Navigate
Courses Tutorials Practice Pricing Tools Get Certified

tidyr separate_longer_delim() in R: Split Column Into Rows

The separate_longer_delim() function in tidyr 1.3 splits a column's values by a delimiter and creates ONE NEW ROW per part. It is the tidy-data unstacking idiom: comma-separated lists become individual rows.

By Selva Prabhakaran  ·  May 2026
⚡ Quick Answer
df |> separate_longer_delim(tags, delim = ",")    # one row per tag
df |> separate_longer_delim(items, delim = "; ")
df |> separate_longer_delim(c(col1, col2), delim = ",") # multiple cols
tidyr::separate_rows(df, tags, sep = ",")          # superseded predecessor
df |> separate_wider_delim(...)                    # different: into columns

Need explanation? Read on for examples and pitfalls.

📊 Is separate_longer_delim() the right tool?
STARTcomma-list -> rowsseparate_longer_delim()comma-list -> columnsseparate_wider_delim()old code with separate_rowsworks but supersededpositional split into rowsseparate_longer_position()list-column to rowstidyr::unnest_longer()

What separate_longer_delim() does in one sentence

separate_longer_delim(data, cols, delim) splits each value of cols by delim and creates ONE NEW ROW per resulting part. Other columns' values are duplicated for each new row.

This unstacks delimiter-separated lists into the standard tidy "one observation per row" format.

Syntax

separate_longer_delim(data, cols, delim). delim is a literal string (not regex).

Run live
Run live, no install needed. Every R block on this page runs in your browser. Click Run, edit the code, re-run instantly. No setup.
RComma-separated tags to rows
library(tidyr) library(dplyr) df <- tibble(id = 1:2, tags = c("a,b,c","d,e")) df |> separate_longer_delim(tags, delim = ",") #> id tags #> 1 1 a #> 2 1 b #> 3 1 c #> 4 2 d #> 5 2 e

  

  





Tip
This is a key tidy-data operation: convert "list inside a cell" to "one cell per item, one row per cell". Standard input prep for many downstream operations.



Five common patterns



1. Comma-separated tags




  

    
REach tag becomes a row

    
df <- tibble(post_id = 1:2, tags = c("R,dplyr,tidyr","ggplot2,R"))
df |>
  separate_longer_delim(tags, delim = ",")

    

      
      
    

    

  

  





2. Multi-character delimiter




  

    
R'; ' separator

    
df |>
  separate_longer_delim(items, delim = "; ")

    

      
      
    

    

  

  





3. Multiple columns at once




  

    
RBoth authors and keywords

    
df |>
  separate_longer_delim(c(authors, keywords), delim = ",")

    

      
      
    

    

  

  





4. Trim whitespace post-split




  

    
RCombine with stringr

    
library(stringr)
df |>
  separate_longer_delim(tags, delim = ",") |>
  mutate(tags = str_trim(tags))

    

      
      
    

    

  

  





5. Empty / NA handling




  

    
RNAs and empties

    
df <- tibble(id = 1:3, tags = c("a,b","", NA))
df |>
  separate_longer_delim(tags, delim = ",")
#> id 1 produces 2 rows (a, b)
#> id 2 produces 1 row (empty string)
#> id 3 produces 1 row (NA)

    

      
      
    

    

  

  





Key Insight
**separate_longer_delim() is the modern replacement for separate_rows().* Same semantics; cleaner name and consistency with separate_wider_. The _longer_ and _wider_ naming aligns with pivot_longer / pivot_wider conventions.



separate_longer_delim() vs separate_rows() vs unnest()







Function
Status
Input
Best for






separate_longer_delim()
Recommended
String column
Modern delimiter-to-rows




tidyr::separate_rows()
Superseded
String column
Same; older API




tidyr::unnest()
Recommended
List column
Already-list-shaped data




unnest_longer()
Recommended
List column
Single column to rows







When to use which:



    


  • separate_longer_delim for delimiter-separated strings.
    • 


  • unnest for list columns (e.g., from JSON).
    • 


  • Avoid separate_rows in new code.
    • 





A practical workflow



Use separate_longer_delim for "tags" or "tags-like" data normalization.




  

    
RInteractive R

    
posts |>
  separate_longer_delim(tags, delim = ",") |>
  mutate(tags = trimws(tags)) |>
  count(tags, sort = TRUE)

    

      
      
    

    

  

  





Tag frequency analysis from comma-separated tag strings. Common pre-processing for text data.



Common pitfalls



Pitfall 1: forgetting to trim whitespace. "a, b, c" splits to "a", " b", " c". Use mutate(col = trimws(col)) after.



Pitfall 2: confusing wider vs longer. wider splits into COLUMNS. longer splits into ROWS. Pick by what you want.



Warning
separate_longer_delim() requires tidyr 1.3+. Earlier versions only have separate_rows(). Check packageVersion("tidyr") if confused.



Try it yourself





Try it: Split a genres column of comma-separated movie genres into one row per genre. Save to ex_long.



  

    
RYour turn: split genres

    
df <- tibble(
  movie = c("Inception","La La Land"),
  genres = c("Sci-Fi,Thriller","Drama,Musical,Romance")
)

ex_long <- df |>
  # your code here

nrow(ex_long)
#> Expected: 5 (2 + 3)

    

      
      
    

    

  

  





Click to reveal solution



  

    
RSolution

    
ex_long <- df |>
  separate_longer_delim(genres, delim = ",")

ex_long
#>   movie       genres
#> 1 Inception   Sci-Fi
#> 2 Inception   Thriller
#> 3 La La Land  Drama
#> 4 La La Land  Musical
#> 5 La La Land  Romance

    

      
      
    

    

  

  





Explanation: Each comma-separated genre becomes its own row; movie name is repeated.










After mastering separate_longer_delim, look at:



    


  • separate_longer_position(): split into rows by character position
    • 


  • separate_wider_delim(): split into columns
    • 


  • separate_rows(): superseded predecessor
    • 


  • unnest_longer(): list column to rows
    • 


  • tidyr::nest() / unnest(): list-column workflows
    • 





For combining rows back into delimiter-separated strings, use summarise(col = paste(col, collapse = ",")) after grouping.



FAQ



What does separate_longer_delim do in tidyr?



Splits a column's values by a delimiter and creates one row per resulting part. Other columns are duplicated to match.



What is the difference between separate_longer_delim and separate_rows?



separate_rows is superseded by separate_longer_delim (and separate_longer_position). Same semantics, modern naming.



What is the difference between separate_longer_delim and separate_wider_delim?



longer creates new ROWS (one per part). wider creates new COLUMNS (one per part). Same input; different output shape.



How do I trim whitespace after split?



mutate(col = trimws(col)) after the split. Or strip whitespace with strsplit + trimws before separate.



What happens with NAs?



NA values produce one row with NA. Empty strings produce one row with "". Filter after if you don't want them.



          
Prefer to learn by doing? Every track on the learning roadmap opens with a free section of interactive lessons, where you write the R yourself and the page tells you when it is wrong.