Molcas Forum

s.savi.lorenzo

Member

Registered: 2025-09-15

Posts: 2

[SOLVED] Problem with HDIAG

Hi,

I have a very dumb question but i am going crazy

I have 2 RASSCF states and I want to diabatize them using DQVD, Boys method. However I want to change the diagonal energies using CASPT2 results. Unfortunately I can not save Jobmix nor the wavefunction as the molecule is too big.

I tried using HDIAG but I do now understand how to actually change the energies. This is an example of the input (but I tried like 20 combinations of keyword-position of the key values, like HDIAG=0.0 1.0 or others):

&RASSI
HDIAG
0.0000 1.0000
TDMN
0.001
DQVD
ALPHA=0.0
BETA=0.0

This is the output:

       EIGENSTATES OF A SPIN-FREE HAMILTONIAN WILL BE COMPUTED BASED ON:

  A Hamiltonian matrix computed by RASSI.
  In addition, the diagonal energies of the hamiltonian matrix will be replaced by either the user (HDIAG keyword) or read from the wavefunction file(s).

  Nr of states:                    2

   State:        1   2
  JobIph:        1   1
 Root nr:        1   2

  HAMILTONIAN MATRIX FOR THE ORIGINAL STATES:

  Diagonal, with energies
  -9048.30050056  -9048.19767399

      OVERLAP MATRIX FOR THE ORIGINAL STATES:

  Diagonal, with elements
      1.00000000      1.00000000

  USER-MODIFIED HAMILTONIAN FOR THE ORIGINAL STATES:
 (With user shifts, and/or replaced diagonal
  elements, including overlap corrections.)
      0.00000000
     -0.00000000      0.00000000

::    RASSI State    1     Total energy:       0.00000000000000
::    RASSI State    2     Total energy:       0.00000000000000

...
...
...
++ DQV Diabatization section
   -------------------------

  The TRQ matrix in DQVDiabat
  mat. size =     2x    2
     -859.5874273865474606    -16.5190751408854517
      -16.5190751408854517   -657.3077810944332668

 Converged in this many iterations:
                    2


  Diabatic Coefficients
  mat. size =     2x    2
        0.9969505529523335     -0.0780358569379262
        0.0780358569379262      0.9969505529523335

  Weights of adiabatic states
  mat. size =     2x    2
        0.9939104050319635      0.0060895949680365
        0.0060895949680365      0.9939104050319635

 The eigenvectors may no longer be in
 energetic order


  Diabatic Hamiltonian
  mat. size =     2x    2
       -0.0000000000651370     -0.0000000004134600
       -0.0000000004134600      0.0000000000669560

Thanks in advance

Last edited by s.savi.lorenzo (2025-09-17 08:02:10)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,141

Re: [SOLVED] Problem with HDIAG

Try explicitly writing the states:

&RASSI
NROFJOB
1 2 * 1 file (so 1 line below), 2 states in the file (so 2 numbers in the 1st line)
1 2 * States 1 and 2 (from file 1)
HDIAG
0.0000 1.0000
TDMN
0.001
DQVD
ALPHA=0.0
BETA=0.0

s.savi.lorenzo

Member

Registered: 2025-09-15

Posts: 2

Re: [SOLVED] Problem with HDIAG

Dear Ignacio,

that's working!!!!!
Thank you so much!

Lorenzo