[AccessD] bust out the math books

Stuart:

Fortran resolved these sort of issues by using 64/128 byte floating numbers.
The values were so large that a round-off error would disappear if a result
was displayed with a mere 10 significant digits. 

Of course these errors may simply be the result of bad coding but I took the
stance that if the programming was correct then core round-off errors were
the logical culprit and whether the issue was displayed in 20 digits or 2 is
irrelevant. 

Regards
Jim 

-----Original Message-----
From: accessd-bounces at databaseadvisors.com
[mailto:accessd-bounces at databaseadvisors.com] On Behalf Of Stuart McLachlan
Sent: January 25, 2006 8:52 AM
To: Access Developers discussion and problem solving
Subject: Re: [AccessD] bust out the math books

On 25 Jan 2006 at 7:24, Jim Lawrence wrote:

 
> When tax totals are generated then all the tax subtotals are just added
> together. In theory, all the subtotals could be added and the total tax
> could be calculated but the answer would be wrong. This is the same
problem
> experience when trying to calculate the average by item and then expecting
> the average for the whole group to add up. It never will.
> 

No, the two are completely separate problems.

Tax totalling problems are caused by rounding errors inherent in the 
process of working with currency which has a minimum sized unit (a cent or 
whatever).  Averaging problems are caused by false assumptions about the 
underlying maths by the  person writing the code.



-- 
Stuart


-- 
AccessD mailing list
AccessD at databaseadvisors.com
http://databaseadvisors.com/mailman/listinfo/accessd
Website: http://www.databaseadvisors.com