Good afternoon,
I'm wondering if anyone else has encountered this: One of our big monthly journals wouldn't import today, because it was unbalanced. I've seen journal imports fail before because of rounding errors, and usually if you Round the values in Excel, paste values, and re-import, you're fine.
But this time it was different. Round, RoundUp, Trunc, and MRound--all of them failed to get the debits and the credits to match. We were still off by 0.001. Looking through the values, none had any value displayed beyond the second decimal place.
I think this is a floating point error, by which I mean there's some number in our import file that isn't representable in Excel's 15 decimal place implementation of binary.
My question is, does anyone know how to figure out which number that is? I tried Dec2Bin in Excel, but it can't handle values over 511 or so.
We can successfully load the journal by rounding all values to one decimal place, thus getting rid of whatever number is throwing things. But rounding an entire 500 line journal to one decimal place from two has more impact than I'd like.
I'd prefer to find the non-representable number and change that one by 0.01, which would probably do it.
Anyone have suggestions, or worked with this before?
Thanks,
Evan