How to use a variable in the i position to filter rows

Question

Example data

d<-data.table(a=c(1,2,3,4,5))

I want this result:

d[a < 3]

How do achieve the same Using my.column?

my.column<-'a'
d[my.column < 3]
#Empty data.table (0 rows) of 1 col: a

Although the following works, I hope there is a better way:

setnames(d,my.column,"my.column")
d[my.column<3,]
setnames(d,'my.column',my.column)

Note that this also does not work

setkey(d,my.column)

A similar question with some additional options: http://stackoverflow.com/q/29564002/1191259 — Frank, Aug 30 '15 at 13:55

score 0 · Accepted Answer · answered Aug 30 '15 at 12:54

As pointed by David in comment you can use

d[get(my.column) < 3]

The other way, which I personally prefer, is to build a call for i instead of getting the variable:

d[eval(as.name(my.column)) < 3]

Not sure about the performance when using in i. In case of using in j arg I believe the eval(as.name( would be marginally faster - depends on amount of columns AFAIK.

How to use a variable in the i position to filter rows

1 Answers1