CODEWITHSUNDEEP
rfeatherbit64integer64
LeGeniusII
LeGeniusII

Reputation: 960

R feather integer64 preserve large numbers

I have a dataset with some large numbers:

require(pacman)
p_load(bit64,
       feather,
       data.table)

sampleDT <- data.table(ID = c(1,2,3),
                       Val = c(11111111111,2222222222,3333))
sampleDT[, Val := bit64::as.integer64(Val)]     # convert it to integer64 type
sampleDT              # result looks fine

   ID         Val
1:  1 11111111111
2:  2  2222222222
3:  3        3333

However, when I use R feather to save this data and reload, the large numbers are not preserved.

feather::write_feather(x = sampleDT, path = "C:/Users/xxx/Downloads/test.feather")
resultDT <- feather::read_feather(path = "C:/Users/xxx/Downloads/test.feather")
resultDT
# A tibble: 3 x 2
     ID       Val
  <dbl>     <dbl>
1     1 5.49e-314
2     2 1.10e-314
3     3 1.65e-320

Is this a genuine issue with integer64 type? What should I do to preserve this large numbers?

Upvotes: 1

Views: 93

Answers (0)

Related Questions