MathGroup Archive 2004

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

Search the Archive

Re: Re: Programming style

  • To: mathgroup at smc.vnet.net
  • Subject: [mg47627] Re: [mg47608] Re: Programming style
  • From: schmitther at t-online.de (Hermann Schmitt)
  • Date: Mon, 19 Apr 2004 04:32:58 -0400 (EDT)
  • References: <c5qj4q$fvh$1@smc.vnet.net> <200404180815.EAA18000@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

Hello,
The assumption, that functional programs are better than procedural programs
has exceptions:
A recursive program is often more expensive than using a loop instead.
See also:
R.E. Maeder: "Computer Science with Mathematica." p.36
Hermann
----- Original Message -----
From: "Mark Westwood [EPCC]" <markw at epcc.ed.ac.uk>
To: mathgroup at smc.vnet.net
Subject: [mg47627] [mg47608] Re: Programming style


> Lorenzo
>
> Here's my five-pennyworth, I expect that many others in this newsgroup
> will chip in with their own strongly-held, entirely reasonable and
> mutually contradictory opinions:
>
> Yes, it is worth the effort to learn the Mathematica way of doing
> things, the 'functional' method as you put it, just as it is worth the
> effort to learn the object-oriented way when learning Java or Smalltalk
> etc.  Their are two reasons for making the effort:
>
> 1) equivalent programs written in the functional style and the
> procedural style usually execute faster, in Mathematica, in the
> functional style;
>
> 2) functional programs are often shorter, easier to write and easier to
> understand than procedural programs - once you have enough experience of
> Mathematica that is;
>
> 3) writing functional-style programs is much more natural in Mathematica
> than writing procedural programs; when I try to write procedural
> programs in Mathematica I always feel that I am fighting against the
> system rather than with it.
>
> (OK, so that make's three reasons ...)
>
> In my second paragraph I place the word functional in quotation marks, I
>   won't be surprised to read other answers to your enquiry which deny
> that Mathematica is a functional language - pure functional languages
> don't do assignment for instance.  I think it's functional enough to be
> considered a functional language.  But you should also make some effort
> to get your head around the concept of Mathematica as a term rewriting
> system, transforming expressions in one form into an equivalent (usually
> simpler) form.
>
> If the book you have is the one by Nancy Blachman then stick with it.  I
> think it is the best introductory text for general purpose Mathematica.
>   Once you've finished with it you will be ready to digest The
> Mathematica Book itself.
>
> good luck
>
> Mark
>
>
> lorenzo.keegan at handbag.com wrote:
> > Hello,
> >
> > I am an experienced computer programmer having used ADA, C, Visual Basic
> > etc. for years.
> >
> > Recently I bought the book "Mathematica - A Practical Approach", where
different
> > styles of programming are discussed, namely
> > Functional programming versus Procedural programming.
> >
> > The book seems to suggest that most Scientist and "professional"
Mathematica
> > users prefer the Functional programming approach.
> >
> > For years I have been working with the "Procedural" method.
> >
> > What are your feelings ? Is it worth the effort to learn the Functional
> > method ? Does it matter ?
> >
> > Thanks for your time
> >
> > Best Wishes
> > Laurence Keegan
> >
>


  • Prev by Date: Re: Elliptic Curves and Cryptography Questions
  • Next by Date: Re: plot combination
  • Previous by thread: Re: Programming style
  • Next by thread: Re: Re: Programming style