From: Nuralanur@...
Date: 2006-07-05T03:49:46+09:00
Subject: Re: ed borasky and his second favourite programming language

-------------------------------1152038972
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit

Dear Matthew,
 
you are running into open doors there - I did not write my last
email to criticize anybody, but to ask why Ed Borasky likes
to use R - that's what the link Ara posted was promising
to tell me - and I have no answer for that yet.
That question is of course different from the question,
"Why does Matthew Smillie like to use R?", "Why do I use R?"
 
Your statement 
 
> I can't speak for Ed, but for me (I use it as well), it does   
>precisely what it needs to do, and it does it succinctly and  well.   
>Which, not coincidentally, are the same reasons I use  Ruby for the  
>things I use Ruby for, and the same reasons I use  Prolog for the  
> things I use Prolog for.

does not tell me what ordering of preference you would
give to R, Ruby, or Prolog in any application nor for what  reason.
That sentence is just no more informative than saying,
"I use language X". That obviously means that you'll have a
reason for that - it could be "it's the  easiest/ the most
convenient tool for doing task ...., because ...../I know of no other
language (to do this task properly)/I've used that language for so-and-so  
many years and
I'm too old to change now/it can output 'hello world' faster than language  
Y",
 but you'll have to tell me, because I don't accept all of those on an  equal 
footing.
I have talked to people who gave me any of those reasons for choosing
their preferred language - maybe with a slight bias towards the "I'm  too old
to change now" argument, which I personally don't find acceptable at  all.
 
Hearing those reasons might be interesting because one could then profit  
from 
other people's experience with any of these languages for a future  project.
 
Regarding R, I'd like to hear how it compares to Gnuplot for plotting in  
other people's
opinion.
 
For scientific computations, I use the language kan/sm1.
You can do Groebner basis calculations in differential ideals in it,
(that's to do with algebra) not so much because I like its syntax so  dearly,
but I prefer it over Maple and Mathematica, because I like their  syntax
even less and they don't have an efficient implementation of a b-function,  
which
is something I need quite dearly ( and am too lazy to implement  myself).
I also use Ruby to put together programs for kan/ms1, and when I say that  
Ruby is "more general" than R (notice the quotes!), I mean that it has been  
designed as a general
computer language, not just for one particular field of applications, as 
statistical evaluations. That's no criticism of R, though.
 
This generality implies that you can use bindings for R from Ruby, and  thus 
use R without having to learn R's syntax from scratch (maybe just for one  
dataset evaluation - which could be as far you can tell when you first use  R).

 
If you use Ruby and kan/sm1, there are several packages on Rubyforge
to let you define multivariate polynomials in Ruby and calculate with them  
in Ruby, so 
Ruby's possibilities go beyond invoking some external program and  reading
out that program's output, but you can treat that result in Ruby.
That would not be possible for R for this particular task, so R is less  
general
than Ruby. But I don't mind - that language wasn't designed for this  
purpose. R does statistics and plotting, and it's good at that - but I still  have no 
answer to the
question we started with.
 
Best regards,
 
Axel 
 

-------------------------------1152038972--