how to get each column as data.frame (instead of a vector) from a data.frame?

Question

Normally when you get a column, it is a vector. How can I keep it as the data.frame with the same row names and corresponding column name?

possible duplicate of [1-dimensional Matrix is changed to a vector in R](http://stackoverflow.com/questions/9949202/1-dimensional-matrix-is-changed-to-a-vector-in-r) — joran, Apr 06 '12 at 19:08
@joran Sorry didn't see the possible duplicate. I'll leave my response and if this question is closed removal of my answer is a moot point. — Tyler Rinker, Apr 06 '12 at 19:10
@TylerRinker Don't worry about it. This is just a really common question, that's all. — joran, Apr 06 '12 at 19:35

Thraupidae · Accepted Answer · 2014-01-11T18:14:45.820

16

Instead of calling the desired column with a comma i.e. data.frame[,i] use data.frame[i] to preserve the class as data.frame and also retain row names.

data.frame[,i] #As a vector
data.frame[i] #As a data.frame

edited Jan 11 '14 at 18:14

answered Apr 06 '12 at 21:43

Thraupidae

665
2
5
14

Nice, I didn't know you could do that +1 – Tyler Rinker Apr 07 '12 at 13:45
I guess that adds some value to my question post and deserve not to be closed :) – RNA Apr 08 '12 at 00:02

score 6 · Answer 2 · answered Apr 06 '12 at 19:09

6

use the argument drop = FALSE as in:

mtcars[, 1, drop = FALSE]

answered Apr 06 '12 at 19:09

Tyler Rinker

108,132
65
322
519

score 0 · Answer 3 · answered Apr 06 '12 at 19:19

If you specify a single number when subsetting a data.frame, you get a one-column data.frame. This is different than matrix subsetting, which requires a "missing" i argument to return the entire column (which it then converts to a vector).

# mtcars is a data.frame
mtcars[1]       # first column
str(mtcars[1])  # is still a data.frame
# 'data.frame':   32 obs. of  1 variable:
#  $ mpg: num  21 21 22.8 21.4 18.7 18.1 14.3 24.4 22.8 19.2 ...
# MTCARS is a matrix
MTCARS <- as.matrix(mtcars)
as.matrix(MTCARS)[1]        # only the first element
# [1] 21
str(as.matrix(MTCARS)[,1])  # the first column, as a vector
 Named num [1:32] 21 21 22.8 21.4 18.7 18.1 14.3 24.4 22.8 19.2 ...
 - attr(*, "names")= chr [1:32] "Mazda RX4" "Mazda RX4 Wag" "Datsun 710" ...

how to get each column as data.frame (instead of a vector) from a data.frame?

3 Answers3