I want to add a shorter column to my dataset with the function "merge",
it
should be filled with NAs wo be as long as the other colums, like this:
id age
9 46
8 56
6 52
5 NA
4 NA
3 NA
1 NA
i did this:
pa1 <- merge(pa1, an1, by="mergeid")
and it says "'by' must specify uniquely valid column(s)"
Hi,
You could use sort=FALSE in ?merge()
d1<-data.frame(id=c(9,8,6,5,4,3,1))
d2<-data.frame(id=c(9,8,6),age=c(46,56,52))
?d<-merge(d1,d2,all.x=TRUE,sort=FALSE)
#or
library(plyr)
join(d1,d2,by="id",type="full")
#? id age
#1? 9? 46
#2? 8? 56
#3? 6? 52
#4? 5? NA
#5? 4? NA
#6? 3? NA
#7? 1? NA
A.K.
----- Original Message -----
From: Pete Brecknock <Peter.Brecknock at bp.com>
To: r-help at r-project.org
Cc:
Sent: Friday, December 7, 2012 9:12 AM
Subject: Re: [R] how to add a column from another dataset with "merge"
kiotoqq wrote
I want to add a shorter column to my dataset with the function "merge",
it
should be filled with NAs wo be as long as the other colums, like this:
id? ? age
9? ? 46
8? ? 56
6? 52
5? NA
4? NA
3? NA
1? NA
i did this:
pa1 <- merge(pa1, an1, by="mergeid")
and it says "'by' must specify uniquely valid column(s)"