Selecting only numeric columns from a data frame

Question

Suppose, you have a data.frame like this:

x <- data.frame(v1=1:20,v2=1:20,v3=1:20,v4=letters[1:20])

How would you select only those columns in x that are numeric?

mdsumner · Accepted Answer · 2022-05-16T11:30:43.843

374

EDIT: updated to avoid use of ill-advised sapply.

Since a data frame is a list we can use the list-apply functions:

nums <- unlist(lapply(x, is.numeric), use.names = FALSE)

Then standard subsetting

x[ , nums]

## don't use sapply, even though it's less code
## nums <- sapply(x, is.numeric)

For a more idiomatic modern R I'd now recommend

x[ , purrr::map_lgl(x, is.numeric)]

Less codey, less reflecting R's particular quirks, and more straightforward, and robust to use on database-back-ended tibbles:

dplyr::select_if(x, is.numeric)

Newer versions of dplyr, also support the following syntax:

x %>% dplyr::select(where(is.numeric))

edited May 16 '22 at 11:30

answered May 02 '11 at 22:28

mdsumner

29,099
6
83
91

12

`x[nums]` or `x[sapply(x,is.numeric)]` works as well. And they always return `data.frame`. Compare `x[1]` vs `x[,1]` - first is `data.frame`, second is a vector. If one want to prevent conversion then must use `x[, 1, drop=FALSE]` . – Marek May 03 '11 at 11:46
Any way to select continuous data only? This method returns continuous as well as integer. – derelict Aug 04 '16 at 19:28
When there is no numeric column, the following error arise `undefined columns selected`. How do you avoid it ? – Yohan Obadia Aug 13 '16 at 14:15
@SoilSciGuy continuous data should be as.numeric. Perhaps you have factor data that's in numeric form? You should open a new question. – Brandon Bertelsen Sep 02 '16 at 23:38
1

@YohanObadia You can use a `tryCatch()` to deal with this. Please consider opening a new question. – Brandon Bertelsen Sep 02 '16 at 23:39
@SoilSciGuy `x[sapply(x, is.numeric) & ! sapply(x, is.integer)]` – Gregor Thomas Mar 23 '20 at 18:34
why if I do that with apply(data,2,is.numeric) it does not work, even if the output is identical to unlist(lapply(x, is.numeric)) @mdsummer – Jenny Jun 16 '20 at 17:47
apply is for arrays, not for data frames that can have columns of different types - if you apply() you get all columns the same type, and so some are character and then all are character – mdsumner Jun 17 '20 at 00:07

score 95 · Answer 2 · answered Nov 25 '16 at 16:08

95

The dplyr package's select_if() function is an elegant solution:

library("dplyr")
select_if(x, is.numeric)

answered Nov 25 '16 at 16:08

Sharon

3,676
3
23
20

1

select_if has been superseded, any idea how to do this with the current version? – Max F Nov 01 '21 at 03:27
2

select(where(is.numeric)) https://tidyselect.r-lib.org/reference/where.html – bryn Jan 21 '22 at 06:27

Kevin Zarca · Answer 3 · 2019-10-02T09:13:24.647

57

Filter() from the base package is the perfect function for that use-case: You simply have to code:

Filter(is.numeric, x)

It is also much faster than select_if():

library(microbenchmark)
microbenchmark(
    dplyr::select_if(mtcars, is.numeric),
    Filter(is.numeric, mtcars)
)

returns (on my computer) a median of 60 microseconds for Filter, and 21 000 microseconds for select_if (350x faster).

edited Oct 02 '19 at 09:13

answered Nov 09 '16 at 10:31

Kevin Zarca

2,572
1
18
18

This solution doesn't fail when no numeric columns are present. Are there any drawbacks to using it? – bli Nov 22 '16 at 10:10
Filter only applies to rows of a dataframe rather than columns. As such, this solution wouldn't give the correct result. – Michael Jan 18 '17 at 11:45
4

@Michael don't confuse Filter from the base package and filter from dplyr package! – Kevin Zarca Feb 01 '17 at 14:45
1

@bli I can't see any drawback of using Filter. Its input is a data.frame object and it return a data.frame – Kevin Zarca Feb 01 '17 at 14:47
Just chiming in here for reference: what ```Filter()``` doesn't work for here is replacing, e.g. ```Filter(is.numeric,iris) <- 0.5*Filter(is.numeric,iris)``` won't work. – Mobeus Zoom Jun 06 '20 at 07:05
This one also works with data.table, thx! – Akantor Nov 18 '22 at 12:05

score 9 · Answer 4 · answered Apr 05 '18 at 09:44

9

in case you are interested only in column names then use this :

names(dplyr::select_if(train,is.numeric))

answered Apr 05 '18 at 09:44

user3065757

475
1
5
14

AlexB · Answer 5 · 2020-10-20T09:38:54.897

9

iris %>% dplyr::select(where(is.numeric)) #as per most recent updates

Another option with purrr would be to negate discard function:

iris %>% purrr::discard(~!is.numeric(.))

If you want the names of the numeric columns, you can add names or colnames:

iris %>% purrr::discard(~!is.numeric(.)) %>% names

edited Oct 20 '20 at 09:38

answered Oct 20 '20 at 07:30

AlexB

3,061
2
17
19

negating `discard()` is pretty much the same as using `keep()`. – GuedesBF Jul 16 '21 at 21:22

Enrique Pérez Herrero · Answer 6 · 2016-11-13T19:54:45.657

8

This an alternate code to other answers:

x[, sapply(x, class) == "numeric"]

with a data.table

x[, lapply(x, is.numeric) == TRUE, with = FALSE]

edited Nov 13 '16 at 19:54

answered Nov 13 '16 at 16:11

Enrique Pérez Herrero

3,699
2
32
33

3

This is more of a comment to the selected answer, rather than aunique answer. – Brandon Bertelsen Nov 13 '16 at 17:55
2

Columns can have more than one class. – Rich Scriven Nov 13 '16 at 19:56

score 6 · Answer 7 · edited Mar 23 '20 at 18:28

6

library(purrr)
x <- x %>% keep(is.numeric)

edited Mar 23 '20 at 18:28

Brandon Bertelsen

43,807
34
160
255

answered Mar 23 '20 at 15:50

Yash Khokale

61
1
1

user1 · Answer 8 · 2017-11-13T15:48:57.900

3

The library PCAmixdata has functon splitmix that splits quantitative(Numerical data) and qualitative (Categorical data) of a given dataframe "YourDataframe" as shown below:

install.packages("PCAmixdata")
library(PCAmixdata)
split <- splitmix(YourDataframe)
X1 <- split$X.quanti(Gives numerical columns in the dataset) 
X2 <- split$X.quali (Gives categorical columns in the dataset)

edited Nov 13 '17 at 15:48

answered Nov 13 '17 at 15:42

user1

391
3
27

score 1 · Answer 9 · answered Jan 06 '17 at 00:19

1

If you have many factor variables, you can use select_if funtion. install the dplyr packages. There are many function that separates data by satisfying a condition. you can set the conditions.

Use like this.

categorical<-select_if(df,is.factor)
str(categorical)

answered Jan 06 '17 at 00:19

서영재

96
1
9

2

Looks like a duplicate of this earlier answer http://stackoverflow.com/a/40808873/170352 – Brandon Bertelsen Jan 06 '17 at 04:07

score 0 · Answer 10 · edited Oct 09 '18 at 07:04

0

Another way could be as follows:-

#extracting numeric columns from iris datset
(iris[sapply(iris, is.numeric)])

edited Oct 09 '18 at 07:04

greg-449

109,219
232
102
145

answered Oct 09 '18 at 06:00

Ayushi

9
1

1

Hi Ayushi, this probably was downvoted because it's a repeat of the first answer, but this method has some issues that were identified. Take a look at the comments in the first answer, you'll see what I mean. – Brandon Bertelsen Oct 09 '18 at 18:25

score 0 · Answer 11 · edited Jul 31 '20 at 01:10

0

Numerical_variables <- which(sapply(df, is.numeric))
# then extract column names 
Names <- names(Numerical_variables)

edited Jul 31 '20 at 01:10

Brandon Bertelsen

43,807
34
160
255

answered Jul 30 '20 at 21:08

Mohamed Ali Hefnawy

21
7

Hey Mo! I recommend you try this in the console because I don't think it's going to give you what you think it will. – Brandon Bertelsen Jul 31 '20 at 01:12

score -1 · Answer 12 · answered Mar 29 '18 at 16:32

This doesn't directly answer the question but can be very useful, especially if you want something like all the numeric columns except for your id column and dependent variable.

numeric_cols <- sapply(dataframe, is.numeric) %>% which %>% 
                   names %>% setdiff(., c("id_variable", "dep_var"))

dataframe %<>% dplyr::mutate_at(numeric_cols, function(x) your_function(x))

Selecting only numeric columns from a data frame

12 Answers12

Linked

Related