Part of the point of the discussion is that your series is incorrect, in that
it arbitrarily includes a datapoint that shouldn't be there -- and without
that datapoint, your claim of a supposed imbalance in rounding methods falls
I could just as easily pick a different, but equally arbitrary, series to
"prove" that an *opposite* imbalance exists, but that "proof" would be no
more, or less, valid than yours.
ROTFLMAO -- in other words, don't bring in anything that would demonstrate
I'm still waiting for you to cite a source for your claims about the Cray XP.
Doug Miller (alphageek at milmac dot com)
On Dec 15, 10:46 pm, email@example.com (Doug Miller) wrote:
Rounding. And I cite:
With all rounding schemes there are two possible outcomes: increasing
the rounding digit by one or leaving it alone. With traditional
rounding, if the number has a value less than the half-way mark between
the possible outcomes, it is rounded down; if the number has a value
exactly half-way or greater than half-way between the possible
outcomes, it is rounded up. The round-to-even method is the same except
that numbers exactly half-way between the possible outcomes are
sometimes rounded up-sometimes down.
Although it is customary to round the number 4.5 up to 5, in fact 4.5
is no nearer to 5 than it is to 4 (it is 0.5 away from either). When
dealing with large sets of scientific or statistical data, where trends
are important, traditional rounding on average biases the data upwards
slightly. Over a large set of data, or when many subsequent rounding
operations are performed as in digital signal processing, the
round-to-even rule tends to reduce the total rounding error, with (on
average) an equal portion of numbers rounding up as rounding down. This
generally reduces the upwards skewing of the result.
Round-to-even is used rather than round-to-odd as the latter rule would
prevent rounding to a result of zero.
3.016 rounded to hundredths is 3.02 (because the next digit (6) is 6 or
3.013 rounded to hundredths is 3.01 (because the next digit (3) is 4 or
3.015 rounded to hundredths is 3.02 (because the next digit is 5, and
the hundredths digit (1) is odd)
3.045 rounded to hundredths is 3.04 (because the next digit is 5, and
the hundredths digit (4) is even)
3.04501 rounded to hundredths is 3.05 (because the next digit is 5, but
it is followed by non-zero digits)
I keep seeing you reply to yourself and had to look. As I suspected, you
may as being talking to a mirror. Doug is relentless and does not know how
to loose gracefully. He is one of those type people that cannot pass up a
good argument regardless on which side he is on. You are wasting your time
trying to explain any thing to him if he has set his mind to ignore facts.
Rob: Here, I have a 12" oak stick.
Doug: Your stick should be 11.5"
Rob, Don't change the argument.
Doug: Cite where you say your stick is oak. It is pine.
Rob: (after a couple of tries of trying to bring Doug back to reality,
that this stick, in fact MY stick, *I* made it, *IS* both 12" and made
from oak.) realizes Doug is a troll.
Doug: (Realizing he doesn't have a leg to stand on) ": It is not a
stick, it is a baton, cite where your stick isn't a baton.)
Rob: Wants to toss the stick one more time, but Doug has decided to
chase an 11.5" pine stick instead, so Rob won't play any more.
Rob: I have a qt of stain and it is enough for this table.
Doug: When painting ocean liners, a qt won't be enough and stain won't
The man is a troll.
Dunno what he claimed, but all the gory details are here:
Note that there are decimal floating point values that cannot
be represented in IEEE 754 floating point (which is implemented
by pretty much every processor in existence today).
Note that bankers generally do _NOT_ use binary floating
point for financial calculations, but rather use fixed-point
arithmetic (or even integer arithmetic denominated in pennies,
hundreths of a penny, or mils).
Many of the early mainframes used BCD arithmetic for this.
On Dec 15, 8:50 pm, firstname.lastname@example.org (Doug Miller) wrote:
If you knew what it was called, why didn't you tell us, Doug?
Are you trying to tell me that it wouldn't be able to?
No programmer could make a Cray round in any way?
No way? In financial or scientific models, there couldn't be any
BTW, a bank if 1100 G5 Macintosh computers blew away a Cray a few years
ago. (THIS time, go look it up before shooting your mouth off again.)
I rest my case. Another strawman up in flames.
You asserted that id *did*, i.e. that it was constructed that way. Cite,
Could, yes. Would, no -- because it would be incorrect. Standard rounding is
that anything between .00 and .499999.... gets rounded down, .50 to .99999...
gets rounded up. That's the way software rounding works -- and hardware
rounding, too, in the machines that have it.
Doug Miller (alphageek at milmac dot com)
Doug Miller (in wVJgh.13775$ email@example.com)
| Could, yes. Would, no -- because it would be incorrect.
|| I rest my case. Another strawman up in flames.
Seems to me that both of you are missing the point. Rounding is
nothing more than a convenience for dealing with _errors_ - and it's
ocurred to me that an argument over the /correctness/ of an error is
almost guaranteed to produce a lot more heat than light.
DeSoto, Iowa USA
First off, your "sequence" should stop at 10.9, on 11.0 it begins a
new sequence, just at your example started at 10.0
Second, Crays are not used for routine financial transactions like
interest calculations, they would be done on run-of-the-mill
mainframes or AS400 type systems.
Third, maybe you're just joking?
When the game is over, the pawn and the king are returned to the same box.
Larry Wasserman - Baltimore Maryland - firstname.lastname@example.org
On Dec 16, 2:10 am, email@example.com () wrote:
LOL.. I know, Larry. 12 wickets, 11 spaces in between.
That would have been a different sequence than I presented. The
sequence I presented (which could have ended in 10.9999999999999999999,
does illustrate, and magnifies greatly the errors made in rounding.
That's all it is supposed to do. It cracks me up that a simple
illustration which says that there are many ways to deal with rounding
errors, which the attendant at the gas-bar (I think in Mr. Magan's
post) may have applied (humourous in its unlikelyness) has evolved,
thanks to Mr. Miller, into a flap about very little.
I guess I'm guilty of 'working' Mr. Miller a little, but he needs to
stop drinking coffee.
Okay, let me re-phrase. When shoving a lot of really big calculations
through a really big computer, rounding errors count for something, and
not all rounding methods end up with the same results. Would an AS400
as an example of a big computer been as recognizable as a Cray? All
Miller did, was to jump all over one word, out of a whole topic, in the
faint hopes that he could demonstrate his vast intellect so that people
would not become hip to his small penis.
I never joke.
okay... maybe almost ( 96.334 % oops, make that 96.4 %) of the time.
I often joke around, but Miller just isn't funny. There are a couple of
people in here who have no sense of humour. Now they're both in the
I like that line. Both black and white chess pieces also end up in the
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.