[R] convert a character string to a name

[jpm miao]

Hi,
From time to time I need to do the aggregation. To illustrate, I present
a toy example as below. In this example, the task is to aggregate x and y
by z with the function mean.
Could I call the aggregation function with x_test, where
x_test=c("x","y")? Thanks

Miao


> dftest<-data.frame(x=1:12, y=(1:12)%%4, z=(1:12)%%2)
> dftest
x y z
1 1 1 1
2 2 2 0
3 3 3 1
4 4 0 0
5 5 1 1
6 6 2 0
7 7 3 1
8 8 0 0
9 9 1 1
10 10 2 0
11 11 3 1
12 12 0 0
> aggregate(cbind(x,y)~z, data=dftest, FUN=mean)
z x y
1 0 7 1
2 1 6 2
> x_test=c("x","y")
> aggregate(cbind(x_test)~z, data=dftest, FUN=mean)
Error in model.frame.default(formula = cbind(x_test) ~ z, data = dftest) :
variable lengths differ (found for 'z')
a1aggregate(cbind(factor(x_test))~z, data=dftest, FUN=mean)
Error in model.frame.default(formula = cbind(factor(x_test)) ~ z, data =
dftest) :
variable lengths differ (found for 'z')
> aggregate(factor(x_test)~z, data=dftest, FUN=mean)
Error in model.frame.default(formula = factor(x_test) ~ z, data = dftest) :
variable lengths differ (found for 'z')

[[alternative HTML version deleted]]

______________________________________________
R-h...@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[arun]

 with(dftest,aggregate(cbind(x,y),list(z),FUN=mean))
#  Group.1 x y
#1       0 7 1
#2       1 6 2


#or
library(plyr)
ddply(dftest,.(z),numcolwise(mean))
#  z x y
#1 0 7 1
#2 1 6 2
A.K.

[Blaser Nello]

If you want to use the character string:

attach(dftest)
aggregate(cbind(sapply(x_test, get))~z, data=dftest, FUN=mean)
# or
with(dftest,aggregate(cbind(sapply(x_test, get)),list(z),FUN=mean))
detach(dftest)

Cheers,
Nello

[jim holtman]

try this:

> dftest<-data.frame(x=1:12, y=(1:12)%%4, z=(1:12)%%2)

> aggregate(cbind(x,y)~z, data=dftest, FUN=mean)
z x y
1 0 7 1
2 1 6 2
> x_test=c("x","y")

> a <- formula(paste0('cbind('
+ , x_test[1]
+ , ','
+ , x_test[2]
+ , ') ~ z'
+ ))
> a
cbind(x, y) ~ z
> aggregate(a, data = dftest, FUN = mean)

--
Jim Holtman
Data Munger Guru

What is the problem that you are trying to solve?
Tell me what you want to do, not how you want to do it.

[arun]

Sorry, didn't read your question properly

#Just a modification without attach():
 aggregate(cbind(sapply(x_test,get,dftest))~z,data=dftest,FUN=mean)

#  z x y
#1 0 7 1
#2 1 6 2

#if you need to aggregate() all the columns except the grouping column

 aggregate(.~z,data=dftest,FUN=mean)

[Greg Snow]

Here are a couple of approaches:

> dftest<-data.frame(x=1:12, y=(1:12)%%4, z=(1:12)%%2)

> x_test=c("x","y")

> aggregate( dftest[,x_test], dftest['z'], FUN=mean )

z x y
1 0 7 1
2 1 6 2
>

> ### Or
>
> tmp.f <- as.formula( paste( 'cbind(',
+ paste( x_test, collapse=',' ),
+ ') ~ z' ) )
> aggregate( tmp.f, data=dftest, FUN=mean )

z x y
1 0 7 1
2 1 6 2

The first just uses x_test to subset the data frame and sends the
constructed subset to aggregate. The second constructs the formula from
the strings and passes the formula to aggregate.

--
Gregory (Greg) L. Snow Ph.D.
538...@gmail.com