HP 42s Value Display

[cylurian]

On the DM42 calculator, when I type 5/10, it gives me 0.5, which is great. The display is set to "ALL." However, when I type 12/148, I get 8.108108×10^-3. If I use the "Shift" key and select "FIX 9," it shows 0.081081081, which is what I want. But then, if I do something like 12 + 12, I get 24.000000000, which is also correct since it's showing nine digits after the decimal. Is there a way to get 0.081081081 without having to use the FIX setting every time?

[Steve Simpkin]

I do not believe so. ALL display mode appears to switch to scientific mode automatically if there are more digits beyond the 12 digits display capacity. For example, in ALL mode, you can enter .33333333333 (. followed by eleven 3's) and it will display 0.33333333333 on the screen. If you take the reciprocal of 3 however, you will see 3.33333333333E-1 on the screen. I always found this odd and inconvenient.

   


Note the HP 35s is worse in this respect. 1/3 will show 3.33333333333E-. You have to use the right cursor key to see the 1 at the far right.

   

[dm319]

I also agree that there could be another mode here. A mode that shows you 'up to' a certain number of 'interesting' (i.e. not trailing zero) digits.

It's unlikely to happen as free42 is built as a faithful recreation of the HP-42s and this might mess with it, but maybe not, and to be fair NSTK mode would be a bigger change.

C47 has a feature called 'Sig', but I'm not sure if that includes trailing zeros or not.

[dm319]

It would be nice if we could set a range where scientific notation is not used. I.e. an exponent between -3 and +3 would default to regular notation.

[AnnoyedOne]

My 2c.

Most HP calculators, and their clones, especially scientific models are intended to be used by engineers and/or scientists who have no problem with 42 displaying as 42.0000 or whatever. Or scientific/engineering notation.

If one has a problem with this perhaps TI, or dare I say it Casio, would be better brands to use.

A1

[naddy]

(10-22-2024 02:16 PM)AnnoyedOne Wrote:  Most HP calculators, and their clones, especially scientific models are intended to be used by engineers and/or scientists who have no problem with 42 displaying as 42.0000 or whatever.

Right. When you are dealing with physical systems as opposed to purely mathematical objects, numbers represent measurements that always have limited accuracy, so you never have 42, you have e.g. 42.000 +/- 0.005, and you can set the display accuracy accordingly. When those values come from measuring tape and multimeter, a total of three significant digits is already aspirational.

[Steve Simpkin]

(10-21-2024 11:37 PM)Steve Simpkin Wrote:  I do not believe so. ALL display mode appears to switch to scientific mode automatically if there are more digits beyond the 12 digits display capacity. For example, in ALL mode, you can enter .33333333333 (. followed by eleven 3's) and it will display 0.33333333333 on the screen. If you take the reciprocal of 3 however, you will see 3.33333333333E-1 on the screen. I always found this odd and inconvenient.

Note the HP 35s is worse in this respect. 1/3 will show 3.33333333333E-. You have to use the right cursor key to see the 1 at the far right.

This unique display behavior seems to have started with the Pioneer series and "ALL" display mode. The HP-28/48/49/50 series and HP Prime do not do this. They have an equivalent display mode called "Standard" (or STD). In that mode, if you take the reciprocal of 3 on any of those models, you will see .333333333333 on the screen.

[AnnoyedOne]

FYI, my HP-20S (a Pioneer model) manual

HP-20S Owner's Manual (00020-90001, Edition 5, 8-92)

states on p17 (19 in the PDF)

Quote:To set your calculator to display numbers as precisely as
possible, press [>][ALL]. Trailing zeros are not displayed.

I've never used it and probably never will (now).

A1

[dm319]

(10-22-2024 02:16 PM)AnnoyedOne Wrote:  If one has a problem with this perhaps TI, or dare I say it Casio, would be better brands to use.

This is, funny enough, my second burn today.

[Thomas Okken]

The point of ALL mode on the 42S is to show all 12 digits of the mantissa, without trailing zeros. It switches to SCI mode when that is necessary to show all 12 digits, which is why you'll get SCI mode when you do 3 1/X, but not for, say, 8 1/X or 9.9 1/X.

Note that the way this works is slightly different than how it works in the 48 series, because the 42S won't suppress the integer part if it is zero, while the 48 series will. So, 3 1/X is displayed like 3.33333333333E-1 on the 42S, and as .333333333333 on the 48. On the 42S, the zero before the decimal would only leave room for 11 fractional digits, so it has to switch to SCI mode. In my 9.9 1/X example, the switch to SCI mode is not necessary, because the least significant digit happens to be zero in that case, so nothing is lost by not having room to display it.

[cylurian]

True, you can place it in SCI mode, but then when you type a 2, you get 2.000000000000E0, talk about significant digits! I'm so spoiled with those TI calculators!

[rprosperi]

(10-24-2024 10:08 PM)cylurian Wrote:  True, you can place it in SCI mode, but then when you type a 2, you get 2.000000000000E0, talk about significant digits! I'm so spoiled with those TI calculators!

Yeah, no need to focus on those meaningless digits, no need to think. "Trust us... we know what answers you want to see... you can trust TI...".

If they had the intestinal fortitude to actually document how they changed results to make them "easier to understand" I would have respected TI a lot more. I mean folks that don't care about such details likely weren't reading all the manuals, whereas folks that did care, and could understand, were denied the chance. To this day, I'd guess (but I've spent literally no time checking) that these details remain undocumented.

[AnnoyedOne]

(10-24-2024 11:50 PM)rprosperi Wrote:  Yeah, no need to focus on those meaningless digits, no need to think. "Trust us... we know what answers you want to see... you can trust TI...".

Exactly. I'm ignorant and want to remain so. Don't bother me with details!

A1

[dm319]

The secret behaviour of some calculators to show you what it thinks you wanted to see is infuriating!

The HP-12c Rogue was doing this, and you really had to work to uncover what it didn't want to show you. Turns out Casio and TI do a lot of this as standard. The one making me pull my hair out was subtracting two large and very similar numbers together. It calls it zero to save you the bother.

When the internet archive is back up again you can look at this document on what the TI-83/84 was doing:
http://web.archive.org/web/2017042821103...ecrets.pdf

And this is quite a popular video on a Casio calculator deciding that an answer is actually a rational fraction of Pi:
https://youtu.be/7LKy3lrkTRA?si=5LfTmnr6zDd2MJ-E

All of this makes me think that any fakery in the display of the result of your calculation is a bad thing. I would rather the calculator doesn't try to guess whether a result is a fraction of Pi, or close enough to an irrational number multiple of sqrt(2), or small enough to tell me that what I really want to see is '0'.

In fact, taking a step back, that's why I'm here. Why I think RPN is the best. The calculator doesn't do the interpretation that my brain is capable of. I don't have to argue with my calculator about the 'correct' way to interpret arithmetic formulae - I get to decide.

Calculators should be there to do the bit my brain doesn't do - the actual calculating. But let me do the thinking.

[Steve Simpkin]

The above goes with the HP traditional philosophy of showing you the computed answer based on the number of significant digits stored by the hardware without using invisible "guard-digits" to round off the answer as TI and other calculator manufactures typically do.

There is some interesting history behind the decision on what to show the calculator user when the true result using fixed digit math is not what they expect to see. William Kahan had the following to say about the apparent "increased accuracy" of TI models when he worked as a consultant for HP.

"Hewlett-Packard had come out with a beautifully engineered job called the HP-35, which was the first scientific calculator with all the scientific functions instead of just the add, subtract, multiply, divide, and maybe a square root. And then they came out with the HP-45, which was an improved version. It had more functionality. But in the meantime, Texas Instruments came out with a calculator that was a great deal cheaper, and here’s how they advertised their calculator. So TI had this advertisement in the papers. It was a full-page advertisement. It said, “Type in your telephone number. Now,” they said, “Take the logarithm.” The logarithm turns out to be a number form ten-point-something, or nine-point something, actually. “Now hit the exponential key. Do you get your phone number back? You do on our calculator.” HP knew that it was the target of this advertisement because it did that on an HP-45, which carried ten digits. You type in your ten-digit phone number, take the log, take the exponential, and the last digit or two would change but, apparently, not on the TI calculator. HP was very worried about this, because it seemed to impugn the integrity of their beast."

This and what he recommended HP do is discussed starting around page 144 on the following interview with William Kahan - August 2005
https://drive.google.com/file/d/1Jlg9EWQ...zwcol/edit

[dm319]

(10-25-2024 11:45 AM)Steve Simpkin Wrote:  This and what he recommended HP do is discussed starting around page 144 on the following interview with William Kahan - August 2005
https://drive.google.com/file/d/1Jlg9EWQ...zwcol/edit

That's brilliant. The difference between making something really good and making something look really good.

[AnnoyedOne]

(10-25-2024 11:45 AM)Steve Simpkin Wrote:  This and what he recommended HP do is discussed starting around page 144 on the following interview with William Kahan - August 2005...

From the Kahan interview. P144 of the PDF.

Quote:But in the meantime, Texas Instruments came out with a calculator that was a great deal cheaper, and here's how they advertised their calculator. So TI had this advertisement in the papers. It was a full-page advertisement. It said, "Type in your telephone number. Now,” they said, “Take the logarithm.” The logarithm turns out to be a number form ten-point-something, or nine-point-something, actually. “Now hit the exponential key. Do you get your phone number back? You do on our calculator."

and P146

Quote:After you've done it seven times, your telephone number changes. Do you feel that that's honest? Is this an honest ad?

This, boys and girls, is known as Marketing. Bending the truth or outright lying to sell your stuff. HP used to be an engineering company and didn't do this often.

A1

[AnnoyedOne]

(10-25-2024 11:45 AM)Steve Simpkin Wrote:  ...interview with William Kahan - August 2005...

Thanks sooo much for posting that link Steve. You have no idea how much I'm enjoying reading it! And Kahan didn't have much respect for marketing people either. His stories are similar to my experiences

A1

[Jase]

(10-25-2024 11:45 AM)Steve Simpkin Wrote:  This and what he recommended HP do is discussed starting around page 144 on the following interview with William Kahan - August 2005
https://drive.google.com/file/d/1Jlg9EWQ...zwcol/edit

I've been reading for 50 minutes now, with more to go. This is fascinating.
Thank you.
Jase.

[Albert Chan]

Kahan Wrote:After you've done it seven times, your telephone number changes. Do you feel that that's honest? Is this an honest ad?

I think TI's ad is honest, showing usefulness of exposed excess precision.

exp(ln(x) ± ε) = x * exp(±ε) ≈ x * (1±ε)

x is 10-digits telephone number --> ln(x) = ln(1e9 .. 1e10) ≈ 20.7 .. 23.0

HP with 10 digits precision, max(ε) = 50 ULP, where ULP = 1e-10

Example, in HP12C

9985550000
LN   --> 23.02440488
E^X --> 9985549951

Compare this to TI, e^ln(e^ln(e^ln(e^ln(e^ln(e^ln(e^ln(x))))))) only move at most 1 ULP!
(If it did move more than 1 ULP, Kahan would not have said 7 times)