Junhao Li
Junhao Li
I am not familiar with the detailed algorithm of coupled cluster, but I can try...
I followed your blog post on CCSD (http://joshuagoings.com/2013/07/17/coupled-cluster-with-singles-and-doubles-ccsd-in-python/), but the energy I got is slightly different from Molpro and PySCF by several microHartree, even though the convergence criterion is 1...
I noticed that part and I thought maybe you were using a different convention. But, besides the nuclear repulsion, the CCSD correction itself is different, as mentioned in the previous...
The code is here: https://github.com/jl2922/McMurchie-Davidson/blob/ccsd/mmd/postscf.py The SCF energy is exactly the same as Molpro. Only the CCSD correction is different. In this case, what might be the reason?
Thank you! These are quite helpful.