Alejandro Varas edited geometry optimization.tex  over 9 years ago

Commit id: eec609b3f2fe59f8b7a51032dee4a97c7b5be147

deletions | additions      

       

Among the variety of local minimizers available on the literature, the so-called FIRE (Fast Inertial Relaxation Engine) algorithm, put forward by Bitzek {\it et al.}~\cite{Bitzek2006}, has become a robust state of the art local minimizer. It has proven to have a competitive performance compared with both standard implementations of the conjugate gradient method, and more sophisticated schemes typically used in {\it ab initio} calculations. A recent article shows also the FIRE as the most convenient algorithm do to its speed and precision in order to reach the nearest local minimum starting from a given initial configuration~\cite{Asenjo2013}.   Until now, all the local minimizers implemented in Octopus were relying upon the GNU Scientific Library (GSL)~\cite{Galassi2009}. However, the FIRE algorithm has not yen been implemented into the GSL, so we have implemented it enterely inside the Octopus Code without relying on any external library.