Subsetting data.table columns using []

Question

This is very strange. When I try to select columns on my data.table by doing

df1[, 30]

It just gives me 30, or whatever number I put in there. Not column 30.

I've checked, and everything works properly when I just produce a test data.frame:

df2 <- data.frame(x = 1:3, y = 3:1, z = 7:9)
> df2[, 2]
[1] 3 2 1

Any ideas on what might be happening?

data frame or a data table? the behavior you described looks like a data table one. — SabDeM
– SabDeM, Commented May 4, 2015 at 18:53
My bad, it is a data table. I'll convert it into a data frame and hopefully that will fix this! — Anshu Chen
– Anshu Chen, Commented May 4, 2015 at 18:55
Just ignore my comment. I thought you want it as a column. If it is a data.table, df2[,2, with=FALSE] — akrun
– akrun, Commented May 4, 2015 at 18:55
For reference, it's the first question in the faq :) github.com/Rdatatable/data.table/wiki/vignettes/… You may want to edit the question title (since it says "data frame") — Frank
– Frank, Commented May 4, 2015 at 18:56

MichaelChirico · Accepted Answer · 2020-04-22 12:30:53Z

8

When working with data.table you need to use following to chose column by numbers:

df2[, 2]
df2[, .SD, .SDcols=2]

This will still return data.table, not a vector.
As always on list you can also use below to return a vector:

df2[[2]]

MichaelChirico

34.9k17 gold badges122 silver badges209 bronze badges

answered May 4, 2015 at 18:56

jangorecki

16.8k5 gold badges86 silver badges169 bronze badges

Sign up to request clarification or add additional context in comments.

1 Answer 1