More about Aggregation by Group in R

December 24, 2012
By

(This article was first published on Yet Another Blog in Statistical Computing » S+/R, and kindly contributed to R-bloggers)

Motivated by my young friend, HongMing Song, I managed to find more handy ways to calculate aggregated statistics by group in R. They require loading additional packages, plyr, doBy, Hmisc, and gdata, and are extremely user-friendly. In terms of CPU time, while the method with summarize() is as efficient as the 2nd method with by() introduced yesterday, summaryBy() in doBy package seems the slowest.

“Learn as if you were to live forever” – Mahatma Gandhi

> # METHOD 5: USING DDPLY()
> library(plyr)
> summ5 <- ddply(df, .(SELFEMPL, OWNRENT), summarize, INCOME = mean(INCOME), BAD = mean(BAD))
> print(summ5)
  SELFEMPL OWNRENT   INCOME        BAD
1        0       0 2133.314 0.08470957
2        0       1 2881.201 0.06293210
3        1       0 2742.247 0.06896552
4        1       1 3487.910 0.05316973
> 
> # METHOD 6: USING DOBy()
> library(doBy)
> summ6 <- summaryBy(INCOME + BAD ~ SELFEMPL + OWNRENT, data = df, fun = c(mean), keep.names = TRUE)
> print(summ6)
  SELFEMPL OWNRENT   INCOME        BAD
1        0       0 2133.314 0.08470957
2        0       1 2881.201 0.06293210
3        1       0 2742.247 0.06896552
4        1       1 3487.910 0.05316973
>
> # METHOD 7: USING SUMMARIZE()
> library(Hmisc)
> summ7 <- summarize(df[c('INCOME', 'BAD', 'SELFEMPL', 'OWNRENT')], df[c('SELFEMPL', 'OWNRENT')], colMeans, stat.name = NULL)
> print(summ7)
  SELFEMPL OWNRENT   INCOME        BAD
1        0       0 2133.314 0.08470957
2        0       1 2881.201 0.06293210
3        1       0 2742.247 0.06896552
4        1       1 3487.910 0.05316973
> 
> # METHOD 8: USING FRAMEAPPLY()
> library(gdata)
> summ8 <- frameApply(df, by = c('SELFEMPL', 'OWNRENT'), on = c('INCOME', 'BAD'), fun = colMeans)
> rownames(summ8) <- NULL
> print(summ8)
  SELFEMPL OWNRENT   INCOME        BAD
1        0       0 2133.314 0.08470957
2        0       1 2881.201 0.06293210
3        1       0 2742.247 0.06896552
4        1       1 3487.910 0.05316973

Efficiency Comparison

> test5 <- function(n){
+   for (i in 1:n){
+     summ5 <- ddply(df, .(SELFEMPL, OWNRENT), summarize, INCOME = mean(INCOME), BAD = mean(BAD))
+   }
+ }
> system.time(test5(10))
   user  system elapsed 
  0.524   0.068   0.622 
>
> test6 <- function(n){
+   for (i in 1:n){
+     summ6 <- summaryBy(INCOME + BAD ~ SELFEMPL + OWNRENT, data = df, fun = c(mean), keep.names = TRUE)
+   }
+ }
> system.time(test6(10))
   user  system elapsed 
  1.800   0.060   1.903 
> 
> test7 <- function(n){
+   for (i in 1:n){
+     summ7 <- summarize(df[c('INCOME', 'BAD', 'SELFEMPL', 'OWNRENT')], df[c('SELFEMPL', 'OWNRENT')], colMeans, stat.name = NULL)
+   }
+ }
> system.time(test7(10))
   user  system elapsed 
  0.236   0.020   0.274 
> 
> test8 <- function(n){
+   for (i in 1:n){
+     summ8 <- frameApply(df, by = c('SELFEMPL', 'OWNRENT'), on = c('INCOME', 'BAD'), fun = colMeans)
+     rownames(summ8) <- NULL
+   }
+ }
> system.time(test8(10))
   user  system elapsed 
  0.580   0.008   0.668 

To leave a comment for the author, please follow the link and comment on his blog: Yet Another Blog in Statistical Computing » S+/R.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.