In R, how do I add a max by group? [duplicate]

Answer 1

We can use data.table assignment (:=) to create a column in place without copying

library(data.table)
setDT(x)[, max.per.group := max(replicate), by = group]
x
#     group replicate max.per.group
#1:     A         1             5
#2:     A         2             5
#3:     A         3             5
#4:     A         4             5
#5:     A         5             5
#6:     B         1             2
#7:     B         2             2
#8:     C         1             3
#9:     C         2             3
#10:    C         3             3

data

x <- data.frame(group,replicate)

Answer 2

you can use the plyr package:

library(plyr)
> ddply(x, .(group), transform, max.per.group=max(replicate))
   group replicate max.per.group
1      A         1             5
2      A         2             5
3      A         3             5
4      A         4             5
5      A         5             5
6      B         1             2
7      B         2             2
8      C         1             3
9      C         2             3
10     C         3             3
> 

Answer 3

Try

# This is how you create your data.frame
group<-c("A","A","A","A","A","B","B","C","C","C")
replicate<-c(1,2,3,4,5,1,2,1,2,3)
x<-data.frame(group,replicate) # here you don't need c()

# Here's my solution    
Max <- tapply(x$replicate, x$group,max)
data.frame(x, max.per.group=rep(Max, table(x$group)))
 group replicate max.per.group
1      A         1             5
2      A         2             5
3      A         3             5
4      A         4             5
5      A         5             5
6      B         1             2
7      B         2             2
8      C         1             3
9      C         2             3
10     C         3             3

Answer 4

Here is an other base R solution:

cbind(x, cummax=unlist(tapply(x$replicate, x$group, function(x) rep(max(x), length(x)))))
   group replicate cummax
A1     A         1      5
A2     A         2      5
A3     A         3      5
A4     A         4      5
A5     A         5      5
B1     B         1      2
B2     B         2      2
C1     C         1      3
C2     C         2      3
C3     C         3      3

Answer 5

The shared reproducible example suggests that you have the columns as factors. We need to convert them to numeric first.

We can try with base R ave. Finding out the max in each group.

x$max.per.group <- ave(x$replicate, x$group, FUN = function(x) max(as.numeric(x)))

#   group replicate max.per.group
#1      A         1             5
#2      A         2             5
#3      A         3             5
#4      A         4             5
#5      A         5             5
#6      B         1             2
#7      B         2             2
#8      C         1             3
#9      C         2             3
#10     C         3             3

---

Another option with dplyr

library(dplyr)
x %>%
   group_by(group) %>%
   mutate(max.per.group = max(as.numeric(replicate)))

Answer 6

If you redefine x first (the cbind makes both columns factors),

x<-data.frame(group,replicate)

you can use this:

merge(x,aggregate(replicate~group,x,FUN=max),all.x=TRUE,by="group")
   group replicate.x replicate.y
1      A           1           5
2      A           2           5
3      A           3           5
4      A           4           5
5      A           5           5
6      B           1           2
7      B           2           2
8      C           1           3
9      C           2           3
10     C           3           3