Issue with int64 columns tolerance is not taken into account

[hannes101]

Dear all,

first of all thanks a lot for this package, it really makes it easy to detect differences in data.frames.
I would like to report an issue with int64 columns, which are part of the bit64 package.

library(bit64)
library(data.table)
df1.int64 <- structure(list(date = 20220908, val = structure(2.76818691336063e-313, class = "integer64")), row.names = c(NA, 
-1L), class = c("data.table", "data.frame"))
df2.int64 <- structure(list(date = 20220908, val = structure(2.76818691341003e-313, class = "integer64")), row.names = c(NA, 
-1L), class = c("data.table", "data.frame"))

df1.num <- structure(list(date = 20220908, val = 56028726884), row.names = c(NA, 
-1L), class = c("data.table", "data.frame"))
df2.num <- structure(list(date = 20220908, val = 56028726885), row.names = c(NA, 
-1L), class = c("data.table", "data.frame"))

diffdf(df1.num, df2.num, tolerance = 1, scale = 1)
No issues were found!
diffdf(df1.int64, df2.int64, tolerance = 1, scale = 1)
Differences found between the objects!

A summary is given below.

Not all Values Compared Equal
All rows are shown in table below

  =============================
   Variable  No of Differences 
  -----------------------------
     val             1         
  -----------------------------


All rows are shown in table below

  ===================================================
   VARIABLE  ..ROWNUMBER..     BASE        COMPARE   
  ---------------------------------------------------
     val           1        56028726884  56028726885 
  ---------------------------------------------------

Warning message:
In diffdf(df1.int64, df2.int64, tolerance = 1, scale = 1) : 
Not all Values Compared Equal

It was not really obvious to me, that it doesn't work at the beginning.

[gowerc]

We are planning a minor overhaul of the package in Q4 so will look to include this change in then. Thank you for raising.