This is not an aproximation, just an inconsequent way of rounding error. If
you perform same calculations by excel formulas in cells you never get such a
bombastic approximation error. In your case, the value shown as 33% becomes
33.3 % when you increase the decimal. Here, what the problem is that 31.5 %
becomes 34 % upon decreasing decimal or vice versa, which denies the
theoritical rounding rules. When calculated in cells it is normal and
correct. But charting the same values gives such kind of errors.
"Bernard Liengme" wrote:
> What sort of comment do you want?
> If I divide a pie in three part, father gets 33%, mother gets 33% and baby
> get 33%. But that is only 99%; did Goldilocks eat the other 1%. Oh no, you
> say, every one gets 33.333333% but that still adds to 99.999999%. This time
> G gets a crumb.
> When ever you make approximations, you get approximations.
> best wishes
> --
> Bernard V Liengme
> www.stfx.ca/people/bliengme
> remove caps from email
>
> "Murat Gulbay" <Murat (E-Mail Removed)> wrote in message
> news:948E7A02-0EF8-4E01-91F9-(E-Mail Removed)...
> > (Excel 2003)
> > Suppose sales values for January to June are 25,50,90,45,100, and 75,
> > respectively.
> > Bar of pie chart with a second plot containing the sales values whose
> > percentage values are less than 15%. The problem is that, Excel draws the
> > pie
> > chart with a total percentage of 102 %. The percentage of the total sales
> > included in the bar is 33%. Now if you increase its decimal, it is seen
> > that
> > its actual value is 31.2 %. How Excel 2003 rounds 31.2 % as 33 %. Now
> > changing sales value for April from 45 to 47, it is seen that bar
> > percentage
> > becomes 34 %, and total percentage for the pie becomes 103 %... When
> > increasing the decimal for 34 % it has changed to 31.5 percent. Chart
> > percentages fail for some particular data like this.. Any comments about
> > this?
>
>
>