tidyr uncount() in R: Expand Frequency Counts Into Rows

The uncount() function in tidyr expands a data frame where one column holds frequency counts into one row per count unit. It is the opposite of dplyr::count().

By Selva Prabhakaran · Published May 11, 2026 · Last updated May 11, 2026

⚡ Quick Answer

df |> uncount(n)                    # repeat each row n times
df |> uncount(n, .remove = FALSE)   # keep the n column
df |> uncount(n, .id = "row_id")    # add row index
dplyr::count(df, ...)                # opposite operation

Need explanation? Read on for examples and pitfalls.

📊 Is uncount() the right tool?

What uncount() does in one sentence

uncount(data, weights, .remove = TRUE, .id = NULL) repeats each row of data weights times, expanding a frequency-count table into individual rows. It reverses dplyr's count().

Syntax

uncount(data, weights, .remove = TRUE, .id = NULL). weights is the count column.

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.

RExpand a frequency table

library(tidyr) library(dplyr) freq <- tibble( product = c("X","Y","Z"), n = c(3, 2, 1) ) freq |> uncount(n) #> product #> 1 X #> 2 X #> 3 X #> 4 Y #> 5 Y #> 6 Z

Tip

Use uncount when you have count-form data and need one-row-per-observation form. Common after working with pre-aggregated data.

Five common patterns

1. Standard expansion

RCounts to rows

freq |> uncount(n)

2. Keep original count

R.remove = FALSE

freq |> uncount(n, .remove = FALSE) #> Each row repeated n times AND original n column kept

3. Add row index

R.id assigns 1..n within each group

freq |> uncount(n, .id = "instance") #> Adds 'instance' column: 1, 2, 3 for first product; 1, 2 for second; etc.

4. After dplyr count

RRound-trip

df <- tibble(g = c("a","a","a","b","b")) df |> count(g) |> uncount(n) |> identical(df)

5. Use to simulate / replicate

RMultiply rows by a weight

df <- tibble(item = c("X","Y"), wt = c(2, 3)) df |> uncount(wt) #> X repeated 2 times, Y repeated 3 times

Key Insight

uncount is dplyr count's INVERSE. count rolls rows into a frequency table; uncount expands a frequency table back into rows. Useful when downstream code needs row-per-event format.

uncount() vs base rep() vs count()

Function	Direction	Input
`uncount(data, n)`	Expand	Frequency table
`dplyr::count(data, g)`	Collapse	Row-per-event
`base::rep(x, times)`	Expand	Vector

When to use which:

uncount for data frame frequency expansion.
count for row aggregation.
rep for vector / non-data frame.

A practical workflow

Use uncount when bridging summary data to row-level analysis.

RInteractive R

survey_summary |> uncount(respondent_count) |> mutate(id = row_number())

Each summary row expands to N rows; row_number adds a unique id.

Common pitfalls

Pitfall 1: confusing direction. uncount EXPANDS; count COLLAPSES. They are opposites.

Pitfall 2: zero or negative counts. uncount with n = 0 produces no rows for that input. Negative weights error.

Warning

uncount can produce HUGE output. A row with n = 1,000,000 expands to a million rows. Check sizes first.

Try it yourself

Try it: Expand a frequency table of cars by gear count. Save to ex_expanded.

RYour turn: expand gear frequencies

freq <- mtcars |> count(gear, name = "n") ex_expanded <- freq |> # your code here nrow(ex_expanded) #> Expected: 32 (matches original mtcars)

Click to reveal solution

RSolution

ex_expanded <- freq |> uncount(n) nrow(ex_expanded) #> [1] 32

Explanation: count produced (gear, n) rows; uncount(n) expands to 32 rows total.

After mastering uncount, look at:

dplyr::count(): opposite (collapse to frequency)
dplyr::add_count(): count without collapsing
base::rep(): vector replication
tidyr::complete(): fill missing combinations

FAQ

What does uncount do in tidyr?

uncount(data, n) repeats each row n times, expanding a frequency-count table into individual rows. Opposite of dplyr's count.

What is the difference between uncount and count?

count COLLAPSES rows into a frequency table. uncount EXPANDS a frequency table back into rows. Opposite operations.

How do I keep the original count column?

Pass .remove = FALSE. Each row is still repeated; the count column survives.

How do I add an id within each expansion?

Pass .id = "instance". Adds a column with 1, 2, 3, ... per repetition.

What if a count is 0?

That row contributes 0 rows to the output. Negative weights error.

Navigate

Tidyverse packages

Deep dives

Wrangling & EDA

Statistics

Machine Learning

Time Series

By Industry

Reporting & Apps

Levels

tidyr uncount() in R: Expand Frequency Counts Into Rows

What uncount() does in one sentence

Syntax

Five common patterns

1. Standard expansion

2. Keep original count

3. Add row index

4. After dplyr count

5. Use to simulate / replicate

uncount() vs base rep() vs count()

A practical workflow

Common pitfalls

Try it yourself

FAQ

Navigate

Tidyverse packages

Deep dives

Wrangling & EDA

Statistics

Machine Learning

Time Series

By Industry

Reporting & Apps

Levels

tidyr uncount() in R: Expand Frequency Counts Into Rows

What uncount() does in one sentence

Syntax

Five common patterns

1. Standard expansion

2. Keep original count

3. Add row index

4. After dplyr count

5. Use to simulate / replicate

uncount() vs base rep() vs count()

A practical workflow

Common pitfalls

Try it yourself

Related tidyr / dplyr functions

FAQ