Am Mittwoch, den 20.08.2008, 12:20 +0200 schrieb Luca Favatella: > 2008/8/20 Thomas Weber <thomas.weber.mail@gmail.com>: > [...] > > The current DESCRIPTION doesn't mention the 'miscellaneous' package > > anymore. Don't you need it? > > Exact, current version doesn't need miscellaneous. > > > > Rafael, I'm going to revert some of your unreleased changes to the > > package. The package is misclassified as > > Architecture: any > > but it contains only .m files. > > Exact, only .m files. Hmm, the unit tests fail (nr unknown). Maybe the applied patch should be pushed into SVN? (ignore the newline stuff, that's unimportant). Thomas
Index: __ga_initial_population__.m
===================================================================
--- __ga_initial_population__.m (Revision 5250)
+++ __ga_initial_population__.m (Arbeitskopie)
@@ -41,7 +41,7 @@
CreatedPopulation(1:options.PopulationSize, 1:GenomeLength) = \
options.CreationFcn (GenomeLength, FitnessFcn, options);
Population(1:options.PopulationSize, 1:GenomeLength) = vertcat \
- (options.InitialPopulation(1:nr, 1:GenomeLength),
+ (options.InitialPopulation(1:nrInitialPopulation, 1:GenomeLength),
CreatedPopulation(1:(options.PopulationSize - nrInitialPopulation),
1:GenomeLength));
elseif (nrInitialPopulation == options.PopulationSize)
@@ -67,4 +67,4 @@
%! GenomeLength = 2;
%! options = gaoptimset ("InitialPopulation", [1, 2; 3, 4; 5, 6]);
%! Population = __ga_initial_population__ (GenomeLength, @rastriginsfcn, options);
-%! assert (size (Population), [options.PopulationSize, GenomeLength]);
\ No newline at end of file
+%! assert (size (Population), [options.PopulationSize, GenomeLength]);