[Date Index] [Thread Index] [Author Index]

Re: How should I start with mathematica?

• To: mathgroup at smc.vnet.net
• Subject: [mg85412] Re: How should I start with mathematica?
• From: David Bailey <dave at Remove_Thisdbailey.co.uk>
• Date: Sun, 10 Feb 2008 05:08:32 -0500 (EST)
• References: <fobjom$mqj$1@smc.vnet.net> <foekpa$ieq$1@smc.vnet.net>

David Park wrote:
> 1) Go to Help, Documentation Center, and read and type. Especially Core 
> Language, Mathematics and Algorithms, Visualization and Graphics.
> 
> 2) Actually type in the Help commands in your own notebook, try variations 
> that you can think of, and make sure they work.
> 
> 3) Pay special attention to the Functional Programming commands. Just keep 
> pushing the envelop.
> 
> 4) In graphics don't overlook the use of graphics options, which at first 
> may seem complicated and confusing.
> 
> 5) Take the time to learn the syntax and basic commands before undertaking a 
> major project. Then try to use Mathematica 'flying solo' on simple topics 
> you already know the answer to.
> 
> 6) Use MathGroup. There are NO silly questions. Silly people wouldn't even 
> use Mathematica in the first place. Mathematica is complex enough that 
> nobody knows everything about it and everybody deserves help.
> 
> 7) Learn how to use the sectional organization of notebooks and learn 
> especially the use of Text cells to document and explain what you are doing. 
> Text cells are almost as important as Input/Output cells.
> 
> 8) As you become more advanced try to develop and calculate everything 
> actively within Mathematica. Write the definitions you need and use them. 
> You will have to do this. Everything that might be convenient for your work 
> will not be directly in Mathematica. This requires some effort but the 
> payoff is enormous in terms of understanding, consistency and self-proofing.
> 
> 9) Try not to think of Mathematica as just a 'super calculator' or as a 
> 'programming language'. And especially don't waste time trying to morph 
> Mathematica into some other programming language. Think of a Mathematica 
> notebook as a piece of paper on which you are writing and developing your 
> technical ideas. Once you become resonable proficient with Mathematica you 
> can think in terms of your subject matter and not in terms of computer 
> science. We have already paid WRI good money to do the computer science.
> 
> 10) So this is what I wish for you: when you click the Mathematica icon it 
> will be just like when Euler reached for pen and paper. Maybe even better.
> 
> 
I would just like to add a couple of suggestions to David's excellent list.

As you may be aware, prior to version 6, Mathematica used to come with a 
  notoriously thick book. If you can get hold of a copy of one of these 
books - say for version 5 - read it. All the basic operations are the 
same except that some new functions have been added. The graphics and 
frontend programming have changed somewhat, however, as have the 
built-in packages.

Someone here (I wish I could remember their name) wrote a program to 
stitch together the 6.0 documentation into a version of this tome - if 
you can get hold of that, it would certainly be useful.

Learning Mathematica is a bit like learning a language - reading other 
people's code is essential. At first, other people's code can seem very 
daunting indeed (some of it remains that way :) ). One useful way to 
explore other people's code is to click on a part of an expression, and 
press control-. one or more times. This progressively expands the 
selection to cover mathematically meaningful portions of the code (try 
it first on a few simple expressions of your own - or something from the 
documentation. Picking portions of code and testing what they do in 
isolation can be a great way to get into otherwise obscure code.

David Bailey
http://www.dbaileyconsultancy.co.uk