<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><span style="font-family: monospace;">Dear all,<br><br>I have a general question regarding the search radius parameter <br>for this optimizer. <br><br>Some articles and codes I have read choose a positive search radius; I have yet to see one with a negative value. <br><br>My gut feeling is that a negative search radius will search the parameter space in the other direction....matrix A ( see code below) would be negative and thus, the delta value would be negative as well, which seems to be fine in my opinion.<br><br>Any insight on this would be appreciated.<br><br><br><br>*******************************************************<br>lines of code where radius is defined in optimizer<br><br>A.set_identity();<br>for(unsigned int i = 0; i < spaceDimension; i++) <br>{<br>A(i,i) = m_InitialRadius / scales[i];<br>}</span><br><br> delta = A *
f_norm;<br> child = parent + delta;<br><br><span style="font-family: monospace;"><span style="font-family: arial;">convergence criterion: f-norm of A < epsilon_A</span></span><br><span style="font-family: monospace;">*******************************************************</span><br><br><br><br><br><br></td></tr></table><br>