> As output I now have 12 .xml files - I specified 12 nodes for input.
> Each of them seems to be capable of producing an output file with
> tablix_output. Which one should I be using?
Choose the one that you think it is the best. All solutions satisfy your
restrictions. There are only minor differences between them. I usually
do the following:
$ grep grade result?.xml
This will show the grade (fitness value) of each solution. You can then
choose the solution that has the lowest grade (and is probably the
best).
> Supplementary question - how does one produce a graph like the ones on
> the Tablix web page?
See "plotme" and "plotmore" utilities. They are in the test/
subdirectory in the source distribution or in the
/usr/share/doc/tablix/examples in debian packages. Example usage:
./plotme ./
will plot data from all conv?.txt files in the current directory.
./plotmore conv0.txt xmlfile.xml
will plot the number of errors for each module used versus time (actualy
generations).
Hope that helps
Tomaz Solc
This archive was generated by hypermail 2.1.8 : Tue Aug 16 2005 - 20:42:14 CEST