Symmetry dependent results

Dear MRCC community,

I noticed an interesting result obtained with the mrcc code that I thaught might be worth sharing.
For the H2He+ molecule in a nuclear arrangement of C2v point group symmetry, I get different FCI excitation energy when the mrcc program identifies and uses the symmetry of C2v, compared to when I declare the lower symmetry of Cs to be used (cmpgrp=Cs).

The input was the following:
# H2He+ input file R1 R2 R3 6.94789 3.32212 6.94789
basis=aug-cc-pVDZ
calc=FCI
mem=3000MB
dens=2
nstate=2
charge=1

geom=xyz
3

H 0.0000 0.0000 0.0000
H 0.0000 1.7580 0.0000
He 0.0000 0.8790 3.5700

Dear Tamas,
The reason for that is quite simple. You search for totally symmetric roots in both cases. However, in Cs the B1 irrep of C2v reduces to the totally symmetric irrep, and thus, in Cs what you get is the lowest root of B1 symmetry.
If you must use Cs point group for some reason, you should give a good initial guess to get the correct root (see the description of keywords ciguess and rest in the manual).

Dear Mihály,

thank you for the explanation, I was not aware, that the roots of only a single irrep are searched for, my mistake. Fortunately, I don't need to worry about a good initial guess, as using the Cs symmetry gives the (lower energy) root I need.

Many thanks,
Tamás