The Best Graphing Calculator on the Market? 724
aaronbeekay asks: "I'm a sophomore in high school taking an honors chem course. I'm being forced to buy something handheld for a calculator (I've been using Qalculate! and GraphMonkey on my Thinkpad until now). I see people all around me with TIs and think 'there could be something so much better'. The low-res, monochrome display just isn't appealing to me for $100-150, and I'd like for it to last through college. Is there something I can use close to the same price range with better screen, more usable, and more powerful? Which high-tech calculators do you guys use?"
Are you kidding? (Score:1, Interesting)
First Post?
HP 48 4-Life!!! (Score:5, Interesting)
RPN is very nice for long equations. Once you get used to it, you'll be more accurate and efficient. You'll never want to go back to algebraic entry. It has a lot of features, and still stands up pretty well to modern offerings. Unless they've made calculus problems a lot harder, you won't need anything more functionality wise.
The built in equation library is very nice. There is a plethora of available programs to download. The IR sensor is just cool and the keys have the best tactile feel of any calculators ever, and the batteries last about 20 months. Oh, and you could probably dip it in motor oil, and it would still work. The screen while having good contrast, is very fragile however. That's one bad thing.
Expect to pay $250 on ebay for a 48GX unless you get lucky. (The 128K expandable model. Original MSRP was $159 I think) You can probably get a 48G (32KB non expandable model) in your price range though.
Re:HP (Score:2, Interesting)
I must agree with you in many ways here. But to me (also a physicist) the HP-33 has too much clutter and is too slow. The TI-36X does 99.5% of all the calculations that I need and since I've used it so long I no longer need to look at the keypad while entering commands. It also has the benefit of only allowing you to only input one calculation at a time which will help prevent errors (like missing a parentheses or accidentally evaluating an exponent on only the numerator). For more complicated calculations using matrix transformations or graphs of non-linear systems the HP-48/49 and soon to be HP-50 series can't be beat.
For anyone who is planning to be a physical scientist or an engineer, a powerful calculator is a handicap and will hurt you in the long run. The ease of solving problems in low level math courses will come to haunt you when you take a course that includes something like Laplace transforms or complex analysis.
Re:RPN Baby! (Score:5, Interesting)
A friend of mine at MIT had an HP-48, and I had a TI-81, we used to do a lot of engineering problem sets together and would often race on entering calculations. Averaged over time the competition was a draw. Although the HP-48 definitely wins from a "cool" factor perspective (where cool=geek).
Speaking of the TI-81, I bought mine in 1991 for $82, and I'm still using it every day.
Re:Save your money (Score:4, Interesting)
And you never know when being able to do things by hand is going to save your ass.
I recall a physics exam my freshman year of college, fairly simple mechanics stuff: find how long something takes to slide down a ramp, that sort of thing. About 10 minutes into the hour long exam my calculator blew up. Something in the LCD burst, it was a paperweight.
This was the kind of tech school where the professors just don't give a shit about your issues, and where too many missed exams counted against you heavily; leaving in the middle of one without completing it was the same thing. I was fast enough to get everything but one problem finished with 40 minutes to spare even without the calculator. Only problem was that the answer involved multiplying by the sine of an angle.
I had a couple of sin and cos values memorized: 30 degrees, 60 degrees. Had memorized the square roots of 1 through 5 to a few places, and happened to know how to compute those by hand as well.
Ever come across these formulas?
sin(x/2) = ± sqrt([1 cos x] / 2)
cos(x/2) = ± sqrt([1 + cos x] / 2)
sin(a±b)=sin(a)*cos(b)±sin(b)*cos(a)
Well, if you know sin(30) and cos(30), from these you can compute the values at 15 degrees with a few mathematical operations, then 7.5, then 3.75, etc. Build that little table, and then you can add or subtract things together to reach other values, and maybe throw in a little linear interpolation. Eventually I build an estimate answer using this approach that was close enough to get most of the points for the problem. Got dinged for not using enough significant digits, as if I'd made a rounding error, but got most of the credit.
When time was called I was in the middle of trying to check my answer against the results of a Taylor Series computed with Horner's Rule [dattalo.com]. Converting degrees to radians by hand is a snap once you've memorized Pi to a thousand places [youtube.com]...
Re:TI-89 Platinum (Score:3, Interesting)
I recently got a TI-89 Platinum for use in several science (and calculus >_</) courses over the next few years. Despite the fact that the HP-48 and HP-50 are technically superior, and RPN is the fucking win, I chose the TI anyway, and for one reason: software.
There is TONS of homebrewed software out there for TI calcs, and I'm already relatively familiar with m68k assembly, from coding on my C=64 back in the day (though I'm horribly rusty), so I don't have to learn to write for ARMs for the HPs. I also looked for homegrown softs for HP calcs, and found the results wanting.
I have several incredibly useful and easy-to-use chemistry tools, and lots of other good stuff for my TI, and there is a huge community. Not to mention the link software is actually well designed, and easy to use~
Link to huge amounts of TI calc software:
http://www.ticalcs.org/
Re:HP (Score:5, Interesting)
strong statement as to their durability.
Re:TI nspire (Score:4, Interesting)
If you can prove me wrong, and show that the nspire is as accessible as the TI-85, I might buy one just for day-to-day field engineering needs.
Re:WHy any? (Score:5, Interesting)
I'm also a Ph.D. student in math (defending my dissertation next month), and I've found the exact opposite to be true. There's no better way to develop a deeper understanding of something than to play with it. As regards calculus and functions, this means plotting functions, composing them, zooming in on them, adding them, differentiating them, multiplying them, etc. This is especially relevant with polar and parametric equations, which can take some time to get the hang of.
The newer calculators even let you play with systems of differential equations and trace out solutions, flow lines, etc. What a great way to learn to visualize otherwise abstract concepts! If students would just sit and play with equations and see what the solutions would look like, they would have a much better grasp of what to expect when they encounter something new. Otherwise, it can tend to be a matter of memorizing a cook book of solution techniques.
Of course, there are times when the calculator can be a hinderance. In particular, the built-in symbolic differentiation and integration can become a crutch. (On the other hand, it's a great way to check your answers.) However, most of the associated problems can easily be dealt with by properly writing your curriculum. (e.g., giving calculator-free exams to test differentiation knowledge, splitting them into two-part exams (without calculator, then with calculator), giving weekly 5-minute self-quizzes, etc.)
At the end of the day, a graphing calculator is just another tool that can be used to help or hinder education. How it goes depends on a combination of student motivation and the leadership and guidance they receive from their professors and teaching assistants. (i.e., you) -- Paul
No you have to use TI (Score:4, Interesting)
Oh come off it. (Score:4, Interesting)
* Quick factoring of integers, radicals, polynomials
* Term collection and simplification
* Handling of arbitrarily large values without loss of precision (esp w.r.t. factorials)
* Substitution of variables or expressions in general formulas (user-provided function)
It really can't "solve" very much other than 4th degree polynomial roots. It's really just there to help you manipulate a complex expression without making a mistake (but you really need to be doing the manipulations... which of course requires a bit of knowledge, don't it?)
BTW I distinctly remember adding the incomplete beta and gamma functions to my TI-89, and I think error function too. They would simplify to trivial expressions if they could (to promote further manipulation) or returned numerical solutions if so coerced. I thought it was pretty slick...
Re:PDA? (Score:2, Interesting)
Re:TI nspire (Score:2, Interesting)
I've considered getting a TI-89 Titanium just so I can use a USB cable instead of the grey GraphLink cable (yeah, I have one of those old things). But I don't think I will get one because I don't like the keyboard and form factor. The original 89 is easier to hold and the keyboard is easier to use, in my opinion. So if the answer to my first question is yes, my real question is have these things been considered? A mouse like interface is "cool" but I don't see how it will aid a student who is trying to do some quick calculations. Remember, a student has only fifty minutes to seventy-five minutes (at least, at my university [clayton.edu]). Are you guys testing that calculator under such conditions?
Re:RPN Baby! (Score:4, Interesting)
Then your friend was slow -- or you were very quick. Take some complex expressions and write out the keystrokes required in RPN and infix notations, and you'll see that RPN almost always wins. However, the big win isn't the keystrokes, it's the mental complexity. With infix, you have to maintain too much state in your head -- with particularly nasty expressions, you basically have to keep track of the whole expression in order to enter it all correctly, closing the parentheses at the right times. With RPN, you think about it differently, "collapsing" subexpressions as early as possible, minimizing the amount of you have to hold.
My friends and I ran a series of tests in college, specifically to determine which was more efficient. Not only did the postfix evaluations typically have 10-20% fewer keystrokes, the person writing the postfix version typically finished writing the evaluation while the person writing the infix was still figuring out how to express it. What finally convinced the doubters in our little experiment to buy HPs was that the infix evaluation got the wrong answer much more often than the postfix evaluation did -- usually because of some miscounted parentheses.
RPN is faster, easier and more accurate on complex expressions.
Re:HP 48GX is an Amazing Calculator (Score:2, Interesting)
'y=x*x-3' enter
'y=22' enter
'x' solve
(or something like that) and it would solve the equation for you, giving 'x=5'. A few other examples:
'y=x*x-3' enter
3 +
sqrt
would leave 'sqrt(y+3)=x' on the stack.
It could also do bracket expansion, simultaneous equations, formulaic integration / differentiation ('y=x*x' 'y' differentiate = 'dy/dx = 2x'), had a full hierarchical filesystem and was fully programmable (there are even assemblers for it on the net). Though programming in RPN is difficult at best.
Oh, it also did unit conversion.
Re:IA32 + Matlab R13 (Score:2, Interesting)
I don't get the graphing stuff - can someone help? (Score:3, Interesting)
The only time I have ever seen it used is to show the multple zeros of an equation, but even that was just a curiosity. If you can't get a pretty printout, why bother? Furthermore, you need the exact numbers anyway whenever you want to solve something. If you want to estimate, do it in your head.
Admittedly, I own an HP48, so I use the screen as a visual stack. Again, all of the graphing fuctions are pretty, but not practical unless you happen to be using it for a game, or calendar, or as a help screen in an equation (and if you need a help screen, imo you don't know the equation well enough to be using a calculator).
So, are there really useful or computationally practical reasons for a graphing calulator, or does everyone just want them because they are "cool"?
Re:WHy any? (Score:2, Interesting)
The learning process is not the same from one student to another, especially with abstract ideals like mathematics (or computers for that matter), and the fact you have a PH.D. means you were interested enough in the subject to train your brain to remember and learn its intricacies, but now that you've learnt them, do you remember the steps you took? do you remember how much time you spent on it?
I've often been faced with bad teachers because they think that PH.D. = good teacher.. it doesn't.. a good teacher needs these qualities:
1) understand the subject
2) understand the progress from a fresh student not knowing anything about the subject to a student who does, know what concepts and exercises he should be able to work on at the end.
3) be able to empathise with students regarding issues with the subject. Abstract thinking means you have to connect ideals with things in your own experience (like many of us do) or create a totally new experience around it (which I believe many of the more capable abstract thinkers do - they spend the time, and have the motivation to create this experience) - this means you have to understand how students think about the subject, and how they understand it.
4) be able to inspire students to keep trying until they succeed.
Learning maths should be no different to a motivated student then a child motivated to learn to walk, it'll take time, effort and a little guidance. And so what if he uses a nearby chair to support himself while taking those first steps, all you have to do is guide them a little, and any support will then have merely been a stepping stone and not a crutch.
Of course, if motivation is not at hand, it's up to you to decide whether you want to motivate people or not, and if the answer is no, then maybe teaching isn't your forte.
K.
Maybe offtopic: The Curta (Score:2, Interesting)
Re:WHy any? (Score:3, Interesting)
"Regular" calcs are not PN calcs (Score:3, Interesting)
The "regular" calculators with equal sign are not PN calculators.
Re:IA32 + Matlab R13 (Score:3, Interesting)
I'm not sure about the models from only 12 years ago, but the HP-41C I bought around 1982 or so still works perfectly. Truth be told, it works better than new, thanks to it's accepting add-on modules (of which all four slots are permanently full). The newer HPs don't seem to be quite as solidly build as that, but they're still quite a bit better than the TIs and Sharps.
The comments about RPN being difficult really are nonsense. If you can't figure out RPN, my advice is to forget even politics and just live under a bridge -- though you'll probably have to fight for a spot, since the other homeless people will neither respect nor wish to associate with you.