You are not logged in.
During the spring semester I studied circular quantum dots, using variational Monte Carlo, in a 10 credit course.
This summer I looked a bit into autoconf, and in total I ended up with my very own C++ library.
https://aur.archlinux.org/packages.php?ID=51778
The library contains an abstract class for wavefunctions (WaveFunc.h), and in addition a superclass for the Monte Carlo integration part (MC.h).
Currently I have subclassed WaveFunc with two classes, one for a 2 particle closed shell quantum dot and another one for more than two particles.
(The 2 particle class has even basic support for OpenCL...)
MC has two subclasses, one simply calculate energies for different combination of variational parameters (this is a great tool to get a clue where to start a search for energy minima), and another class for a more `intelligent' conjugate gradient search.
If someone who studies physics have interest in this, it would be cool to extend it to other cases (atoms etc.), and implement other minimisation teqniques.
It would also be cool to make the OpenCL integration better.
Anyone who knows physics, C++, OpenCL, autoconf and/or PKGBUILDs are welcome to look at this.
Comments, fixes, and general feedback are greatly appreciated.
Offline