The reverted ludcmp in <div><br></div><div><a href="http://www.admb-project.org/svn/branches/revert-ludcmp">http://www.admb-project.org/svn/branches/revert-ludcmp</a></div><div><br></div><div>and ADMB-9 (before replacement)</div>
<div><br></div><div><a href="http://www.admb-project.org/redmine/projects/svn-archived/repository/revisions/794/show/trunk">http://www.admb-project.org/redmine/projects/svn-archived/repository/revisions/794/show/trunk</a></div>
<div><br></div><div>Both take ~8minutes and 20 seconds to run, the current trunk takes ~10minutes and 30seconds.</div><div><br></div><div>There a few examples where there are slight .par output differences, but are pretty close.</div>
<div><br></div><div>For Mollie's model, the trunk will not work, but the reverted code does.</div><div><br></div><div>If everyone is okay with the results, I will revert the current trunk's ludcmp function to </div>
<div>ADMB-9.</div><div><br></div><div>Let me know,</div><div>JOhnoel</div>