What are the most popular functions? (Poll until 18.12.)



#36

Imagine the lower six rows of keys of our hypothetical eight-row calculator are filled with functions not listed below. Two rows are left. The following individual functions compete for the remaining space:

1/x SQRT LN e^x x^2 y^x LG 10^x i [theta] [Sigma+] [Sigma-] P>R R>P D>R R>D HR>H.MS H.MS>HR sin cos tan asin acos atan x_bar s y_hat r [parallel] % %CHG x_ROOT(y) LOG[sub_x](y) RAN# H.MS+ H.MS- x! COMB PERM ROUND DISP USER ALPHA DSE ISG ABS IP FP

What shall go there? You may specify up to 12 functions in your vote (one post only, please). You may specify them in pairs (unshifted / shifted) if you want to. You must not specify menus.

This poll will terminate on Tuesday, December 18th, at 12:12 UTC. Please participate.

d:-)


#37

unshifted: 1/x y^x sqrt ln sin cos tan

shifted: e^x [parallel] asin acos atan

#38

Ans, comma, ( , ) , and =. Nothing else matters.

TW

Edited: 14 Dec 2012, 10:08 a.m.


#39

Sorry, Tim, you obviously didn't read the OP thoroughly :-/

#40

P>R, R>P HMS+ HMS- IP FP


SIN COS TAN ASIN ACOS ATAN


most used in navigation formulae.

Thanks, Geoff


#41

Quote:
P>R, R>P HMS+ HMS- IP FP

SIN COS TAN ASIN ACOS ATAN

most used in navigation formulae.


+1 (or in plain text: the same for me please ) :-)

max


Edited: 14 Dec 2012, 4:26 p.m.

#42

P>R,R>P,SQRT,x^2,%,%CHG
SIN,COS,TAN,ASIN,ACOS,ATAN

#43

Shifted: [Sigma-] asin acos atan x^2 y^x
Unshifted: [Sigma+] sin cos tan SQRT 1/x

#44

n, i, PMT, PV, FV, %, Sigma+

... All of a sudden, I feel lonely...

Nevermind, I love your project :-)

Etienne


#45

Quote:

n, i, PMT, PV, FV, %, Sigma+

... All of a sudden, I feel lonely...

Nevermind, I love your project :-)

Etienne


You must be kidding, what functions are 'n' and 'i'? What is so important about the super-complicated function '%'? What's more we are discussing here scientific calculator, not a new version of HP12B.

#46

Kidding I was, young padawan.

N, I and % indeed to The Dark Side of The Force belong...

The great RPN Jedis, The Beancounter will fight.

HP-80 rules !

#47

unshifted: sin  cos  tan  LN  LG   x_ROOT(y)
shifted: asin acos atan e^x 1/x y^x
#48

IMHO the 35s has a quite useful keyboard layout. I'd like to have something similar to its third row of keys. However, the 35s has two shift keys. If there is just one, I would say this is the most practical layout for general use:

  shifted: arcsin arccos arctan  x^2    e^x    y^x
unshifted: sin cos tan sqrt ln 1/x
As to the six other functions: Phi(x) and Phi^-1(x) are used in most social sciences, so they are IMHO a "must" on a scientific calculator. While the parallel function can easily be replaced by a short user program (or a few simple keystrokes), this is not the case for the Normal CDF and its inverse. My suggestion for the second row, assuming "disp" sets the display mode and x! is actually Gamma(x+1):
  shifted: Phi^-1   x!    xroot  10^x   Round  User
unshifted: Phi % log_x lg Disp Alpha
No, I do not need any complex number support or rectangular/polar conversion functions.

By the way: what about an INV key?

Dieter

#49

Quote:
Imagine the lower six rows of keys of our hypothetical eight-row calculator are filled with functions not listed below. Two rows are left. The following individual functions compete for the remaining space:

1/x SQRT LN e^x x^2 y^x LG 10^x i [theta] [Sigma+] [Sigma-] P>R R>P D>R R>D HR>H.MS H.MS>HR sin cos tan asin acos atan x_bar s y_hat r [parallel] % %CHG x_ROOT(y) LOG[sub_x](y) RAN# H.MS+ H.MS- x! COMB PERM ROUND DISP USER ALPHA DSE ISG ABS IP FP

What shall go there? You may specify up to 12 functions in your vote (one post only, please). You may specify them in pairs (unshifted / shifted) if you want to. You must not specify menus.

This poll will terminate on Tuesday, December 18th, at 12:12 UTC. Please participate.

d:-)


Unshifted: SQRT, SIN, COS, TAN, LN, >RECT


Shifted: x^2, ASIN, ACOS, ATAN, e^x, >POL

Edited: 14 Dec 2012, 3:45 p.m.

#50

Unshifted: sin cos tan sqrt e^x y^x

Shifted: asin acos atan x^2 ln 1/x

Edited: 14 Dec 2012, 3:53 p.m.

#51

HYP, HYP<superscript>-1, L.R.

#52

Only 12 functions? The premise of this poll seems premature. The fundamental question of whether to load the keyboard top row or dedicate them to menu soft keys is still open, is it not?

It seems to me that the question(s) of keybpard loading are fundamental to the operation (or operability) of the calculator. My preference would be to "load" the keyboard. I would propose blue and yellow functions (in addition to the primary functions) for the second row. The top row would be re-assignable soft keys, but absent a menu call these keys would have assigned primary and yellow-shifted functions.

Loading the top row with a primary and shifted function (in addition to the soft key assignment) and adding a second shifted function to the second row would give us 30 spots (or 32, counting USER/ALPHA on the third row) and would cover most of the above list of potential functions. This is the question that should be answered first: load the keyboard, or not? Then, we debate the 18, 24, 30, or 32 most popular functions and their placement.

In case I'm alone in this, here are my 12 functions:

1. 1/x (I'd actually make this a data entry key along with +/-, EEX, and backspace, but surely >almost< all at least agree that it has to be on the keyboard. I'd make this a primary key.)

2. SQRT (This seems fundamental to me, and I'd like it to be primary. However with only 6 keys, I'd make it the yellow-shift of 1/x.)

3-8. SIN, COS, and TAN with shifted ASIN, ACOS, and ATAN.

9-10. LN with shifted e^x.

11. LG

12. y^x as the shifted function of LG. (I'd prefer 10^x be located on the LG key with y^x as a separate, primary key but real estate dictates.)

With the exception of #1, these are primary functions in that no closed-form calculation exists. With these functions and some stack, we can calculate most anything else that's missing, like H.MS, rect/polar, degrees/radians, TVM, parallel, and IP, FP. Some of these will be bigger losses in terms of convenience/usability than others, and this will vary by person. We >can<, however, get to them without menus. Unfortunately (and I see this as a major compromise) complex numbers, probability, and statistics are completely relegated to menus.

-Bill


#53

I kind of agree with Bill here. There is so much not being asked, and much implied that many seem to have missed.

Since functions like USER and ALPHA are mentioned in the poll, this implies they are not assigned below already. They seem to be crucial to a sophisticated calculator like this is clearly going to be!

Given the layouts shown in other posts, I'm guessing many are assuming those are already present somewhere!?

Here are my choices given the constrains in the poll:

shifted: DISP   i    y^x  x^2   LN  SIGMA-
primary: USER ALPHA 1/x SQRT e^x SIGMA+

If in fact USER, ALPHA, DISP, or i are pre-assigned below as has been hinted at before, I'd try to fit the trig functions, giving up lower priority stuff like x^2, and pushing the SIGMAs to a menu, but as it is I couldn't fit some trigs but not others.

I sincerely hope the basic trig functions, as well as LG and 10^x are also available on the keyboard somehow. Since you only allowed me 12, I made the difficult choice.


#54

The simple reason is: I've to sum up your votes at the end of this poll; and I don't want to add let's say 40 times 50 functions - 40 times 12 is more convenient for me.

Side effect: this limit lets you feel the difficulty we feel all the time - we simply haven't enough real estate for all those nice functions you want to have, so we have to select. Same for you here.

Thanks for your understanding.

d:-)

#55

1/x SQRT x^2 y^x P>R R>P HR>H.MS H.MS>HR H.MS+ H.MS- USER ALPHA

#56

unshifted SQRT LOG  LN   y^x  1/x  sincos
shifted x^2 10^x e^x y^-x x/y PI
Note:sincos calculates sin(x) and cos(x) simultaneously and stores
sin(x) to y and cos(x) to x.

Lyuka


#57

y-x and x/y seem weird. x<>y / is the same as x/y and in two keystrokes still -- okay slightly different LastX behaviour. yx 1/x or +/- yx are the same as y-x again subject to LastX.

Why not x-y as a function (reverse subtract) as well?


At least nobody has suggested xy yet :-) Must maintain HP 35 compatibility.


- Pauli

#58

First thought:

sqrt, 1/x, x2, yx, USER, ALPHA, sigma+, sigma-, y1/x.

Rounds off with LOG, TRIG, CPX.

Unshifted: USER ALPHA, sigma+, LOG, TRIG. Probably CPX or 1/x.


But why not put all that powers into a menu too. We still get two key access to them which is no worse than a shift.

Unshifted: USER, ALPHA, sigma+, POWER, LOG, TRIG.
Shifted: USER1, ALPHA1, sigma-, GAMMAs, HYP, ELLIPT.

Maybe swap out sigma+ for STATS or 1/x. Maybe call POWER PARTS or COMMON, SIMPLE-MATHS-FUNCTIONS.


My all time favourite HP, the 34c has almost all the mathematical functions shifted -- a top level menu and one further press is no different. The WP 34S also shifts the maths functions and we've no had major complaints about that either.


- Pauli


#59

Pauli,

I admit I still don't understand why we should need all those four. Maybe I'm wrong but as far as I got it:

  • USER shall toggle user mode, so the user assignments come to the front.
  • ALPHA shall toggle alpha mode, so the grey letters come to the front.
  • ALPHA1 shall switch to alpha mode for input of a single character only, then return to the mode set before.
  • USER1 shall switch to user mode for input of a single operation only, then return to the mode set before.
Please confirm or correct.

d:-?


#60

I disagree about how the distinction between USER and USER1 (or ALPHA and ALPHA1) occur. I prefer the HP-48 method: USER or USER1 are the same key. The behavior depends on the value of a flag.

The default after reset is that a single press of the USER key sets displays a USER1 annunciator and USER mode is active for a single following key sequence (other than another USER press), after which the USER1 annunciator is turned off, and normal mode is resumed.

If, from normal mode, USER is pressed twice consecutively, the first press sets USER1 as described above, but the second press changes the annunciator to USER (no "1"), and USER mode is "locked" on until the USER key is pressed again later, returning to normal mode.

If a "USER lock" flag is set, then a single press of the USER key enters USER lock, and the next USER press returns to normal.

I'm not really sure that ALPHA1 makes a lot of sense on this calculator, but if it is to be done, it should work similarly to USER1, except that I think the default setting should be the opposite (ALPHA lock enabled), and the user should have to clear the ALPHA lock flag to get ALPHA1 behavior.

USER1 is not very useful if it is a separate, shifted function, as it doesn't save keystrokes over the use of a locking USER key.


#61

Yeah, the 48 method of press or double press is probably better than shifting one of the two flavours. Either way, a flag to toggle which occurs first is sensible I think.


- Pauli


#62

I second that.

#63

Correct for all four.

They are all useful in different ways.


With well thought our menus (including user defined), USER1 might end up redundant. ALPHA1 is probably less useful here than on the RPL machines it is derived -- this will depend on how we address named registers and locals.


- Pauli

#64

Quote:
My all time favourite HP, the 34c has almost all the mathematical functions shifted -- a top level menu and one further press is no different. The WP 34S also shifts the maths functions and we've no had major complaints about that either.

Both (34C and 34s) keyboards have only seven key rows, so even the most important math functions have to be shifted. That's what I like about the 41C and 35s - their additional 8th key row with these functions makes them much more convenient in everyday use. If the 43s hardware offers these additional keys as well, please use them for the most important math functions.

Dieter

#65

For me two-key access through a menu is >much< worse than a shift. The 28 was the only calculator I ever owned that I actively disliked. You're right in that it's the same number of keystrokes, so maybe it's just me.

Re the WP 34S: it's an amazing accomplishment, a technical wonder. The completeness of its function set and the accuracy it achieves on a limited hardware platform are remarkable. And it provides access to whole menus full of functionality I'll never need with the same number of keystrokes as the more basic functions that I use every day. Unfortunately, after some effort, I was not able to get comfortable with it. This marvel sits unused--I pick up a 15, a 42, a 48, or even a 35 instead. It's a shame.

I think my problem is this: I'm no mathematical slouch, but there among those whole menus of functions that I'll never use are many that I don't even know how to use, that I've never even heard of being used. Instead of being buried, they're just as accessible as the functions I use all day every day. They get in the way-- they get in >my< way, anyhow-- and negatively impact this wonder's usability. I thought it rude to "complain," but since you brought it up, this is how I feel.

If I had a solution to offer short of user-replaceable keys, I would. Lacking one, however, I'm left to advocate for what I see as essential: 1/x, sqrt, y^x, trigs, logs, and their inverses. While I don't use them regularly, sigma+ and i probably need permanent and prominent (primary key) placement to be really useful for those who do.

The WP 43S is proof that you can't give everybody everything that they want. But take the basic on-keyboard functionality of the 15, 41, or 42, add access to the advanced functionality of the WP 34S, and a user menu and you've made a living legend. Don't worry if some of the functions are a few levels deep-- if we need something esoteric we'll find it and with the ability to customize the menus, this becomes a non-issue. The 42 was almost there with two fewer keys overall, only one shift key, and a lesser display. You can do it!

-Bill


#66

The WP 34S does not feature menus and it never will. This is due to its breathtaking display (as mentioned here frequently). It provides catalogs instead. Cataloged functions may be accessed rapidly using the alphabetic search method described in the manual. This method works so well IMHO that we could have put all the functions in one large catalog there after all. We didn't do that, however, partly for historic reasons, partly since the practical knowledge of the alphabet is ... ummh ... found to be not as extensive as one should think.

The 43S will feature menus and only one catalog. A menu and a catalog are two completely different cups of tea. Please check the information given in earlier posts and you'll know what a menu is.

d:-)

#67

From this description, it sounds like we shouldn't even be trying to design a 43S that you'll want to use. We're almost certainly going to have more features and functions than than the 34S, to cram those in we'll have to have menus of some kind -- things will have to be left off the keyboard I'm afraid. We're fairly committed to a user mode and an alpha mode and keeping everything else uncluttered would be nice as well. USER mode lets you put want you want where which ought to satisfy more people than us fixing the layout.

The 42S also has some shortcomings -- programming it can be painful, alpha entry is painful and there are many non-obvious functions in its menu system too.


- Pauli


#68

Walter: I apologize for my sloppy use of terminology and stand corrected. Know that your educational efforts are not wasted; see how readily I've adapted to LG instead of LOG.

Pauli: I certainly didn't mean to offend. You've already designed a calculator that I >want< to use! I'm just offering my suggestions on how your design might be improved in its next incarnation. The WP 34S improves on the 42S's alpha entry without significant sacrifice. This is a major improvement! The improvements in programming are less useful to me because I'm willing to put up with a bit of pain while programming, but I see how this might be more important to others.

It sounds like the new machine will also be "improved" by omitting the third shift. Because my eye is really drawn to the green-shifted labels, I see this as a big improvement, but I recognize that the tradeoff is an effective loss of keyboard real estate. I think the team is on the right track.

One "improvement" relative to the 42S that's been discussed is not pre-loading the top row, instead keeping these keys clear for soft keys. This is something that I never found to be an issue on the 42S. When I'm in its ASSIGN or any other menu I see right past the key labels. With a primary and two shifted functions on each of the six second-row keys and a primary and either one or two (based on earlier information from Walter) shifted functions on the top row, there is enough real estate for a "good" and uncluttered layout. It's just a matter of defining the criteria for "good" that satisfy "most" people. After that, USER will be icing on the cake.

-Bill


#69

I think, the reason why the 42S has the main functions of the top row printed on the keys is the relatively small display. In order to show two stack levels on the screen the menu line must be hidden most of the time. Without a menu line, the top row is useless, except if decorated with a set of default functions.

Compare this to the large screen devices (48 family) and the 17b or 27s. Both have a blank top row and permanent menus. Having permanent markings on these keys just doesn't make sense when the menu line is always shown.


#70

Of course, top row keys under permanent menu line of the display shouldn't be labelled. I'm under the impression that the development team are trying to create the best of two worlds rather than amended HP42s


#71

Just you wait a bit ;-)

#72

Pauli,

I do have difficulties counting your votes. Please read the OP again and clarify. Remember: no menus, 12 functions maximum.

d:-)

#73

unshifted: 1/x x^2 % USER ALPHA DISP

Shifted: sin cos tan asin acos atan

Dave


#74

Do you really want to see % as primary function?


#75

Seems he does. People are different. Else a function like %T never made it ;-)

#76

I do. I think % is more useful to more people than the trig functions. I think high end calculators should add to the functionality of low end ones without taking anything away. % is a common function on low end calculators for a good reason.

Dave

#77

And x^2 instead of sqrt? For the cost of a stack level and (assuming x^2 is primary instead of shifted) one keystroke, I can get to the square of x. I don't know another way (especially if logs and antilogs are absent) to get to the square root of x.

Not arguing, just curious.

-Bill

#78

P>R R>P

Kermit

#79

Unshifted: Sigma+  1/x  SQRT  LN   i     DISP
Shifted: Sigma- y^x x^2 e^x USER ALPHA
#80

12 (10  2)  1/x           
12 (12 0) SQRT
11 ( 2 9) x^2
11 sin cos tan / asin acos atan
10 ( 4 6) y^x
9 LN / e^x
7 P>R / R>P
5 ( 4 1) ALPHA
5 ( 3 2) USER
3 [Sigma+] / [Sigma-]
3 H.MS+ / H.MS-
3 ( 2 1) DISP
3 LG
3 i
3 %
2 IP / FP
2 %CHG
1 HR>H.MS / H.MS>HR
1 10^x
1 [theta]
1 [parallel]
1 x_ROOT(y)

d:-)


#81

Two of the most important prefix keys (ALPHA and USER) only on place 8 and 9?

And the complex entry 'i' even far behind??

Well, I guess this 43s will not be a calculator for me ... :-(

Franz

Edited: 17 Dec 2012, 10:55 a.m.


#82

If ALPHA and USER don't make it as primary I'll give it a miss too.

Looks like the biggest offender are trig functions - 6 out of 12 taken by them. I see no problem at all saving some 'real estate' by putting then in a menu.


Edited: 17 Dec 2012, 2:49 p.m.


#83

That's the inevitable risk of polls: sometimes the results don't meet your expectations. Bad luck ;-) But there's still enough time to push some of your friends to the polling station.

And remember it won't be a calculator designed by a committee - we'll be looking to the results but won't follow them blindly.

d:-)


#84

Quote:
But there's still enough time to push some of your friends to the polling station.

Ok, so here's my vote for those 12 keys:
ALPHA ALPHA ALPHA ALPHA ALPHA ALPHA
USER USER USER USER USER USER
:-)

Franz


#85

no further text


#86

ALPHA

and

USER

-- I agree. These really need to be primary functions, IMO.

#87

A problem is, a lot of people, myself included, would really like the trig functions to be on the keyboard. But they must* come or go as a group, so if you want them, that takes up six of the 12 available spots. I was willing to put them in a TRIGS menu, which would probably be active almost all of the time for me. That, plus my earlier choices in the poll, would cover about 99% of my normal use.


* - Obviously they do not have to come or go as a group, but most would expect to see all or none.


#88

Unshifted: sin cos tan sqrt 1/x ln
Shifted: arcsin arccos arctan ??? y^x e^x

As noted by others: x^2 can be easily accomplished, whereas sqrt(x) isn't quite as easy. If there is something more "valuable" it can be shifted sqrt.

% - I've never used this in forty years of calculator use.


#89

I also would agree that square root as primary makes more sense to me than x^2


#90

and the end of the day and after endless polls and bizantine discussions it'll turn out the best keyboard is the 41's - and they didn't obliterate anybody's opinion, go figure


#91

Quote:
... it'll turn out the best keyboard is the 41's ...

It cannot be the best because it has the basic arithmetic keys on the wrong side ;-) The best keyboard ever is the Ti59. If only Ti had been able to manufacture proper key contacts...


#92

@Maximilian: Schönheit liegt im Auge des Betrachters.

@Angel: You don't know anything about the polls and discussions at HP before the 41 was launched, do you?

d:-)


#93

Quote:
You don't know anything about the polls and discussions at HP before the 41 was launched, do you?

Much more than you can possibly think...


#94

:-) I miss-read you comment like:

"Much more than you can possibly drink..."

Who knows, what really happened at HP at that time :-)

#95

un-shifted: 1/x  sqrt(x)  e^x   ln  y^x  i
shifted: ?? x^2 10^x lg; ?? < (polar entry of complex umbers)


Edited: 17 Dec 2012, 7:41 a.m.

#96

Unshifted:  %     y^x     i     IP  P>R  H.MS>HR
shifted : %CHG LG [theta] FP R>P HR>H.MS
#97

Unshifted / Shifted:
Sin / Asin
Cos / Acos
Tan / Atan
x^2 / sqrt x
y^x / xth root of y
1/x / ??

No strong views on the last space, those above cover what I use most.

#98

15 (14  1)  SQRT 
15 (12 2) 1/x
14 ( 7 7) y^x
13 sin cos tan / asin acos atan
13 ( 3 10) x^2
11 LN / e^x
8 P>R / R>P
7 ( 6 1) ALPHA
7 ( 4 3) USER
5 i
5 LG
4 %
3 [Sigma+] / [Sigma-]
3 H.MS+ / H.MS-
3 [theta]
3 DISP
3 IP / FP
3 %CHG
2 10^x
2 HR>H.MS / H.MS>HR
2 x_ROOT(y)
1 [parallel]

... if I counted correctly. Thanks for participating.

d:-)


Possibly Related Threads…
Thread Author Replies Views Last Post
  [41CL] New Extra Functions version Monte Dalrymple 0 1,158 11-08-2013, 04:32 PM
Last Post: Monte Dalrymple
  HP Prime: in need of help with defining functions Alberto Candel 14 4,514 10-27-2013, 10:48 AM
Last Post: Alberto Candel
  HP Prime spreadsheet functions SanS 0 1,946 10-04-2013, 04:23 AM
Last Post: SanS
  Stats functions on the HP34S Nicholas van Stigt 5 2,102 09-24-2013, 02:45 AM
Last Post: Nick_S
  Trig Functions Howard Owen 11 3,709 09-16-2013, 02:53 PM
Last Post: Fred Lusk
  50g piecewise functions Kurt Fankhauser 6 2,427 09-15-2013, 08:01 PM
Last Post: Kurt Fankhauser
  Back to School/Work/Other Calculator Poll Eddie W. Shore 63 17,367 09-12-2013, 06:03 AM
Last Post: bhtooefr
  Missing functions on the HP Prime!!!??? :-( Namir 6 2,395 08-22-2013, 08:40 AM
Last Post: Gilles Carpentier
  HP41 Functions Address Table (F.A.T.) Antoine M. Couëtte 6 2,278 07-21-2013, 02:48 AM
Last Post: Antoine M. Couëtte
  Poll: SandMath and 41Z Overlays anyone? Ángel Martin 20 5,385 05-20-2013, 04:13 PM
Last Post: aurelio

Forum Jump: