Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!zaphod.mps.ohio-state.edu!rpi!bu.edu!orc!inews!iwarp.intel.com!psueea!eecs!warren
From: warren@eecs.cs.pdx.edu (Warren Harrison)
Newsgroups: comp.software-eng
Subject: Re: complexity question
Message-ID: <3196@psueea.UUCP>
Date: 3 Aug 90 08:00:40 GMT
References: <39400115@m.cs.uiuc.edu>
Sender: news@psueea.UUCP
Reply-To: warren@eecs.UUCP (Warren Harrison)
Organization: Portland State University, Portland, OR
Lines: 36

In article <39400115@m.cs.uiuc.edu> marick@m.cs.uiuc.edu writes:
>
>I have a question related to program complexity vs size.  But, first,
>my terms:
>  if (global1 == global2)
>    global1 = a;
>This has 4 data references but only 3 distinct references (since there
>are two duplicate references to global1).
The distinct references are the Software Science "number of unique
operands" - n2 and the data references are the Software Science
"number of total operands" - N2.
>
>Now, if I had to guess, I'd guess that both the number of references
>and the number of distinct references grow linearly with program size
>(measured in lines of code or in number of tokens).  Does anyone know?
Most research suggests that Software Science length is highly correlated
with lines of code (Length= N1 + N2 - where N1 is the total number of
operators used) - this indicates that as the code size grows, more
operators/operands are being referenced (for this I got a PhD? :-).
What is more interesting is the n1/n2 relationship (unique references
to operators/operands). As program size grows, unique operators reach
a limit because a language only has so many operators - or at least a
programmer has a favored "working set of operators". In fact after
some point in size, one might want to consider n1 a constant. However,
program vocabulary (n1+n2) is also correlated with program size, which
suggests as you do that operands are added to a program as it increases
in size.

Warren



==========================================================================
Warren Harrison                                          warren@cs.pdx.edu
Department of Computer Science                                503/725-3108
Portland State University