Selecting a subset of columns in R data.table using a vector with a constant column

user_3915051 0
Viewed 53

In data.table, one can use the with = FALSE to select columns using a vector. For instance:

dt <- data.table(one = 1:3, two = 2:4, three = 3:5)
dt[, .(one, three)]
parlist <- c("one", "three")
dt[, parlist, with = FALSE]

One can also use the j argument to add a new, constant column:

dt[, .(one, new = 10)]

But is it possible to combine the two? I.e., use a vector to select the columns, but such that it includes a new, constant column. Solutions like parlist <- c("one", "new" = 10) do not seem to work.

data.table r
1 Answers

Combine .SDcols with c(.SD, .(new=10)):

dt <- data.table(one = 1:3, two = 2:4, three = 3:5)
parlist <- c("one", "three")

dt[, c(.SD, .(new = 10)), .SDcols = parlist]
#      one three   new
#    <int> <int> <num>
# 1:     1     3    10
# 2:     2     4    10
# 3:     3     5    10
user_33582720
Related