Not recognizing symmetry properly

Dear developers,

I ran a CCSDT/cc-pvtz computation on neo-pentane (2,2-dimethylpropane) and the program recognized it as a D2 point group. I attempted to run a CCSD(T) computation with MP2 natural orbitals and while it recognized the true point group was Td, it used C2v for the computational group rather than lowering to D2. Attached is my input file. Is it possible to tinker with the symmetry cutoff, or force it to look for certain axes to get the correct symmetry?


I am using the April 12,2017 binary.

If you use the same geometry for the CCSDT and the CCSD(T) calculations, the program must recognize the same full/computational point group. Using the version you specified and the attached input file I am getting Td/C2v, which is correct. Please note that both D2 and C2v are Abelian subgroups of Td with the same number of irreps, so either of them can be used as computational point group.

The root of my problem was the program I was doing the optimization with reported it as D2 but giving a structure close enough to C2v. I redid the optimization with the code being told that the symmetry was D2 explicitly and the optimized structure was detected by mrcc as D2.