Re: This calculation is just wrong / computer can't count!

"Joseph M. Newcomer" <newcomer@xxxxxxxxxxxx> wrote in message
news:eo4lg3tqo6os3rglordp6qjbqpb87h5kq7@xxxxxxxxxx
No, there are only two answers you misinterpreted as answering your
question.

The guard bits answer has nothing to do with answering your question.

Your question is erroneous, your understanding is erroneous, and your
computation is
erroneous.

You are being attacked because you have simply failed to grasp the most
elementary aspects
of computer arithmetic, in spite of everyone explaining it to you. You
persist in
thinking a textual representation of a number has meaning. Only the
binary represenation
has meaning, everything else is an illusion. By insisting the illusion is
reality, you
miss the point that the answer you are getting is CORRECT. Only your
attitude is wrong.

No, the answer is wrong. The users of our application type numbers in in
decimal and the answers are shown in decimal. As the computer clearly works
in binary, there are conversions required, but my problem is with the error
that these conversions introduce. By error I am not referring to the
rounding and loss of precision, I am talking about the erroneous final digit
introduced by the conversions. As I said in another post, expressing 1/3 as
a decimal requires a precision decision: we have to decide how many decimal
places to use, but the final digit is a 3. Any other final digit is
mathematically incorrect and wrong and renders and future calculations
unreliable. My original question, which you have completely glossed over was
regarding this inaccuracy. I asked if there was a way to get the computer to
work to only the number of decimal places that can be relied upon. In other
words, how do I get my calculations to work as accurately as the Windows
Calculator or a cheap casio calculator. I have only had 1 reply that sounded
like it would help and it mentioned guard bits, which you have subsequently
told me is nothing to do with it, however you haven't provided an
alternative, simply glossed over the real question and tried to sound
intelligent by referencing years of banking experience and boasting over
losing 50,000 dollars! As I am working on an estimation and project tracking
application, I would be shot if a project lost 50,000 dollars, so need
accurate results, not extra 7s here and there!!!


.

Relevant Pages

  • Re: This calculation is just wrong / computer cant count!
    ... the error that these conversions introduce. ... but the final digit is a 3. ... how do I get my calculations to work as accurately ... project tracking application, I would be shot if a project lost 50,000 ...
    (microsoft.public.vc.mfc)
  • Re: An uncountable countable set
    ... but it may depend on the ordering of the calculations. ... the diagonal number it does not matter what digit you calculate first. ... to which every path will lead is always the same: The set of rationals ... You state, explicitly, that they were not binary representations, ...
    (sci.math)
  • [ANN] Ruby Units
    ... Many technical applications make use of specialized calculations at some ... conversions and manipulations to ensure an accurate result. ... Unit#floor:: rounds quantity down to next lower integer ...
    (comp.lang.ruby)
  • Re: This calculation is just wrong / computer cant count!
    ... but ONLY BECAUSE OF THE DIGIT HE THINKS IS WRONG!!!!! ... and they correctly reflect the calculations ... point of calculating something that should be zero when we already know with ... Obviously, GT must have some reason for making these calculations, but he ...
    (microsoft.public.vc.mfc)
  • Re: This calculation is just wrong / computer cant count!
    ... extra digit on the end that it can't handle and more importantly - how ... "floating point" in decimals: guess you have 10 digits, ... The computer should surely only store as ... rounding at the last stage, so my final calculations are rounded ...
    (microsoft.public.vc.mfc)
Loading