Summary: Statistical properties of programs.

Some time ago I posted a request for references to studies on the
statistical properties of programs. The references I received in
response to my query were concerned either with the static structure
(of which I am mostly interested) or the dynamic behavior of programs.


These were some of the suggestions I received:


    * You may wish to look into the program complexity metrics
        such as McCabe and Halstead.


    * R. E. Berry and B. A. E. Meekings (1985), "A Style Analysis of C
        Programs", Communications of the ACM, 1985


    * B. W. Kernighan and P. J. Plauger (1978), The Elements of Programming
        Style, McGraw-Hill.


    * See computer performance metrics in
        _Computer Structures: Readings and Examples_
        by C. Gordon Bell and Allen Newell, pp. 49-52.
        They describe (and cite) a couple computer performance studies
        which measure performance based on the ability to execute actual
        programs;


    * @article{SheLi:90,
author = {Zhiyu Shen and Zhiyuan Li and Pen-Chung Yew},
title = {An Empirical Study of {FORTRAN} Programs for
                                  Parallelizing Compilers},
journal = {IEEE Transactions on Parallel and Distributed Systems},
month = {July},
year = {1990},
volume = {1},
number = {3},
pages = {356 -- 364}


    * One book that presents some data on instruction set usage is
        "Computer Architecture - A Quantitative Approach",
        John L Hennessy & David Patterson, Morgan Kaufmann Publishers,
        ISBN 1-55880-069-8.
        Chap 4 is entitled "Instruction Set Examples and Measurements
        of Use." It covers VAX, 360, 8086, and some historical
        architectures of interest.


Thanks to:
      Jim Davies jrbd@craycos.com
      David Spuler spuler@coral.cs.jcu.edu.au
      Neil R. Ormos thssno@iitmax.iit.edu
      Jeff Martens martens@cis.ohio-state.edu
      Allan MacKinnon mondomon@iesun.gatech.edu
      Pete Jinks pjj@cs.man.ac.uk


--
Christian.Collberg@dna.lth.se
Department of Computer Science, Lund University, BOX 118, S-221 00 LUND,
Sweden
--

Return to the comp.compilers page.
Search the comp.compilers archives again.