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

ggplot2 geom_jitter() in R: Scatter With Random Jitter

The geom_jitter() function in ggplot2 adds random horizontal and vertical noise to scatter points to avoid overplotting on discrete or rounded data.

By Selva Prabhakaran  ·  Published May 11, 2026  ·  Last updated May 11, 2026
⚡ Quick Answer
ggplot(df, aes(x, y)) + geom_jitter()
ggplot(df, aes(x, y)) + geom_jitter(width = 0.2, height = 0)
ggplot(df, aes(x, y)) + geom_jitter(alpha = 0.5)
geom_point(position = position_jitter(width = 0.1))  # equivalent
geom_point()                                          # no jitter

Need explanation? Read on for examples and pitfalls.

📊 Is geom_jitter() the right tool?
STARTdiscrete x with overlapping pointsgeom_jitter()categorical-numeric scattergeom_jitter()continuous data, dense overlapgeom_point(alpha = 0.3) or geom_hexbox + raw pointsgeom_boxplot() + geom_jitterexact positions mattergeom_point()

What geom_jitter() does in one sentence

geom_jitter() is geom_point() with random noise added to x and y to spread overlapping points apart. Especially useful for categorical x data.

Syntax

geom_jitter(width = NULL, height = NULL, alpha = 1, ...). width and height control noise amount.

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.
RJitter cyl vs mpg
library(ggplot2) ggplot(mtcars, aes(factor(cyl), mpg)) + geom_jitter(width = 0.2)

  

  





Tip
Set width = 0.2 or smaller for categorical x. Default jitter width is 40% of resolution, often too wide.



Five common patterns



1. Categorical-numeric scatter




  

    
RCyl as factor + mpg

    
ggplot(mtcars, aes(factor(cyl), mpg)) +
  geom_jitter(width = 0.2)

    

      
      
    

    

  

  





2. Combined with boxplot




  

    
RShow distribution + raw

    
ggplot(mtcars, aes(factor(cyl), mpg)) +
  geom_boxplot(outlier.shape = NA) +
  geom_jitter(width = 0.2, alpha = 0.6)

    

      
      
    

    

  

  





3. Vertical-only jitter




  

    
RSpread y but keep x exact

    
ggplot(df, aes(x, y)) +
  geom_jitter(width = 0, height = 0.3)

    

      
      
    

    

  

  





4. With color and alpha




  

    
RMany groups; transparent

    
ggplot(mtcars, aes(factor(cyl), mpg, color = factor(gear))) +
  geom_jitter(width = 0.2, alpha = 0.7)

    

      
      
    

    

  

  





5. Reproducible jitter




  

    
Rset.seed for stable layout

    
set.seed(42)
ggplot(mtcars, aes(factor(cyl), mpg)) +
  geom_jitter(width = 0.2)

    

      
      
    

    

  

  





Key Insight
Jitter adds RANDOM noise; without set.seed, the plot looks slightly different each time. For reproducible reports, set the seed before plotting.



geom_jitter() vs geom_point() vs position_jitter()







Approach
Best for






geom_jitter()
Direct jitter of points




geom_point(position = position_jitter(...))
More control over jitter position




geom_point()
No jitter (exact positions)







A practical workflow



Pair geom_jitter with geom_boxplot to show distribution AND raw points.




  

    
RInteractive R

    
ggplot(mtcars, aes(factor(cyl), mpg)) +
  geom_boxplot(outlier.shape = NA, alpha = 0.5) +
  geom_jitter(width = 0.15, alpha = 0.7) +
  labs(x = "Cylinders", y = "MPG")

    

      
      
    

    

  

  





Box shows summary; jittered points show raw data.



Common pitfalls



Pitfall 1: too much jitter. Default width spreads points beyond their group. Set width = 0.1 to 0.2 for cleaner display.



Pitfall 2: irreproducible plots. Without set.seed, jitter is random per run. Use set.seed() before for reports.



Warning
geom_jitter() MOVES the actual point positions. Don't use it for exact-coordinate plots like maps or precise scatters.



Try it yourself





Try it: Plot mtcars cyl vs mpg with jittered points and a transparent boxplot. Save to ex_plot.



  

    
RYour turn: box + jitter

    
ex_plot <- mtcars |>
  # your code here

    

      
      
    

    

  

  





Click to reveal solution



  

    
RSolution

    
ex_plot <- ggplot(mtcars, aes(factor(cyl), mpg)) +
  geom_boxplot(alpha = 0.4, outlier.shape = NA) +
  geom_jitter(width = 0.15, alpha = 0.7)

    

      
      
    

    

  

  





Explanation: Boxplot for summary, jitter for raw data; outlier.shape = NA prevents double-marking.










After mastering geom_jitter, look at:



    


  • geom_point(): exact positions
    • 


  • geom_boxplot(): distribution summary
    • 


  • geom_violin(): distribution shape
    • 


  • geom_dotplot(): stacked dots
    • 


  • position_jitter() / position_jitterdodge(): positioning helpers
    • 


  • geom_hex() / geom_bin2d(): heatmap for many points
    • 





FAQ



What does geom_jitter do in ggplot2?



geom_jitter() adds random horizontal and vertical noise to scatter points to spread overlapping points apart.



How do I control how much jitter?



Set width and height: geom_jitter(width = 0.2, height = 0). Default is 40% of resolution.



Should I use geom_jitter or geom_point?



geom_jitter for discrete/categorical x with overlapping points. geom_point for continuous x or when exact positions matter.



How do I make jitter reproducible?



Call set.seed(N) before the plot. Without it, jitter is random per render.



Can I combine geom_jitter with geom_boxplot?



Yes. geom_boxplot(outlier.shape = NA) + geom_jitter() shows summary AND raw data without double-marking outliers.



          

        


        
      


      

        

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

        

          About ·
          RSS