id, df1, FUN = paste, collapse=' ')

data

df1 <- structure(list(id = c(1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 3L),
word = c("Lorem", 
"ipsum", "dolor", "sit", "amet", "consectetur", "adipiscing", 
"elit", "Donec")), .Names = c("id", "word"), class = "data.frame", 
row.names = c(NA, -9L))
akrun
  • 874,273
  • 37
  • 540
  • 662