placing mean,total and max together in a table in r

Question

I have some simple commands looking into totals, means and maximums of a variable whilst another variable is an assigned value:

sum(data[data$var1==1,]$var2)
mea         


        

Answer 1

You need to use cbind for that.

cbind(sum(data[data$var1==1,]$var2),mean(data[data$var1==1,]$var2),max(data[data$var1==1,]$var2))

Example using mtcars data

mydata<-mtcars  
  l<-cbind(sum(mydata[mydata$cyl==4,]$mpg),mean(mydata[mydata$cyl==4,]$mpg),max(mydata[mydata$cyl==4,]$mpg))
l<-data.frame(l)
names(l)<-c("sum","mean","max")
> l
    sum     mean  max
1 293.3 26.66364 33.9

There is a ddply function from plyr package that does all for each categories of var1 (here cyl)

library(plyr)
ddply(mydata,.(cyl),summarize, sum=sum(mpg),mean=mean(mpg), max=max(mpg))

 ddply(mydata,.(cyl),summarize, sum=sum(mpg),mean=mean(mpg), max=max(mpg))
  cyl   sum     mean  max
1   4 293.3 26.66364 33.9
2   6 138.2 19.74286 21.4
3   8 211.4 15.10000 19.2

Answer 2

I recommend checking out the data.table package, which is like a beefed-up version of data frames. One thing it does really well (and quickly, if you have a lot of data) is summaries like this.

library(data.table)
as.data.table(mtcars)[, list(sum=sum(mpg), mean=mean(mpg), max=max(mpg)),
                      by=cyl][order(cyl)]
#   cyl   sum     mean  max
#1:   4 293.3 26.66364 33.9
#2:   6 138.2 19.74286 21.4
#3:   8 211.4 15.10000 19.2

If you want to summarize by more than one variable, just use something like by=list(cyl,vs,otherColumnNamesHere).

Answer 3

Look at the tables package, read through the vignette for the package shows how to do exactly what you are asking for.

> tabular( ( factor(cyl) + 1) ~ mpg * (sum + mean + max), data=mtcars )

             mpg             
 factor(cyl) sum   mean  max 
 4           293.3 26.66 33.9
 6           138.2 19.74 21.4
 8           211.4 15.10 19.2
 All         642.9 20.09 33.9