Hi man, 27 12 2010 kl. 16:28 +0100, skrev Simon Börjesson: > For example, when I run the following code several times in a row, I get > different results even though the input should be exactly the same as far as > I can tell. > > %%% testing.m %%%%%%%%%%%%%%% > clear I J > > I = zeros(100,100); > I(30,:) = 1; I(:, 65) = 1; I(35:45, 35:50) = 1; > for i = 1:90, I(i,i) = 1;endfor > > J = houghtf(I); > max(max(J)) > plot(max(J)); > %%%%%%%%%%%%%%%%%%%%%% > > The reported max values seems to increase almost every run and the shape of > the plot differs as well. Sometimes only the scale of the plot is changed > but other times the shape changes drastically between different runs.
I can confirm this on Octave 3.3.54+. The issue seems to be that the output matrix initially contains random values; this surprises me quite a bit as I thought matrices contains zeros when they are initialised. I've committed a fix to SVN, so the next version of the package will have this fixed. Thanks Søren ------------------------------------------------------------------------------ Learn how Oracle Real Application Clusters (RAC) One Node allows customers to consolidate database storage, standardize their database environment, and, should the need arise, upgrade to a full multi-node Oracle RAC database without downtime or disruption http://p.sf.net/sfu/oracle-sfdevnl _______________________________________________ Octave-dev mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/octave-dev
