March 18, 2026

Today

  • Review aggregate()
  • plyr::summarize()
  • ddply()
  • The d_ply() family

Remember aggregate()?

data(warpbreaks)
aggregate(breaks~wool, warpbreaks, FUN=mean)
##   wool   breaks
## 1    A 31.03704
## 2    B 25.25926

Remember aggregate()?

aggregate(breaks~wool+tension, warpbreaks, FUN=mean)
##   wool tension   breaks
## 1    A       L 44.55556
## 2    B       L 28.22222
## 3    A       M 24.00000
## 4    B       M 28.77778
## 5    A       H 24.55556
## 6    B       H 18.77778

aggregate() limitation

Only one aggregation function allowed.

# Does not work
aggregate(breaks~wool+tension, warpbreaks, FUN=c(mean, sd))
## Error in `get()`:
## ! object 'FUN' of mode 'function' was not found
  • What if we want the mean and the sd and the median in one call?

plyr::summarize()

  • Creates new dataframe
library(plyr)
summarize(warpbreaks, mean(breaks), sd(breaks))
##   mean(breaks) sd(breaks)
## 1     28.14815   13.19864

plyr::summarize()

  • Usually add custom names
summarize(warpbreaks, 
  Avg = mean(breaks), 
  StdDev = sd(breaks))
##        Avg   StdDev
## 1 28.14815 13.19864

plyr::summarize()

  • Applied functions can be complex
summarize(warpbreaks, 
  Avg = mean(breaks), 
  Med = median(breaks), 
  Dif = abs(Avg - Med)) #Can use objects created within function
##        Avg Med      Dif
## 1 28.14815  26 2.148148
  • What capability are we missing here with summarize()?

ddply()

  • Also in the plyr package
  • “Split data frame, apply function, and return results in a data frame.” — ?ddply
  • Hadley Wickham (2011). “The Split-Apply-Combine Strategy for Data Analysis”, Journal of Statistical Software, 40(1), 1-29
  • dataframe in, dataframe out, apply

ddply() arguments

ddply() arguments:

  • .data = data frame to be processed
  • .variables = variables to split data frame by
    • a character vector
    • as as.quoted() variables
    • a formula
  • .fun = function to apply to each piece
  • ... = arguments passed to .fun (e.g., na.rm)
    • in our case .fun is summarize()

ddply() in action

  • Often used with summarize() to apply more than one function
ddply(warpbreaks, "wool", summarize,
   Avg = mean(breaks),
   Med = median(breaks),
   Dif = abs(Avg - Med))
##   wool      Avg Med      Dif
## 1    A 31.03704  26 5.037037
## 2    B 25.25926  24 1.259259

ddply(): .() notation

 ddply(warpbreaks, .(wool, tension), summarize,
   Avg = mean(breaks),
   Med = median(breaks),
   Dif = abs(Avg - Med))
##   wool tension      Avg Med       Dif
## 1    A       L 44.55556  51 6.4444444
## 2    A       M 24.00000  21 3.0000000
## 3    A       H 24.55556  24 0.5555556
## 4    B       L 28.22222  29 0.7777778
## 5    B       M 28.77778  28 0.7777778
## 6    B       H 18.77778  17 1.7777778
  • . is a function that simply quotes the names of each variable (see ?.)
    • does not evaluate the variable
    • simply allows you to drop quotes

ddply(): formula interface

 ddply(warpbreaks, ~wool+tension, summarize,
   Avg = mean(breaks),
   Med = median(breaks),
   Dif = abs(Avg - Med))
##   wool tension      Avg Med       Dif
## 1    A       L 44.55556  51 6.4444444
## 2    A       M 24.00000  21 3.0000000
## 3    A       H 24.55556  24 0.5555556
## 4    B       L 28.22222  29 0.7777778
## 5    B       M 28.77778  28 0.7777778
## 6    B       H 18.77778  17 1.7777778

Three ways to specify variables

Method Syntax Notes
Character vector "wool" Quoted, single variable
.() notation .(wool, tension) Unquoted, multiple variables
Formula ~wool+tension Familiar from aggregate()

All three produce the same result. Use whichever you find most readable.

The d_ply() family

What if we want to output a list? Use dlply()…

dlply(warpbreaks, .(wool, tension), summarize,
   Avg = mean(breaks),
   Med = median(breaks),
   Dif = abs(Avg - Med))
## $A.L
##        Avg Med      Dif
## 1 44.55556  51 6.444444
## 
## $A.M
##   Avg Med Dif
## 1  24  21   3
## 
## $A.H
##        Avg Med       Dif
## 1 24.55556  24 0.5555556
## 
## $B.L
##        Avg Med       Dif
## 1 28.22222  29 0.7777778
## 
## $B.M
##        Avg Med       Dif
## 1 28.77778  28 0.7777778
## 
## $B.H
##        Avg Med      Dif
## 1 18.77778  17 1.777778
## 
## attr(,"split_type")
## [1] "data.frame"
## attr(,"split_labels")
##   wool tension
## 1    A       L
## 2    A       M
## 3    A       H
## 4    B       L
## 5    B       M
## 6    B       H

The d_ply() family

What if we want to output an array? Use daply()…

daply(warpbreaks, .(wool, tension), summarize,
   Avg = mean(breaks),
   Med = median(breaks),
   Dif = abs(Avg - Med))
## , ,  = Avg
## 
##     tension
## wool L        M        H       
##    A 44.55556 24       24.55556
##    B 28.22222 28.77778 18.77778
## 
## , ,  = Med
## 
##     tension
## wool L  M  H 
##    A 51 21 24
##    B 29 28 17
## 
## , ,  = Dif
## 
##     tension
## wool L         M         H        
##    A 6.444444  3         0.5555556
##    B 0.7777778 0.7777778 1.777778

The d_ply() family

By far, ddply() is most common because dataframes are easiest to work with.

Other functions that produce dataframes:

  • ldply()
  • adply()
  • mdply()

Practice with ddply()

Import the birthwt data set from the MASS package. For all combinations of race and smoking status, find the:

  • number of observations within each group
  • mean and std. dev. of:
    • mother’s weight at last menstrual period
    • birth weight
  • median number of physician visits during the first trimester

library(MASS)
data(birthwt)
str(birthwt)
## 'data.frame':    189 obs. of  10 variables:
##  $ low  : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ age  : int  19 33 20 21 18 21 22 17 29 26 ...
##  $ lwt  : int  182 155 105 108 107 124 118 103 123 113 ...
##  $ race : int  2 3 1 1 1 3 1 3 1 1 ...
##  $ smoke: int  0 0 1 1 1 0 0 0 1 1 ...
##  $ ptl  : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ ht   : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ ui   : int  1 0 0 1 1 0 0 0 0 0 ...
##  $ ftv  : int  0 3 1 2 0 0 1 1 1 0 ...
##  $ bwt  : int  2523 2551 2557 2594 2600 2622 2637 2637 2663 2665 ...

head(birthwt)
##    low age lwt race smoke ptl ht ui ftv  bwt
## 85   0  19 182    2     0   0  0  1   0 2523
## 86   0  33 155    3     0   0  0  0   3 2551
## 87   0  20 105    1     1   0  0  0   1 2557
## 88   0  21 108    1     1   0  0  1   2 2594
## 89   0  18 107    1     1   0  0  1   0 2600
## 91   0  21 124    3     0   0  0  0   0 2622

ddply(birthwt, .(race, smoke), summarize, 
  N = length(bwt),
  MW.mean = mean(lwt),
  MW.sd = sd(lwt),
  BWT.mean = mean(bwt),
  BWT.sd = sd(bwt),
  FTV.med = median(ftv) )
##   race smoke  N  MW.mean    MW.sd BWT.mean   BWT.sd FTV.med
## 1    1     0 44 138.8409 25.70037 3428.750 710.0989     1.0
## 2    1     1 52 126.3077 30.76383 2826.846 626.4725     0.0
## 3    2     0 16 149.4375 43.40809 2854.500 621.2543     0.5
## 4    2     1 10 142.6000 34.52921 2504.000 637.0568     0.0
## 5    3     0 55 119.1455 19.04422 2815.782 709.3493     0.0
## 6    3     1 12 124.0000 44.58292 2757.167 810.0446     0.0

Wrap in round() for cleaner output:

round(ddply(birthwt, .(race, smoke), summarize,
  N = length(bwt),
  MW.mean = mean(lwt),
  MW.sd = sd(lwt),
  BWT.mean = mean(bwt),
  BWT.sd = sd(bwt),
  FTV.med = median(ftv) ) )
##   race smoke  N MW.mean MW.sd BWT.mean BWT.sd FTV.med
## 1    1     0 44     139    26     3429    710       1
## 2    1     1 52     126    31     2827    626       0
## 3    2     0 16     149    43     2854    621       0
## 4    2     1 10     143    35     2504    637       0
## 5    3     0 55     119    19     2816    709       0
## 6    3     1 12     124    45     2757    810       0

aggregate() vs ddply()

Feature aggregate() ddply()
Package base R plyr
Functions per call One Many
Variable specification Formula (y ~ x) Character, .(), or formula
Output Data frame Data frame (also list, array via dlply, daply)
Custom column names No (uses function name) Yes (name each summary)
Cross-referencing summaries No Yes (Dif = abs(Avg - Med))
  • Use aggregate() for quick one-function summaries
  • Use ddply() when you need multiple summaries in one table

Summary

  • plyr::summarize() applies multiple functions but has no grouping
  • ddply() = summarize() + grouping = the full toolkit
  • dataframe in, dataframe out, apply
  • Three equivalent ways to specify grouping variables
  • The d_ply() family: ddply, dlply, daply for different output types