Discussion:
Mile?
(too old to reply)
Jack
2020-01-12 22:37:09 UTC
Permalink
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.

Does anyone know what
Tony Cooper
2020-01-12 23:19:56 UTC
Permalink
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
A short one by less than a tenth of a mile. A very short one on
water.
--
Tony Cooper - Orlando, Florida
musika
2020-01-12 23:43:07 UTC
Permalink
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
--
Ray
UK
musika
2020-01-12 23:47:08 UTC
Permalink
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Oops, km.
--
Ray
UK
Jack
2020-01-13 21:11:24 UTC
Permalink
Post by musika
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Oops, km.
OK. I see wiktionary has 'Roman mile' as a headword, giving it as 1481
J. J. Lodder
2020-01-13 14:53:19 UTC
Permalink
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,

Jan
Athel Cornish-Bowden
2020-01-14 07:58:44 UTC
Permalink
Post by J. J. Lodder
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
--
athel
J. J. Lodder
2020-01-14 10:36:35 UTC
Permalink
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
To quote wikip:
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."

When setting a standard for a physical quatity,
a meter rod, or the second,
or marching a Roman legion through their paces
there is no specific value.
What matters is how much spread there is
between different attemps to reproduce the standard.

Typical modern example: atomic clocks can't be compared
with anything else, the -are- the second.
You can only observe how they drift wrt each other,

Jan
Athel Cornish-Bowden
2020-01-14 15:01:59 UTC
Permalink
Post by J. J. Lodder
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."
Farbeit from me to argue with Wikipêia, but the way I learned it was
illustrated by the following sort of comparison:

"A Roman mile was about one and a half kilometres" is accurate but not
very precise.

"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.

Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
Post by J. J. Lodder
When setting a standard for a physical quatity,
a meter rod, or the second,
or marching a Roman legion through their paces
there is no specific value.
What matters is how much spread there is
between different attemps to reproduce the standard.
Typical modern example: atomic clocks can't be compared
with anything else, the -are- the second.
You can only observe how they drift wrt each other,
Jan
--
athel
Katy Jennison
2020-01-14 15:05:47 UTC
Permalink
Post by J. J. Lodder
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."
Farbeit from me to  argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
That's how I understand accuracy and precision too.
--
Katy Jennison
Paul Carmichael
2020-01-14 16:03:24 UTC
Permalink
Post by Katy Jennison
"A Roman mile was about one and a half kilometres" is accurate but not very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all accurate.
That's how I understand accuracy and precision too.
Ditto.
--
Paul.

https://paulc.es/elpatio
Eric Walker
2020-01-15 10:46:07 UTC
Permalink
Post by Paul Carmichael
Post by Katy Jennison
"A Roman mile was about one and a half kilometres" is accurate but not very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all accurate.
That's how I understand accuracy and precision too.
Ditto.
I agree.

If the announcer of a ball game tells his audience that the first pitch
was thrown at 7:36 pm and 22 seconds, he is being quite precise; but if
it was a day game that started at around 1 pm, he is horribly inaccurate.
--
Cordially,
Eric Walker
J. J. Lodder
2020-01-14 18:15:42 UTC
Permalink
Post by Katy Jennison
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."
Farbeit from me to argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
That's how I understand accuracy and precision too.
How would you know?
For all you know the Roman mile could have been
1.3895319 kilometres (exactly)
except for the fact that it was never defined exactly.

Compare the statute mile, which is nowadays defined to be
1,609.344 meter, (exactly)
So you can determine for anything claiming to be a statute mile
how accurate it is,

Jan
Peter Young
2020-01-14 18:53:02 UTC
Permalink
On 14 Jan 2020 ***@de-ster.demon.nl (J. J. Lodder) wrote:

[snip]
Post by J. J. Lodder
Compare the statute mile, which is nowadays defined to be
1,609.344 meter, (exactly)
So you can determine for anything claiming to be a statute mile
how accurate it is,
Except in Sweden, where the word "mil", translated confusingly as "mile"
is 10km.

Peter.
--
Peter Young, (BrE, RP), Consultant Anaesthetist, 1975-2004.
(US equivalent: Certified Anesthesiologist) (AUE Hg)
Cheltenham and Gloucester, UK. Now happily retired.
http://pnyoung.orpheusweb.co.uk
musika
2020-01-14 15:21:14 UTC
Permalink
Farbeit from me to  argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
Why do you perpetrate these ad hominem attacks in threads that have
nothing to do with him?
--
Ray
UK
J. J. Lodder
2020-01-14 16:04:03 UTC
Permalink
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by musika
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
The closest is the Roman mile c. 1.48m
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."
Farbeit from me to argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that [-]
I'am afraid so. You can talk about accuracy of a result
only if there is, at least in principle,
a 'true' value that you can relate it to.

There is not one and only true platinum (or even gold)
Roman mile buried somewhere under the Via Aurelia,

Jan
bert
2020-01-14 16:07:22 UTC
Permalink
Post by Athel Cornish-Bowden
"A Roman mile was about one and a half kilometres"
is accurate but not very precise.
"A Roman mile was 1.3895319 kilometres"
is very precise but not at all accurate.
I completely agree with that example.
phil
2020-01-14 17:16:06 UTC
Permalink
Farbeit from me to  argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
I'd say the first is accurate only in the non-technical sense that it's
a true statement, but as JJL has said nearby, there's no absolute
standard Roman mile against which the figure of 1.5 km can be judged.
The second, I'd regard as spurious precision; giving it to eight
significant figures implies that you can measure the value with only a
tiny percentage error compared to the 'true' value.
Peter Moylan
2020-01-15 01:01:28 UTC
Permalink
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."
Farbeit from me to argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
I agree with you. I'm not sure where that Wikipedia definition came
from, but it doesn't agree with what I understand the normal definition
to be. Perhaps it's used in some specialised field. Statistics, perhaps?
--
Peter Moylan http://www.pmoylan.org
Newcastle, NSW, Australia
s***@gmail.com
2020-01-15 04:27:57 UTC
Permalink
[relevant quotes not elided]
Post by Peter Moylan
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."
Farbeit from me to argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
I agree with you. I'm not sure where that Wikipedia definition came
from, but it doesn't agree with what I understand the normal definition
to be. Perhaps it's used in some specialised field. Statistics, perhaps?
Nope, "The field of statistics, where the interpretation of measurements plays a central role, prefers to use the terms bias and variability instead of accuracy and precision: bias is the amount of inaccuracy and variability is the amount of imprecision. "

You should think instead of metrology. The field of measurements.

WP cite, again:

<quote>
Accuracy has two definitions:

More commonly, it is a description of systematic errors,
a measure of statistical bias; low accuracy causes
a difference between a result and a "true" value.
ISO calls this trueness.

Alternatively, ISO defines accuracy as describing a combination of
both types of observational error above (random and systematic),
so high accuracy requires both high precision and high trueness.

Precision is a description of random errors, a measure of statistical variability.

<URL:https://en.wikipedia.org/wiki/Accuracy_and_precision#ISO_definition_(ISO_5725)>

See also ISO 5725-1 (dating to 1994) and the 2008 issue of the
"BIPM International Vocabulary of Metrology" (VIM), items 2.13 and 2.14.[1]

<URL:https://www.bipm.org/en/publications/guides/>
(One of the first links on that page is
"Evaluation of measurement data - Guide to the expression of uncertainty in measurement:
JCGM 100:200 (GUM 1995 with minor corrections)

I would suspect that this is applicable to those measuring reagents
and to those analyzing xray diffraction data,
for just a couple of ways it might come close to some of the expertise
in this group.

(Not so much my expertise; I'm not even doing TDR these days,
and holding light sources up to a piano-sized spectrometer
or running a Kelvin probe are even more distant.
I do pay attention to free file space and to TOP's load measurements)

/dps
Quinn C
2020-01-15 04:52:14 UTC
Permalink
Post by s***@gmail.com
[relevant quotes not elided]
Post by Peter Moylan
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Post by Athel Cornish-Bowden
Post by J. J. Lodder
Some sources do give a Roman mile of 1 482 meter.
I don't think the 'true' Roman mile is known to that accuracy,
or that the Romans could reproduce it to that accuracy,
While we're being pedantic, I think you mean precision.
<outpedanting> I think you are wrong about that.
"In measurement of a set, accuracy refers to closeness of the
measurements to a specific value, while precision refers to the
closeness of the measurements to each other."
Farbeit from me to argue with Wikipêia, but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Does anyone agree, or am I just spouting nonsense that not even "Spains
Harden" would be guilty of?
I agree with you. I'm not sure where that Wikipedia definition came
from, but it doesn't agree with what I understand the normal definition
to be. Perhaps it's used in some specialised field. Statistics, perhaps?
Nope, "The field of statistics, where the interpretation of measurements plays a central role, prefers to use the terms bias and variability instead of accuracy and precision: bias is the amount of inaccuracy and variability is the amount of imprecision. "
You should think instead of metrology. The field of measurements.
<quote>
More commonly, it is a description of systematic errors,
a measure of statistical bias; low accuracy causes
a difference between a result and a "true" value.
ISO calls this trueness.
Alternatively, ISO defines accuracy as describing a combination of
both types of observational error above (random and systematic),
so high accuracy requires both high precision and high trueness.
Precision is a description of random errors, a measure of statistical variability.
<URL:https://en.wikipedia.org/wiki/Accuracy_and_precision#ISO_definition_(ISO_5725)>
Related but somewhat different again, in information retrieval,
precision is a measurement of how reliable extracted results are, i.e.
how likely it is that a retrieved result is a wanted one. In this
context, the sister term is recall, how likely it is that a wanted
result is retrieved.

These are numbers I have to estimate(!) all the time at work; e.g.
these days, for the identification of street addresses in German texts.
--
For spirits when they please
Can either sex assume, or both; so soft
And uncompounded is their essence pure,
-- Milton, Paradise Lost
Peter T. Daniels
2020-01-15 14:17:41 UTC
Permalink
Post by s***@gmail.com
Nope, "The field of statistics, where the interpretation of measurements plays a central role, prefers to use the terms bias and variability instead of accuracy and precision: bias is the amount of inaccuracy and variability is the amount of imprecision. "
How is that "not using" the terms, if their negations are exactly the
definitions of the words they "prefer"?

Does going into statistics preclude statisticians from using English
comprehensibly?

(Does that mean my inability to retain the chi-square concept five
minutes after it's explained to me isn't my fault, but theirs?)

This may be related. In last night's debate, Sanders said "It's
incomprehensible that I would say a woman couldn't win the presidency."
(That may be true, but) He meant "inconceivable."
s***@gmail.com
2020-01-16 22:09:59 UTC
Permalink
Post by Peter T. Daniels
Post by s***@gmail.com
Nope, "The field of statistics, where the interpretation of measurements plays a central role, prefers to use the terms bias and variability instead of accuracy and precision: bias is the amount of inaccuracy and variability is the amount of imprecision. "
How is that "not using" the terms, if their negations are exactly the
definitions of the words they "prefer"?
That's a deeper philosophical question than I was answering
(see PM's post).
Post by Peter T. Daniels
Does going into statistics preclude statisticians from using English
comprehensibly?
I think it happens before specialization.
Post by Peter T. Daniels
(Does that mean my inability to retain the chi-square concept five
minutes after it's explained to me isn't my fault, but theirs?)
I'm sure there's plenty of blame to go around.
Post by Peter T. Daniels
This may be related. In last night's debate, Sanders said "It's
incomprehensible that I would say a woman couldn't win the presidency."
(That may be true, but) He meant "inconceivable."
At least he stayed grammatical.

/dps
Mark Brader
2020-01-15 08:32:37 UTC
Permalink
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Post by Athel Cornish-Bowden
from me to argue with Wikipêia
Quê?
Post by Athel Cornish-Bowden
but the way I learned it was
"A Roman mile was about one and a half kilometres" is accurate but not
very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at all
accurate.
Exactly. And examples like the last one can be called "false precision".
--
Mark Brader "Hacking for 8 years gives a guy a memory.
Toronto If you was with a woman -- I'd've noticed."
***@vex.net PHANTOM LADY
Peter T. Daniels
2020-01-15 14:23:17 UTC
Permalink
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Looks more like Yiddish, where far- is the transliteration of the prefix
corresponding to German ver-.
s***@gmail.com
2020-01-16 22:13:49 UTC
Permalink
Post by Peter T. Daniels
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Looks more like Yiddish, where far- is the transliteration of the prefix
corresponding to German ver-.
Fahrvergnugen.

/dps " or far out, man!"
Peter T. Daniels
2020-01-16 22:26:21 UTC
Permalink
Post by s***@gmail.com
Post by Peter T. Daniels
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Looks more like Yiddish, where far- is the transliteration of the prefix
corresponding to German ver-.
Fahrvergnugen.
/dps " or far out, man!"
I just checked Weinreich's dictionary -- the fahr- words are indeed
spelled the same as the ver- words (<far->). This is the Totally
Official YIVO orthography, introduced in 1936. In the later 19th
century and after, there was a Germanizing orthography of Yiddish
that would have inserted a <h> after the <a>.
Peter Moylan
2020-01-16 00:37:08 UTC
Permalink
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Post by Athel Cornish-Bowden
from me to argue with Wikipêia
Quê?
Post by Athel Cornish-Bowden
but the way I learned it was illustrated by the following sort of
"A Roman mile was about one and a half kilometres" is accurate but
not very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at
all accurate.
Exactly. And examples like the last one can be called "false
precision".
Another common term for that is "delusions of accuracy". A common
student error, especially among those who use a calculator as a
substitute for thinking.
--
Peter Moylan http://www.pmoylan.org
Newcastle, NSW, Australia
J. J. Lodder
2020-01-16 11:27:47 UTC
Permalink
Post by Peter Moylan
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Post by Athel Cornish-Bowden
from me to argue with Wikipêia
Quê?
Post by Athel Cornish-Bowden
but the way I learned it was illustrated by the following sort of
"A Roman mile was about one and a half kilometres" is accurate but
not very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at
all accurate.
Exactly. And examples like the last one can be called "false precision".
Another common term for that is "delusions of accuracy". A common
student error, especially among those who use a calculator as a
substitute for thinking.
In the olden days they did 'precision arithmetic'.
Put a bar over the last significant digit,
and carry it through in the arithmetic
with the standard rules for precision.

I have seen some computer programs that are also capable of it,
but for students it seems to be a lost art,

Jan
David Kleinecke
2020-01-16 19:16:15 UTC
Permalink
Post by J. J. Lodder
Post by Peter Moylan
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Post by Athel Cornish-Bowden
from me to argue with Wikipêia
Quê?
Post by Athel Cornish-Bowden
but the way I learned it was illustrated by the following sort of
"A Roman mile was about one and a half kilometres" is accurate but
not very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at
all accurate.
Exactly. And examples like the last one can be called "false precision".
Another common term for that is "delusions of accuracy". A common
student error, especially among those who use a calculator as a
substitute for thinking.
In the olden days they did 'precision arithmetic'.
Put a bar over the last significant digit,
and carry it through in the arithmetic
with the standard rules for precision.
I have seen some computer programs that are also capable of it,
but for students it seems to be a lost art,
Jan
Back in the old days computers were small and slow and there
was good reason to avoid it but nowadays IMO all computation
should be done in what I know of as "interval arithmetic". One
does not, in this arithmetic, calculate a number. Rather one
calculates upper and lower bounds on the number.

Maybe people are already doing that. I am out of touch.
David Kleinecke
2020-01-16 21:30:08 UTC
Permalink
Post by David Kleinecke
Post by J. J. Lodder
Post by Peter Moylan
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Post by Athel Cornish-Bowden
from me to argue with Wikipêia
Quê?
Post by Athel Cornish-Bowden
but the way I learned it was illustrated by the following sort of
"A Roman mile was about one and a half kilometres" is accurate but
not very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at
all accurate.
Exactly. And examples like the last one can be called "false precision".
Another common term for that is "delusions of accuracy". A common
student error, especially among those who use a calculator as a
substitute for thinking.
In the olden days they did 'precision arithmetic'.
Put a bar over the last significant digit,
and carry it through in the arithmetic
with the standard rules for precision.
I have seen some computer programs that are also capable of it,
but for students it seems to be a lost art,
Jan
Back in the old days computers were small and slow and there
was good reason to avoid it but nowadays IMO all computation
should be done in what I know of as "interval arithmetic".
Au contraire, precision arithmetic was much used by human computers
before the days of computers.
It saves doing superfluous work.
Post by David Kleinecke
One does not, in this arithmetic, calculate a number. Rather one
calculates upper and lower bounds on the number.
That is a cruder way of doing the same thing,
It is the same thing - how otherwise do you calculate the interval?

I strongly doubt that "precision arithmetic" was much used
by human computers. I never worked as a computer but my mother
did the calculating for the fusion project at Livermore Lab and
she never heard of the idea. The physicists and engineers I
discussed the idea with back then had difficulty grasping the
idea and wondered why plus-or-minus wouldn't be just as good
(results can be presented that way after an adjustment but the
computed interval is almost never symmetrical around the computed
single value.)

My attention was called to this situation by attempts to calculate
smog that failed because the round-off errors would drive the
amount of small pollutants below zero on occasion (an easier
solution to that problem is don't calculate the amount of
pollutants - calculate their logs.)
J. J. Lodder
2020-01-17 13:00:49 UTC
Permalink
Post by David Kleinecke
Post by David Kleinecke
Post by J. J. Lodder
Post by Peter Moylan
Post by Mark Brader
Post by Athel Cornish-Bowden
Farbeit
I first read that as being some German word that I didn't know.
Post by Athel Cornish-Bowden
from me to argue with Wikipêia
Quê?
Post by Athel Cornish-Bowden
but the way I learned it was illustrated by the following sort of
"A Roman mile was about one and a half kilometres" is accurate but
not very precise.
"A Roman mile was 1.3895319 kilometres" is very precise but not at
all accurate.
Exactly. And examples like the last one can be called "false
precision".
Another common term for that is "delusions of accuracy". A common
student error, especially among those who use a calculator as a
substitute for thinking.
In the olden days they did 'precision arithmetic'.
Put a bar over the last significant digit,
and carry it through in the arithmetic
with the standard rules for precision.
I have seen some computer programs that are also capable of it,
but for students it seems to be a lost art,
Jan
Back in the old days computers were small and slow and there
was good reason to avoid it but nowadays IMO all computation
should be done in what I know of as "interval arithmetic".
Au contraire, precision arithmetic was much used by human computers
before the days of computers.
It saves doing superfluous work.
Post by David Kleinecke
One does not, in this arithmetic, calculate a number. Rather one
calculates upper and lower bounds on the number.
That is a cruder way of doing the same thing,
It is the same thing - how otherwise do you calculate the interval?
There is no need to calculate intervals. Just carry a few decimals more.
Post by David Kleinecke
I strongly doubt that "precision arithmetic" was much used
by human computers.
Then you are unaware of the way professional human calculators worked.
Doing things like predicting solar eclipse for exanmple,
or planetary positions, with sine tables and hand-cranked calculators.
You stop the calculation when you are no longer adding significance.
Post by David Kleinecke
I never worked as a computer but my mother
did the calculating for the fusion project at Livermore Lab and
she never heard of the idea. The physicists and engineers I
discussed the idea with back then had difficulty grasping the
idea and wondered why plus-or-minus wouldn't be just as good
(results can be presented that way after an adjustment but the
computed interval is almost never symmetrical around the computed
single value.)
Indeed, much calculation is done by incompetents,
then and now, and there are many nonsense results around.
Wheeler formulated his 'First Moral Principle' with good reason.

Jan
--
Wheeler's First Moral Principle: Never make a calculation until you know
the answer.
Lewis
2020-01-13 03:25:26 UTC
Permalink
Post by Jack
A crossword puzzle has the clue: Distance equalling 1,482 meters.
The answer is 'mile'.
Does anyone know what kind of mile might have that size?
Ignorant mile?
--
"Computers are useless. They can only give you answers." - Pablo
Picasso
Loading...