patterns() did not consider the position of provided cols in the original DT

[hongyuanjia]

Please see the MRE below. patterns() returns the matched positions of input cols, which are 1 and 2, resulting using id_1 and id_2 as the measure.vars:

library(data.table)

dt <- data.table(
    id_1 = 1:3,
    id_2 = letters[1:3],
    letter_1 = LETTERS[1:3],
    letter_2 = LETTERS[4:6]
)

melt(dt, measure.vars = patterns("_\\d$", cols = paste0("letter_", 1:2)))
#> Warning in melt.data.table(dt, measure.vars = patterns("_\\d$", cols =
#> paste0("letter_", : 'measure.vars' [id_1, id_2] are not all of the same type.
#> By order of hierarchy, the molten data value column will be of type
#> 'character'. All measure variables not of type 'character' will be coerced too.
#> Check DETAILS in ?melt.data.table for more on coercion.
#>    letter_1 letter_2 variable  value
#>      <char>   <char>   <fctr> <char>
#> 1:        A        D     id_1      1
#> 2:        B        E     id_1      2
#> 3:        C        F     id_1      3
#> 4:        A        D     id_2      a
#> 5:        B        E     id_2      b
#> 6:        C        F     id_2      c