Beginning
in the spring of 2002, I will follow the discipline of
reproducible research. From this date, whenever I include complex computer
generated
figures and numbers in my papers, I will
also make the relevant software code publicly available.
With the code I
make available anyone should be able to duplicate all figures and tables in
my papers.
Why?
The case for ''reproducible research'' is outlined eloquently
in Buckheit and Donoho [1995],
who describe the discipline as follows:
''When
we publish articles containing figures which were generated by computer, we
also publish the complete software environment which generates the figures.''
This will
make the research more credible,
make the research more accurate, since other researchers
will most likely discover any errors I made in the computations,
save time and money for other researchers, and
probably help young researchers.
SHORTCOMINGS
Currently, most code I wrote runs under commercially available
software, including GAUSS, MATLAB, RATS, and S-Plus. Although most universities
has all or some of these available, there will be users that do not have access
to all of these software. However, from now on I will wrote all code in
freely available software R and Ox.
REFERENCES
J. Buckheit and D. L. Donoho
[1995]. Wavelab and reproducible research, in Wavelets
and Statistics, A. Antoniadis, ed., Springer-Verlag, Berlin, New York.
