[HP Prime] - EXP() or e



#8

Hello all

Please forgive me if what follows have already been pointed out here before.

Am I the only one worried, (should I say disturbed ?), by this program-like notation of EXP() ?

I've no problem with this in a program, but I really find it non correct on the display. The right mathematic symbol is e as you all know, and not EXP(). This calculator is supposed to be a math learning tool for kids (among other, but mainly), and I find this really disturbing and confusing. The closer it could be to the right mathematic notation, the best it would be, to my opinion.

More to say, TI and CASIO already use the correct notation e.

Is this due to the fact that the Prime uses a CAS program that uses this EXP() notation ?

Kind regards.

Edited: 23 Aug 2013, 5:11 p.m.


#9

It locks like the CAS environment is quite distinct from the home screen. In the latter, e^ is displayed as an exponential and not as EXP(). The same key in the CAS screen is shown as EXP().

#10

From the calculator's perspective, EXP() and e^ are two different things. One is a much less accurate power function.

Internally in HOME, e^ is rewritten internally to EXP, but still displays as e^. In the CAS side though, it isn't so simple.

Basically, all the complaints people have regarding display of things, or complaints about less strict handling of vectors/matrices/lists, all go away completely in the home screen. It is purely is the CAS screen where these strange things happen and stems directly from the decisions made by the CAS author.

Why haven't all those things been resolved? They just take time. Many other higher priority things were needed for a first release. I've heard from quite a few people how they are extremely impressed with Prime despite having initial misgivings. Usually the main two things mentioned are the "vocal" complaints they saw from a small minority, and then the fact that is isn't exactly like the 50g and so they weren't quite sure what to expect.

There will of course be plenty who disregard/dislike/don't purchase for on reason or another - as is their right to do so.

As a side note, the 50g does this exact thing. e^ gets converted to EXP internally and any time you look at an algebraic object it shows up. It is purely a cosmetic thing there as well. I haven't seen outrage for years about that though... :-)

Would I have liked to have the time needed to polish everything absolutely first? Sure. Reality just doesn't always cooperate.

I would point out though that people (not you, just a comment in general) who immediately dismiss prime as a "student only" machine trying to be an Nspire clone or something should take a closer look at it. Even a casual browse through the matrix menu should demonstrate that isn't the case.

A student will be able to easily learn it, use it, but, more importantly, *continue* to use it in higher ed or beyond. There is plenty of growth built in. It isn't designed to be just a "teaching tool" exclusively like the nspire so obviously is.

TW


Edited: 23 Aug 2013, 5:45 p.m.


#11

Hello Tim

I thank you for your answer. I will wait for next release(s) to have this minor problem fixed. I do understand this is purely cosmetic and that other more important improvements may be done before.
Regards.


#12

Quote:
From the calculator's perspective, EXP() and e^ are two different things. One is a much less accurate power function.

In giac/xcas, e is parsed as exp(1) and exp(1)^x is auto-simplified as exp(x). Therefore EXP(x) and e^x are equivalent.

Quote:
Basically, all the complaints people have regarding display of things, or complaints about less strict handling of vectors/matrices/lists, all go away completely in the home screen. It is purely is the CAS screen where these strange things happen and stems directly from the decisions made by the CAS author.

In fact, when I started the giac project, I decided *not* to make a difference between lists and vectors or lists of lists of the same size and matrices, because of my previous experience with the HP48 Erable and HP49 projects. It was so annoying to have to call array to list to list conversions, and there was no reason to do that since the container are exactly the same.
In addition, I have added the possibility to handle univariate polynomials directly in list form, i.e. extend PROOT/POLYROOT and POLYCOEF to more arithmetic operations. That's why for example [1,2]+3 returns [1,5]: [1,2] is considered as the list of coefficients of the polynomial x+2 (always descending order), you add 3 you get x+5, represented as poly1[1,5]. You can also multiply polynomials like that, take quotient and remainder with quo/rem, find Bezout polynomials with abcuv or egcd, and so on. Coefficients may belong to the rationals, but also to a finite field (Z/pZ with p prime using % on giac or %% on Prime or even a Galois Field extension, command GF).
All the operations that are sometimes commented here as strange have always a meaning. Sometimes there may be two or more ways to interpret an operation on lists, then you can select the right operation by replacing the opening [ of the list with poly1[ or matrix[ or set[ or seq[ or just call another operator (like .+ .* ...)
I can understand that people used to another calculator find it disturbing, and I'm ok to add a few warnings (e.g. for inverting a list of lists elementwise vs inverting a matrix) but I don't believe it would be a good idea to clone the HP49/50 or TI nspire CAS behavior vs lists/vectors because that would probably require a lot of work to add strict checks (and would require a lot more conversions command calls by the users after).
I have never received complains about that by Xcas users (which is now used in many highschools in France as a CAS), therefore I really believe that these complains are mainly caused by old habits, they won't affect new users.
It's in my view much more important to improve the CAS itself (like improve the regroup function for auto-simplification, or extend/speedup some commands).

Quote:
I would point out though that people (not you, just a comment in general) who immediately dismiss prime as a "student only" machine trying to be an Nspire clone or something should take a closer look at it. Even a casual browse through the matrix menu should demonstrate that isn't the case.

A student will be able to easily learn it, use it, but, more importantly, *continue* to use it in higher ed or beyond. There is plenty of growth built in. It isn't designed to be just a "teaching tool" exclusively like the nspire so obviously is.


giac is certainly not a nspire clone, it's designed to be a competitor to major CAS like maple. There are a lot more advanced math functions that (unless I'm mistaken) are not available on the TI nspire CAS like (not limited to)
- integer arithmetic (extended gcd, chinese remainder, modular computation, quadratic sieve for integer factorization)
- polynomial arithmetic (extended gcd, resultant, Groebner basis) also working on Z/pZ and finite fields
- exact linear algebra (kernel, Jordan normal form, symbolic matrix power, quadratic form reduction)
- some symbolic stuff like recurrence relation solver (rsolve command), or special functions like cosine/sine/exponential integral, or Laplace/inverse Laplace transform
giac is clearly not a highschool/first University year only CAS, you can still use it for computations during 3rd/4th year of University or later... This extends to the Prime CAS of course...

#13

Welcome to the HP Museum forum. Thank you for your contribution to HP calculators, and a myriad of other apps that make use of your project.

Best regards,
Pal G.

#14

Thanks Tim Wessman and Bernard Parisse for your clarification.
www.facebook.com/HpPrime. Spanish version:

Tim Wessman:
Básicamente, todas las quejas de visualización, o sobre el manejo menos estricto de vectores/matrices/listas, todas estas desaparecen por completo en la pantalla principal (home). Es en la pantalla CAS donde suceden estas cosas extrañas y se derivan directamente de las decisiones hechas por el autor del CAS.


Bernard Parisse:
De hecho, cuando empecé el proyecto GIAC(Xcas), decidí *no* hacer diferencia entre listas y vectores o listas de listas del mismo tamaño y matrices, debido a mi experiencia previa con los proyectos Erable HP48 (Software algebraico) y la HP49. Era tan molesto tener que llamar o hacer conversiones de matriz a listas, y no había ninguna razón para hacer eso ya que el contenedor era exactamente el mismo. Además, he añadido la posibilidad de manejar polinomios univariantes directamente en forma de lista, es decir, extender PROOT / POLYROOT y POLYCOEF a más operaciones aritméticas. Es por eso que, por ejemplo, [1,2]+3 devuelve [1,5]:[1,2] se considera como la lista de los coeficientes del polinomio x+2 (siempre en orden descendente), le sumas 3 y obtienes x+5, representado como poly1[1,5]. También puedes multiplicar polinomios así, tomar cociente y el resto con quo/rem, encontrar polinomios Bezout con abcuv o EGCD, y así sucesivamente. Los coeficientes pueden pertenecer a los racionales, o también a un campo finito (Z/pZ con p primo usando % en Giac o %% en la HP Prime o incluso una extensión del campo Galois, comando GF). Todas las operaciones que a veces se comentan aquí como extrañas tienen siempre un significado. A veces puede haber dos o más maneras de interpretar una operación en listas, entonces se puede seleccionar el correcto funcionamiento mediante la sustitución de corchete "[" de la lista con "poly1[" o "matrix[" o "set[" o simplemente llamar a otro operador (como .+ .* ...). Puedo entender que personas acostumbradas a otras calculadoras les resulte inquietante, y estoy de acuerdo con agregar algunas advertencias (por ejemplo, invertir una lista de listas de elementos racionales vs invertir una matriz), pero no creo que sea una buena idea clonar el comportamiento de las HP49/50 o la TI Nspire CAS vs el de listas/vectores debido a que probablemente se requeriría mucho trabajo para agregar verificadores estrictos (y se requerirían mucho más comandos de llamadas de conversiones después para los usuarios). Nunca he recibido queja de que los usuarios de Xcas (que ahora es usado en muchas escuelas en Francia como un CAS), más allá de esto, realmente creo que estas quejas se deben principalmente a los viejos hábitos, y no van a afectar a los nuevos usuarios. Es, en mi opinión, mucho más importante para mejorar los CAS en sí mismo (como mejorar la función reagruparse para auto-simplificación, o ampliar/acelerar algunos comandos).


Tim Wessman:
Me gustaría señalar sin embargo, que las personas (sin referirme a alguien en particular, es sólo un comentario en general) que inmediatamente desestiman la HP Prime como una máquina "Sólo para estudiantes" tratando de ser un clon de la Nspire o algo así, deberían echar un vistazo más de cerca. Incluso una breve exploración al menú de matrices les debería demostrar que este no es el caso.
El estudiante será capaz de aprender fácilmente a usarla, pero, más importante aún, *continuará* usándola al acabar la escuela o más allá. Hay un montón de crecimiento dentro. Obviamente no está diseñada para ser exclusivamente sólo un "instrumento de enseñanza" como el Nspire.

Bernard Parisse:
Giac no es un clon de la Nspire, Giac está diseñado para ser un competidor de los principales CAS como el Maple. Hay muchas de estas funciones matemáticas más avanzadas que (si no me equivoco) no están disponibles en la TI Nspire CAS (en particular) como - aritmética de enteros (mcd ampliado, resto chino, computación modular, criba cuadrática para la factorización de enteros) - aritmética polinómica (mcd extendido, resultante, base de Groebner) también trabaja en campos finitos Z/pZ y - álgebra lineal exacta (kernel, Jordania forma normal, potencia en matriz simbólica, reducción de forma cuadrática) - algunas cosas simbólicas como solucionador de relación de recurrencia (comando rsolve ), o funciones especiales como integral de coseno/seno/exponencial, o Laplace/Laplace inversa y transformada de Laplace, Giac claramente no es un CAS solo para la escuela/primeros año de Universidad, se podrá utilizar para cálculos durante 3ro/4to año de la Universidad o después.. . Por supuesto esto se extiende al CAS de la HP Prime ...


Possibly Related Threads...
Thread Author Replies Views Last Post
  HP35 exp(ln(x)) bug, a new number found. Kiwi Geoff 4 439 08-27-2010, 08:45 AM
Last Post: Dieter
  (O.T.) Happy 'XROOT(9,10561572*EXP(2))/sqrtLN(pi)' ! (N.T.) Gerson W. Barbosa 8 791 07-07-2009, 07:05 AM
Last Post: Valentin Albillo
  HP 35 ROM V2 "exp(ln(2.02) = 2", explanation found Jacques Laporte 10 859 03-24-2006, 11:57 AM
Last Post: bill platt
  Scientific calculators, usage of exp Lesley Evennsenn 24 1,650 06-15-2002, 09:26 AM
Last Post: Ellis Easley
  HP75 Memory Exp. Pod Mike Davis 4 428 10-29-2000, 12:02 AM
Last Post: Mike Davis

Forum Jump: