Special Function Sets & HPs

[Matt Agajanian]

Hi all.

I was wondering why HP scientifics did not include extended function sets/constants similar to TI's SR-51 & 51A for example. Why were HP scientific calculators included only the slide-rule and/or statistical functions?

[Gene]

Hi Matt! Welcome back.

The function set of the SR-51/51A was more advanced than the same timeframe HP non-programmables, but that in many ways was a function (ha) of the cheaper memory available to TI compared to HP. TI made memory. HP did not.

That said, HP was much more advanced in programmables in that early 1974-1975 timeframe than TI, so perhaps it is a mixed comparison anyway.

For specifics, the SR-51/51A had these functions most HP's did not in 1974-75:
Hyperbolics
Linear regression
Random numbers
Permutations
20 conversions

Most of that was rectified by the time of the HP 32E in 1978.

[Matt Agajanian]

Good point--HP reinforced its functionality with programming AND, if may infer (from the Classic and Woodstock manuals) packed a heavy set of applications in the manuals. So, that seems a robust compromise.

And I'll say, that 32E is one of my favourite pieces in my ensemble.

[TASP]

The young whippersnappers these days don't realize how rare and precious memory was in the 70s.

It was even worse when I was a kid:

stick = 1

rock = 0


took for ever to code anything

[Paul Dale]

Quote:Hyperbolics

Seriously, does anyone actually use these? I have never required them for anything beyond minimal step challenges. Honestly, I'd prefer to include the Jacobi elliptic functions which are a superset of both hyperbolic and trigonometric functions (these didn't make the cut in the 34S).

Quote:Linear regression

Yes, very useful. Better in later calculators with extra regression modes.

Quote:Random numbers

Given that all calculator random number generators that I'm aware of are pseudos, why? Certainly not a distinguishing feature -- seriously, how many times are you going to press this in a row outside of a programme (where you can include your own)? I'd prefer the extra register on the 11C and 15C e.g.

Quote:Permutations

Yes, useful even if you already have factorial.

Quote:20 conversions

These are useful in two countries in the world these days: USA and Liberia. The third non-metric nation being Myanmar, but they don't use English Imperial measurements. Even back in the day, almost all countries had already converted to metric.


- Pauli

[d b]

Yes Pauli. I second all your points. IMHO; having a calculator with everything but the kitchen sink is desireable UNLESS that takes away from user program memory. Then the ability to loose large un-needed (to me) functions would be great.
As you mentioned; imperial <> metric conversions take space, and they could be memorized by the few of us that need them, if that gains RAM.

[Gene]

Ah, but they were not all metric conversions.

Polar to rectangular

Degrees to DMS

Voltage ratio to decibels

Degrees to radians and grads

Acres to square feet

Miles to Nautical miles

So not just the inches to centimeters stuff :-)

[ttw]

The conversions from various systems matters if one does any work with historical literature. Even metric measures change, calories are passé in thermodynamics (as are Amagats).

It's also important for surveying. Lots of property lines were established (in the US) when the yard was 3600/3937 meter as opposed to the present 36000/39372 meter. The various types of varas are amusing too, 33 inches, 33 and 1/3 inches, etc.

[rprosperi]

(01-31-2016 01:56 PM)Gene Wrote:  Ah, but they were not all metric conversions.
...
Voltage ratio to decibels
....

This is cool and I suspect unique in a calculator, isn't it?

SR-51A was my last pre-HP calculator (I stepped up to an HP-21, but did miss some of these very functions) and I recently got a very nice one, complete with box, manuals, the works for almost nothing (compared to any vintage HP). Fun to reminisce, even with a TI.

[d b]

If you meet a surveyor who needs a dedicated key to devide by 43560, take away his Gunters chain.

TTW: If only all the Vara were that close.....
And what about "smokes"?
Lets not even go into your neighbors and their Arpents.

[emece67]

(01-31-2016 08:29 AM)Paul Dale Wrote:  

Quote:Hyperbolics

Seriously, does anyone actually use these? I have never required them for anything beyond minimal step challenges. Honestly, I'd prefer to include the Jacobi elliptic functions which are a superset of both hyperbolic and trigonometric functions (these didn't make the cut in the 34S).

[...]

Well, as you have asked... I suppose this (and all others) depend on the field each one uses to work. When studying EE I used hyperbolics in two situations (lossy transmission lines and bipolar differential amplifiers). And later, when at work, when dealing with exponential horns. But, instead, I've never seen (this does not mean that they do not exist) an engineering book that uses the Jacobi elliptic functions, but have seen many others using the Bessel ones (one of the things I have missed in my calculators).

(01-31-2016 10:07 AM)Den Belillo (Martinez Ca.) Wrote:  Yes Pauli. I second all your points. IMHO; having a calculator with everything but the kitchen sink is desireable UNLESS that takes away from user program memory. Then the ability to loose large un-needed (to me) functions would be great.
As you mentioned; imperial <> metric conversions take space, and they could be memorized by the few of us that need them, if that gains RAM.

I use the metric system, but many times I need to convert to/fro imperial units. Not being a "native imperial user", I only know, from memory, what feet and inches are, so I currently rely on a wp34s or Emu48 for other conversions. Incidentally, I much appreciate the mpg<->l/100 km conversion on the wp34s!

Surely HP/TI took many compromises when dealing with the scarcity of memory, processor, display and others, apparently both of them took right enough decisions, seeing their success. For me, today, the bottleneck is the interface, so such hypothetical all-included calculator is desirable unless it has a cumbersome interface requiring long key sequences to arrive to the functions I want to use or a cryptic display that turns entering/reading data a nightmare. If not this way, I would prefer to use Matlab/Mathematica on a PC.

[emece67]

(01-31-2016 08:29 AM)Paul Dale Wrote:  

Quote:Random numbers

Given that all calculator random number generators that I'm aware of are pseudos, why? Certainly not a distinguishing feature -- seriously, how many times are you going to press this in a row outside of a programme (where you can include your own)? I'd prefer the extra register on the 11C and 15C e.g.

Incidentally, the processor inside the last incarnation of the 17BII+, an Atmel ATSAM4LC2, does contain a true random number generator. I do not know if the 17BII+ has a RAN# command, much less if it uses such generator.

[Ángel Martin]

(01-31-2016 08:26 PM)emece67 Wrote:  

(01-31-2016 08:29 AM)Paul Dale Wrote:  Seriously, does anyone actually use these? I have never required them for anything beyond minimal step challenges. Honestly, I'd prefer to include the Jacobi elliptic functions which are a superset of both hyperbolic and trigonometric functions (these didn't make the cut in the 34S).

[...]

Well, as you have asked... I suppose this (and all others) depend on the field each one uses to work. When studying EE I used hyperbolics in two situations (lossy transmission lines and bipolar differential amplifiers). And later, when at work, when dealing with exponential horns. But, instead, I've never seen (this does not mean that they do not exist) an engineering book that uses the Jacobi elliptic functions, but have seen many others using the Bessel ones (one of the things I have missed in my calculators).

Both the Jacobi Elliptic functions and the Bessel functions are available in the SandMath - on the 41, that is if you would find a mere 10-digit approach acceptable of course.

BTW another relatively common application of the bare-bones hyperbolic arises in transmission power lines, as the equation of the catenary involves the hyperbolic cosine..

PS. as far as random numbers, involving the time module makes it less pseudo and more like the real thing, does it not??

[Paul Dale]

(01-31-2016 09:32 PM)Ángel Martin Wrote:  PS. as far as random numbers, involving the time module makes it less pseudo and more like the real thing, does it not??

There is some entropy from the jitter reading the time source but I wouldn't expect much -- it is ticking seconds isn't it? Still, better than none.


- Pauli

[Matt Agajanian]

(01-31-2016 01:55 AM)Gene Wrote:  Hi Matt! Welcome back.

The function set of the SR-51/51A was more advanced than the same timeframe HP non-programmables, but that in many ways was a function (ha) of the cheaper memory available to TI compared to HP. TI made memory. HP did not.

That said, HP was much more advanced in programmables in that early 1974-1975 timeframe than TI, so perhaps it is a mixed comparison anyway.

For specifics, the SR-51/51A had these functions most HP's did not in 1974-75:
Hyperbolics
Linear regression
Random numbers
Permutations
20 conversions

Most of that was rectified by the time of the HP 32E in 1978.

Thanks Gene! It feels good to be back.

[Jeff_Kearns]

(01-31-2016 08:29 AM)Paul Dale Wrote:  

Quote:Hyperbolics

Seriously, does anyone actually use these? I have never required them for anything beyond minimal step challenges. Honestly, I'd prefer to include the Jacobi elliptic functions which are a superset of both hyperbolic and trigonometric functions (these didn't make the cut in the 34S).

- Pauli

This comment made me think of one of Valentin's replies back in 2003:
Hyperbolics everywhere ! (almost)
Message #4 Posted by Valentin Albillo on 30 June 2003, 6:38 a.m.,
in response to message #3 by Ernie Malaga
Ernesto wrote:

"How about hyperbolics? So far I haven't learned of a _single_ use for hyperbolic functions in any field."

They're used all the time in some very important Civil Engineering fields, a few assorted examples:

electrical transportation (trains and such, which draw power from a suspended conducting wire). The wire's shape is a curve called catenary, which is a hyperbolic cosine, and thus you need hyperbolics to compute lengths, weights, stress, costs, etc. This also applies to high-voltage power lines and, in general, to any kind of chain, rope or wire suspended from two points that hangs freely under its own weight.
superstructure engineering (suspension bridges and such), where you need to compute the elastic curve and the deflection, also require hyperbolic cosines in spades.
architecture, specially traditional japanese architecture and civil engineering, where curves have various important social and philosophical meanings (temples and other relevant buildings). Japanese and oriental architects frequently use the catenary shape (hyperbolic cosine) for the upward-pointing curves in their constructions instead of the less-quickly-raising parabolas.
aerospatial engineering: engineers (from Boeing, for example) face a problem in applying a surface coating to fighter aircraft. The critical feature is the electrical conductivity at the surface. This physical property determines how an electromagnetic wave will scatter when it hits the aircraft. For some fighter parts, the ideal coating varies along the surface according to the hyperbolic cosine function.
etc, etc. Matter of fact, students of the Civil & Environmental Engineering divisions of most universities are required to be confident with the manipulation of hyperbolic functions and the solution of hyperbolic equations. I guess such an student would find himself as much at a loss using a calculator without hyperbolics as you and me would be using one without division, say.

"There are more things in heaven and earth, Horatio, Than are dreamt of in your philosophy."

--From Hamlet, Prince of Denmark - 1601 - Act I. - Scene 5. - Rows: 166-167

:-)

Best regards.

Edited: 30 June 2003, 6:44 a.m.

Jeff Kearns

[Matt Agajanian]

Jeff,

Your examples are insightful, amazing and quite observant! Yes, these natural, everyday cites although overlooked and taken for granted are quite honestly, for me at least, the most extraordinary and fascinating examples of hyperbolics in implementation!

[emece67]

(02-01-2016 07:45 PM)Jeff_Kearns Wrote:  This comment made me think of one of Valentin's replies back in 2003:
Hyperbolics everywhere ! (almost)
Message #4 Posted by Valentin Albillo on 30 June 2003, 6:38 a.m.,
in response to message #3 by Ernie Malaga
Ernesto wrote:

"How about hyperbolics? So far I haven't learned of a _single_ use for hyperbolic functions in any field."

They're used all the time in some very important Civil Engineering fields, a few assorted examples:

[...]

Thanks Jeff for such interesting insights.

There's a thing that surprises me. In your examples, the revelant hyperbolic function seems to always be the hyperbolic cosine. In my three examples, instead, it is the hyperbolic tangent.

Does anybody know of any field where the relevant hyperbolic function is the hyperbolic sine?

Edit: I have found just now that the hyperbolic functions do also play a role in the Mercator projection. The expressions relating latitude and the y coordinate in a map can be written in terms of the hyperbolic functions, and even using the Gudermannian function.