Well, I'm uncomfortable with the suggested syntax. Suppose instead we add an
optional 'p' followed by a value, to mean the number of digits of precision.
Thus:
y17101e4p3
to say that the precision is three digits.
Thus without the p3 at the end it would be y17101e4, meaning the year
171010000.
With the p3 at the end it would mean:
"Some year in Milenium 171000 and the estimate is 171010000"
Ray
> Original Message
> From: Discussion of the Developing Date/Time Standards
> [mailto:[log in to unmask]] On Behalf Of Edward C. Zimmermann
> Sent: Thursday, September 15, 2011 4:43 PM
> To: [log in to unmask]
> Subject: Re: [DATETIME] End of this phase
>
> On Thu, 15 Sep 2011 13:51:43 0400, Ray Denenberg, Library of Congress
> wrote
> > > y193.9e1 is 1939 with decade precision, the 9 in 1939 is
> "estimated"
> >
> > Still need more clarification. Are you saying
> >
> > (a) "193.9e1" is commonly understood to mean the 9 in 1939 is
> "estimated"
> >
> > (b) "y193.9e1" is proposed to mean that.
>
> Normally scientific notation is used for BIG or small numbers.. In a
> number such as 123.4e20 the 4 is comonly considered an "estimated
> digit"..
>
> see http://en.wikipedia.org/wiki/Scientific_notation
>
> "Significant figures
>
> Ambiguity of the last digit in scientific notation
>
> It is customary in scientific measurements to record all the
> significant digits from the measurements, and to guess one additional
> digit if there is any information at all available to the observer to
> make a guess. The resulting number is considered more valuable than it
> would be without that extra digit, and it is considered a significant
> digit because it contains some information leading to greater precision
> in measurements and in aggregations of measurements (adding them or
> multiplying them together).
>
> Additional information about precision can be conveyed through
> additional notations. In some cases, it may be useful to know how exact
> the final significant digit is. For instance, the accepted value of the
> unit of elementary charge can properly be expressed as
> 1.602176487(40)×10−19 C,[7] which is shorthand for
> 1.602176487±0.000000040×10−19 C."
>
>
> >
> > (c) or both.
> >
> > If (a), I don't see what there is in that expression that denotes
> > estimation. I would read 193.9e1 to mean 1939 (exactly).
> If you wanted to say exactly you would be better off saying 1939..
>
> >
> > > Why the estimate? There are loads of reasons..
> >
> > I don't question the usefulness of this, I question the syntax. If
> > you are
>
> The notation I suggest is inspired by this but without the ambiguity
> and instead of "last digit" I have proposed here ALL the digits right
> of the decimal point build the estimate.. that is right of decimal
> points is an estimate.. and left is "exactly" but, of course, to the
> precision defined by the exponent..
>
> > saying (a) above, then it's news to me, that's not how I learned it,
> > but I will defer to your expertise, and in that case I am fine with
> this change.
> > If on the other hand that's not what you mean, than I am not
> > comfortable with the proposed syntax.
> >
> > Ray
>
> See also
> http://www.chem.tamu.edu/class/fyp/mathrev/mrsigfg.html
> (just found this page but seems a good intro..)
>
> 
>
> Edward C. Zimmermann, NONMONOTONIC LAB
