Navigate
Home Compendium
Exercises
Tidyverse packages
dplyr ggplot2 tidyr stringr lubridate purrr forcats readr broom Tidyverse R for Data Science
Deep dives
dplyr Joins dplyr Window Functions dplyr group_by ggplot2 Themes ggplot2 Facets ggplot2 Color Scales ggplot2 Bar Charts ggplot2 Heatmaps tidyr Pivot tidyr Nest/Unnest Regex
Wrangling & EDA
Data Wrangling Data Cleaning EDA Data Visualization data.table dbplyr / SQL Web Scraping API Calls
Statistics
Hypothesis Testing Linear Regression Logistic Regression Correlation Probability Distributions
Machine Learning
Machine Learning tidymodels Random Forest XGBoost Clustering PCA Cross-Validation
Time Series
Time Series ARIMA
By Industry
Finance Marketing Analytics Healthcare Biostatistics Genomics Sports Analytics A/B Testing
Reporting & Apps
Shiny R Markdown
Levels
R for Beginners R Interview Questions
R r‑statistics.co

dplyr cur_data() in R: Deprecated; Use pick(everything())

The cur_data() function in dplyr returned the current group's data as a tibble inside summarise() or mutate(). As of dplyr 1.1, it is DEPRECATED; use pick(everything()) instead.

By Selva Prabhakaran  ·  Published May 12, 2026  ·  Last updated May 12, 2026
⚡ Quick Answer
df |> summarise(out = some_fn(cur_data()))         # OLD (deprecated)
df |> summarise(out = some_fn(pick(everything()))) # NEW
cur_data_all()                                       # also deprecated
cur_group_id()                                       # different: integer ID
cur_group()                                          # different: group values

Need explanation? Read on for examples and pitfalls.

📊 Should I use cur_data?
STARTnew codeNO. Use pick(everything()) instead.legacy codeworks but emits a deprecation warning.migratingreplace `cur_data()` with `pick(everything())` everywhere.need just group IDcur_group_id()need group valuescur_group()

What cur_data() did in one sentence

cur_data() returned the current group's data (excluding grouping columns) as a tibble, inside dplyr verbs on a grouped tibble. Since dplyr 1.1, it is deprecated in favor of pick(everything()) which is more explicit about what's being selected.

Migration

Replace cur_data() with pick(everything()). The new form is functionally equivalent but reads more clearly.

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.
ROld vs new
library(dplyr) # OLD (deprecated): mtcars |> group_by(cyl) |> summarise(out = nrow(cur_data())) # NEW: mtcars |> group_by(cyl) |> summarise(out = nrow(pick(everything())))

  

  





Tip
cur_data() and cur_data_all() are BOTH deprecated. The first excluded grouping columns; the second included them. Both replaced by pick(...) with explicit column selection.



Five common patterns (legacy)



1. Pass current data to a function




  

    
RCompute something on group data

    
df |>
  group_by(g) |>
  summarise(out = some_fn(cur_data()))

    

      
      
    

    

  

  





Modern: summarise(out = some_fn(pick(everything()))).



2. Row count




  

    
Rcur_data() |> nrow()

    
df |>
  group_by(g) |>
  summarise(n = nrow(cur_data()))

    

      
      
    

    

  

  





Modern: just use n() (or nrow(pick(everything()))).



3. Pass to model fit




  

    
RPer-group model

    
df |>
  group_by(g) |>
  summarise(model = list(lm(y ~ x, data = cur_data())))

    

      
      
    

    

  

  





Modern: summarise(model = list(lm(y ~ x, data = pick(everything())))).



4. Filter inside summarise (rare)




  

    
Rcur_data was the only way pre-1.1

    
df |>
  group_by(g) |>
  summarise(custom = nrow(cur_data() |> filter(value > 10)))

    

      
      
    

    

  

  





Modern: same with pick(everything()).



5. cur_data_all (with grouping cols)




  

    
Rcur_data_all included grouping columns

    
df |>
  group_by(g) |>
  summarise(rows = nrow(cur_data_all()))

    

      
      
    

    

  

  





Modern: still pick(everything()) (dplyr 1.1 includes grouping by default in pick).



Key Insight
The deprecation reflects dplyr's move to more EXPLICIT column selection. cur_data() was opaque about which columns were included; pick(everything()) makes it explicit. Code reviewers can see exactly which columns are passed.



cur_data() vs pick() vs cur_group()



Three approaches in the deprecation transition.







Function
Status
Best for






cur_data()
Deprecated
Migrate to pick




cur_data_all()
Deprecated
Migrate to pick




pick(everything())
Recommended
Modern equivalent




pick(specific, cols)
Recommended
Explicit column subset




cur_group()
Active
Just the group values







When to use which:



    


  • For new code: pick(everything()) or pick(specific, cols).
    • 


  • Reading old code: cur_data and cur_data_all both mean "current group's data".
    • 


  • Just need group ID: cur_group_id().
    • 





A practical migration



Search legacy code for cur_data and replace.




  

    
RInteractive R

    
# Before:
df |> group_by(g) |> summarise(x = some_fn(cur_data()))

# After:
df |> group_by(g) |> summarise(x = some_fn(pick(everything())))

    

      
      
    

    

  

  





For the rare cur_data_all case:




  

    
RInteractive R

    
# Before:
df |> group_by(g) |> summarise(x = some_fn(cur_data_all()))

# After (same):
df |> group_by(g) |> summarise(x = some_fn(pick(everything())))

    

      
      
    

    

  

  





dplyr 1.1's pick(everything()) includes grouping columns by default, matching cur_data_all.



Common pitfalls



Pitfall 1: ignoring the deprecation warning. Code still works but the warning means a future dplyr major version may remove cur_data entirely.



Pitfall 2: confusing cur_data with cur_group_id. They serve different purposes. cur_data was the FULL data; cur_group_id is just an integer.



Warning
cur_data and cur_data_all are deprecated as of dplyr 1.1 (Jan 2023). Code using them emits warnings and may break in a future major version. Migrate now.



Try it yourself





Try it: Convert a cur_data call to its pick equivalent. Save to ex_pick.



  

    
RYour turn: migrate cur_data to pick

    
# Old (avoid):
old_result <- mtcars |>
  group_by(cyl) |>
  summarise(rows = nrow(cur_data()))

# New (write this version):
ex_pick <- mtcars |>
  # your code here

ex_pick
#> Expected: same result, no deprecation warning

    

      
      
    

    

  

  





Click to reveal solution



  

    
RSolution

    
ex_pick <- mtcars |>
  group_by(cyl) |>
  summarise(rows = nrow(pick(everything())))

# Or even simpler:
ex_simple <- mtcars |>
  group_by(cyl) |>
  summarise(rows = n())

    

      
      
    

    

  

  





Explanation: pick(everything()) is the direct cur_data replacement. For row counts, n() is even simpler.










After understanding cur_data's deprecation, look at:



    


  • pick(): modern column-selection inside verbs
    • 


  • cur_group_id() / cur_group() / cur_group_rows(): group introspection (still active)
    • 


  • n(): row count
    • 


  • across(): per-column transformation
    • 


  • c_across(): row-wise across
    • 





For most cur_data uses, n() (for counts) or pick() (for full data) covers the use case.



Why cur_data was deprecated



The dplyr team's reasoning: cur_data was implicit, pick is explicit. With cur_data you couldn't tell from the call site which columns were being operated on; with pick(everything()) the intent is visible. Combined with the new tidyselect support in pick (pick(starts_with("x_")) etc.), the new function is strictly more flexible. The deprecation is part of dplyr's broader push toward making pipelines self-documenting.



FAQ



Is cur_data deprecated in dplyr?



Yes, as of dplyr 1.1.0 (Jan 2023). Use pick(everything()) instead.



What is the modern replacement for cur_data?



pick(everything()). It returns the same result but is more explicit about column selection.



What is the difference between cur_data and cur_data_all?



cur_data EXCLUDED grouping columns; cur_data_all INCLUDED them. Both deprecated. pick(everything()) includes grouping columns by default in dplyr 1.1+.



Will cur_data be removed in a future dplyr version?



Possibly. It is currently soft-deprecated (warning). Future major versions may remove it entirely. Migrate proactively.



Can I use cur_data_id?



There is no cur_data_id. You may be thinking of cur_group_id, which returns an integer. cur_data and cur_data_all both returned a tibble.



          

        


        
      


      

        

        
© 2016-2026 Selva Prabhakaran.
          This work is licensed under the Creative Commons License.
        

        

          About ·
          RSS