Skip to content

Open-shell calculation with BLOCK. Reading input issues. #55

@dithillobothrium

Description

@dithillobothrium

Hello,

I tried to simulate open-shell system, but BLOCK crashes on reading the FCIDUMP file.

For example, I want H2 molecule as open-shell. FCIDUMP file is exported as follows

&FCI NORB=   2,NELEC=   2,MS2= 1,
  ORBSYM=1,1,
  IUHF=.TRUE. ,
  ISYM=0,
/
 4.6190068036607312E-01   1   1   1   1
...
 2.0693987712918233E-01   4   4   4   4
...
-2.5214903899133018E-01   4   4   0   0
 4.3708714060572051E-01   0   0   0   0

it catches an exeption about wrong indexation in
https://github.com/sanshar/StackBlock/blob/f95317b08043b7c531289576d59ad74a6d920741/input.C#L2714
because it creates matrix 2x2, since I have NORB=2 but not 4x4.

But, according to agreements NORB should define number of orbitals for one spin direction. In case of open shell the total number of spin orbitals will be 2*NORB.

IF I change it to 4 orbitals, I get the error

Initial HF occupancy guess: 1 0 0 0 1 0 0 0 
Checking input for errors
Orbital cannot have an irreducible representation of 0  with c1 symmetry

and we see, that the total number of orbitals became 8, so NORB=4 is multiplied somewhere by 2. What is wrong here?

Also, I found in the code that o searches for "IUHF", not for "UHF" in the FCIDUMP file.

Could you help to resolve these problems?

Thanks.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions