Welcome to the MacNN Forums.

wataru · Mar 27, 2003, 04:51 PM

I'm considering purchasing Mathematica, but since it's not exactly cheap (even from my school) I wanted to get some opinions. I'm majoring in physics, and my professors have mentioned it in lectures several times. Is it worth the money? How hard is it to use? How does it compare to MatLab? For integrals and stuff my TI-89 has served me well thus far, but perhaps it's time for bigger and better things.

clarkgoble · Mar 27, 2003, 05:17 PM

If you are majoring in physics it will help you immeasurably in your upper division mechanics, optics and E&M classes. On the other hand, since you *won't* have Mathematic in your exams, doing it the hard way may actually be more beneficial. So if you do buy it, be aware that you are losing that essential practice at guessing how to solve complex equations. That's part of what doing all those rather difficult calculus problems involves. Half the work is guessing the right approach to solving them. After doing a few hundred problems, each with plenty of misstarts, you should develop a fairly good mathematical intuition.

Just like kids who use calculators and never learn arithmetic, using mathematica and never really grasping calculus can be dangerous. A lot of those skills are essential for a physicist - that intuitive feel for the math.

deekay1 · Mar 27, 2003, 05:19 PM

get photoshop instead. it's better. math sux anyway.

Xaositect · Mar 27, 2003, 05:31 PM

Mathmatica: Worth every penny.

I'm a math major (lots of physics, too). The above note about still doing some by hand is vital to remember if you want to pass the exams, so don't ignore it.

Good Luck on the degree!!

BlackGriffen · Mar 27, 2003, 05:58 PM

Welcome to the club! I go to a Maple university, myself, and I actually needed it for one of my courses. It's cost is better than Mathematica, AFAIK, to boot. The down side: not OSX native yet.

Something worth noting, though, is that the syntax of Maple is more like C than Mathematica. It's still a bitch to code for, though, because there are a bunch of little 'gotchas' that I haven't quite figured out that deal with variable creation, initialization, and destruction. I would have done it in C if I had an arbitrary precision library at the time (I was running a simulation that I expected to generate a fractal on my 300 MHz iBook, left the thing running for 5 days straight, and I only managed to get about 15% of the fractal I was looking for despite the low resolution). What's the point of this? Oh, yeah, Maple is a good alternative, give it a look.

Now, if only the would finish the damn graphical front end (the CLI version is apparantly already running).

BlackGriffen

clarkgoble · Mar 27, 2003, 07:03 PM

I'd prefer Maple myself. Most people I know doing relativistic work use Maple over Mathematica. I've actually been out of physics for several years, so I've not kept up on the state of packages. But for a while Maple had some very nice tensor packages that weren't easily available for Mathematica. I believe some are available now, but that they charge you quite a bit for them.

Doing all your mechanics in general tensors is very useful. You also get a much better grasp on what is going on, albeit at a much more abstract level. At my college they actually taught an optional tensor analysis class doing lots of mechanics questions. The trick was to think up the best coordinate system, calculate the metric and problems that may have taken pages of calculus to solve could be solved quickly.

Combine this with Maple and get a nice book of mechanics problems. Practice, practice, practice and you'll have an amazing conceptual grasp on physics that will aid you well on the GRE and also your entry level grad classes. (Not to mention allowing you to actually understand GR and QM much better)

BTW - my comments about doing the equations by hand goes beyond just passing your finals. A lot of physics is pure intuition. The more questions you practice on the better you'll develop your intuitive skills. Further most of the real significant aspects of physics are never taught in your classes. You are expected to pick them up on your own.

We always joked that physics was the one major that required knowledge of all other majors. (i.e. good writing skills for papers, good engineering skills for building your own lab equipment - both mechanical and electrical, good computer skills for running your simulations and controls, excellent math skills for solving the damn problems, some philosophy if you want to actually understand what is going on) Yet I was always amazed that the underlying problem solving skills are never formally taught. You sort of pick them up as you go along. Yet those skills are what separate the physicist from the mathematician or a person just quoting facts about the universe.

gorgonzola · Mar 27, 2003, 07:12 PM

I haven't used Maple, so I can't really compare, but Mathematica is an extremely nice program that is extraordinarily powerful while still being easy to use. Although the help system is nonstandard, it's quite thorough and you won't have much trouble finding what you need. You can try out Mathematica for integrals at http://integrals.wolfram.com .

The syntax is pretty simple...stuff like Sqrt[Sin[Pi]] and whatnot (obviously, it gets more complex, but you get the idea). I dunno about Maple, again, but Mathematica has a LaTeX export feature that's rather useful.

I don't really know whether specific packages you may want are available or whether Maple is better at certain specific advanced things (relativity theory or whatever), but Mathematica is definitely a *VERY* powerful program and is worth every cent.

In addition, it's stable and fast on OS X and is quite a good OS X citizen. It makes good use of OS X technologies, and the math typesetting it does looks really beautiful. You can get an idea for the kind of stuff it can do and what it looks like at Physics World and MathWorld.

Cipher13 · Mar 27, 2003, 07:30 PM

My uni is also Maple-centric. I'm not into hardcore maths (I despise it), but I still have to do it, and Maple seems to do quite a good job, all considered. I've never used Mathematica, but given my experiences with Maple, I wouldn't hesitate to just get it without trying Mathematica.

BlackGriffen · Mar 27, 2003, 07:50 PM

Originally posted by gorgonzola:
I haven't used Maple, so I can't really compare, but Mathematica is an extremely nice program that is extraordinarily powerful while still being easy to use. Although the help system is nonstandard, it's quite thorough and you won't have much trouble finding what you need. You can try out Mathematica for integrals at http://integrals.wolfram.com .

I will say that I have used both, recently, and I prefer Maple overall. Well, final assessment still pending the OSX release. That being said, the help systems are actually similar in the two packages.

The syntax is pretty simple...stuff like Sqrt[Sin[Pi]] and whatnot (obviously, it gets more complex, but you get the idea).

sqrt(sin(Pi)); check.

I dunno about Maple, again, but Mathematica has a LaTeX export feature that's rather useful.

check, haven't used it, but it's there.

There are only two strikes against Maple I can think of: one, no OSX native version (grrr!); none of the versions I've used (6, 7) support visual clues for closing of parentheses. I really like how Mathematica highlights any parentheses, brackets, or whatnot that don't have a matching partner. I must say, though, that if you now anything about coding in C, coding in Maple is very easy (the look control is a bit different, and like I said, there's some weirdness with variables).

BlackGriffen

OreoCookie · Mar 28, 2003, 10:14 AM

Originally posted by wataru:
I'm considering purchasing Mathematica, but since it's not exactly cheap (even from my school) I wanted to get some opinions. I'm majoring in physics, and my professors have mentioned it in lectures several times. Is it worth the money? How hard is it to use? How does it compare to MatLab? For integrals and stuff my TI-89 has served me well thus far, but perhaps it's time for bigger and better things.

Being a physicist myself, I have mainly worked with Maple and Mathematica. MatLab is more focussed on numeric stuff.

wataru · Mar 28, 2003, 10:24 AM

Wow, some good feedback here.

Since I don't need anything right now, I think I'll just wait for Maple to release their OS X native version (with GUI) and then read reviews and whatnot.

funkboy · Mar 28, 2003, 09:02 PM

I'm currently using Mathematica for a class in Differential Equations (use of the program is required for class), and I've found it quite a neat little program. Extremely powerful, much more power than I need for a little 200-level undergraduate math class.

I hope they update the OS X version again - it uses the "old" anti-aliasing, not the totally smooth Quartz stuff, but that may even be due to the non-standard way it has to format some text.

If you really are interested in doing some powerful math stuff, Mathematica's not a bad place to start. I believe it costs around $99 at my school's store. I don't think I'll use the program much after this math class I'm currently in, though. Electrical engineering doesn't involve using it directly much I don't think.

OreoCookie · Mar 29, 2003, 04:14 AM

Originally posted by wataru:
Wow, some good feedback here.

Since I don't need anything right now, I think I'll just wait for Maple to release their OS X native version (with GUI) and then read reviews and whatnot.

I prefer Maple's syntax to Mathematica's. Do you know when Maple for OS X will be released?

nickdman · Mar 29, 2003, 04:56 AM

Yes couldn't Integrate without it

wataru · Mar 29, 2003, 02:27 PM

Originally posted by OreoCookie:
I prefer Maple's syntax to Mathematica's. Do you know when Maple for OS X will be released?

As has been mentioned before, they have a CLI version working right now. According to their FAQ, the full GUI version will be done before the next academic year: http://www.maplesoft.com/support/Faq...eneral/1.shtml