how to deal with BigINT in R

Question

I have a BigInt number.If I try to store it in R

  R> a <- 9223372036854775807
  R> a
   [1] 9.223372e+18

As you can notice its loosing info of last few digits. I tried multiple other ways to solve this but no luck like increasing options(digits = 22) or changing to numeric, double, integer.

 > as.integer(9223372036854775807)
  [1] NA
  Warning message:
  NAs introduced by coercion to integer range 
 R> as.numeric(9223372036854775807)
  [1] 9.223372e+18
 R> as.double(9223372036854775807)
  [1] 9.223372e+18

Can anyone help me with this problem. I want to retain the same original value. I also do not want to install any external package.

You may need to check `gmp` package or `bit64` i.e. `library(gmp);as.bigz(a)` — akrun, Apr 04 '18 at 04:42
Try `options(scipen = 99)` but I think it would still loose precision. — Ronak Shah, Apr 04 '18 at 04:44
@RonakShah: Its still losing precision R> options(scipen = 99) R> as.numeric(9223372036854775807) [1] 9223372036854777856 R> as.double(9223372036854775807) [1] 9223372036854777856 — shivank agrawal, Apr 04 '18 at 04:45
@akrun : External package unfortunately can not be installed — shivank agrawal, Apr 04 '18 at 04:46

score 4 · Answer 1 · answered Apr 04 '18 at 04:46

4

We can use as.integer64 from bit64

library(bit64)
as.integer64(as.character(a))
#integer64
#[1] 9223372036854775807

answered Apr 04 '18 at 04:46

akrun

874,273
37
540
662

1

Unfortunately, I can't install external package as mentioned in question. – shivank agrawal Apr 04 '18 at 04:48
@shivankagrawal Then, probably you may have to adjust with the reduced precision, Even if you try with `as.character`, it will reduce the accuracy i.e. `as.character(a)# [1] "9223372036854775808"` Check the last digit it is rounded – akrun Apr 04 '18 at 04:48

how to deal with BigINT in R

1 Answers1

Related