use runif to calculate a column in data.table

Question

I would like to generate a random number using runif for each row in a data table. Unfortunately, I end up with the same number in every row.

require(data.table)
dx <- data.table( min=1:10, max=seq(10,100,10))
dx[, sum:=min+max]
dx[, runif:=runif(n=1,min,max)]

> dx
    min max sum    runif
 1:   1  10  11 5.086488
 2:   2  20  22 5.086488
 3:   3  30  33 5.086488
 4:   4  40  44 5.086488
 5:   5  50  55 5.086488
 6:   6  60  66 5.086488
 7:   7  70  77 5.086488
 8:   8  80  88 5.086488
 9:   9  90  99 5.086488
10:  10 100 110 5.086488

Why does runif behave this way but addition works just fine? Can someone suggest how to use the min and max on each row to generate a uniformly distributed random number for that row?

Answer 1

I think just

dx[, runif := runif(.N, min, max)]

should do it. By setting n=1 you were requesting that a single value be generated, which was then recycled to the appropriate length. min and max are still handled in the appropriate, vectorized way.