Research code dump!04 Dec 2013
I’ve finally gotten around to cleaning up my home-brewed research code (at least a little bit), and I’ve hosted it here on my GitHub. I’m calling the project
pyqchem, and it contains Hartree Fock, MP2, coupled cluster and a smattering of excited state methods. I’d love it if you check it out!
You’ll notice that I still don’t have a way of reading in molecular geometry and basis set, and so I am still reliant on parsing G09 outputfor the atomic orbital integrals. Oh well, it’s a future project :) I’d also love to learn how to make it object oriented, as well as optimize the lower level routines in
C/C++. It is so far far from optimized right now it is ridiculous. I think my EOM-CCSD code scales as N^8 instead of N^6, just because I haven’t factored the equations right. But it works, and it has helped me learn so much about electronic structure methods. If you get the chance to look at it, I would love input for the future. I’m still developing as a programmer and I could use all the help I can get!
I hosted a lot of premade molecules (and their AO integrals), so all you have to do is execute
pyqchem.py followed by the folder of interest.
Would execute the
pyqchem.py script on the folder
h2_3-21G, which contains all the precomputed atomic-orbital basis integrals. The type of calculation is changed by editing the
pyqchem.py script itself. Say I wanted to perform an MP2 calculation on H2 in a 3-21G basis. I open
pyqchem.py, and edit at the top:
And you’ll see the pretty output dump to your terminal :)
That’s all there is to it! Enjoy!