[DFTB-Plus-User] eigenvec.out

[Alessandro Landi]

Dear Balint, 

Thank you for your help. 

Unfortunately, I still have problems. 

I actually did some calculations using either "ReadInitialCharges = Yes"
or "ReadInitialCharges = No", but surprisingly I got the same overlap
matrix (whereas the Hamiltonian in "hamsqr.dat" was different in the two
cases).

Obviously, I still have the problem that C^T SC is not an identity
matrix, since S and C are the same. 

If it may help, I report below the input for H2O calculation: 

Geometry = GenFormat {
3  C
 O H
     1    1    0.00000000000E+00  -0.10000000000E+01   0.00000000000E+00
     2    2    0.00000000000E+00   0.00000000000E+00   0.78306400000E+00
     3    2    0.00000000000E+00   0.00000000000E+00  -0.78306400000E+00
}

Driver = ConjugateGradient {
  MovedAtoms = 1:3
  MaxForceComponent = 1.000000000000000E-008  # Extremely small!
  MaxSteps = 100
  OutputPrefix = "geom.out"
}

Hamiltonian = DFTB {
  SCC = Yes
  SCCTolerance = 1.0E-010  # Extremely small!
  MaxSCCIterations = 1000
  MaxAngularMomentum = {
    O = "p"
    H = "s"
  }
  Charge = 0.000000000000000E+000
  SpinPolarisation = {}
  Eigensolver = Standard {}
  SlaterKosterFiles = {
    O-O = "O-O.skf"
    O-H = "O-H.skf"
    H-O = "O-H.skf"
    H-H = "H-H.skf"
  }
  ReadInitialCharges = Yes
}

Options = {
  WriteAutotestTag = Yes
  WriteHS = Yes
}

Analysis = {
WriteEigenvectors=Yes
EigenvectorsAsTxt=Yes
}

ParserOptions = {
  ParserVersion = 4
} 

Do you have any idea what am I doing wrong? 

Best regards, 

Alessandro 

Il 2017-10-17 08:25 Bálint Aradi ha scritto:

> Dear Alessandro,
> 
>> Let me clarify: if I have the following file "eigenvec.out" (H_2 O
>> calculation):
> 
>> Are the eigenvector listed by columns? I mean, is the eigenvector matrix
>> as following?
> 
> Yes, the first column represents the components of the respective
> eigenvector.
> 
>> I am puzzled about this point because I tried to compute C^T SC (using
>> the overlap matrix in the file "oversqr.dat" as S) but I did not obtain
>> an identity matrix, as I expected.
> 
> Please note, that if you switch on the writeHS option, there program
> just builds H and S and writes them and stops. There are no SCC-cycles
> made to reach convergence. Therefore, if you obtained your eigenvectors
> from an SCC-calculation, you should make sure, that your read in the
> charges.bin of that calculation (ReadInitialCharges = Yes), when asking
> DFTB+ to write H and S. Then, the overlap you obtain should be
> compatible with the eigenvectors and give the expected norm.
> 
> Best regards,
> 
> Bálint 
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at mailman.zfn.uni-bremen.de
> https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user

-- 

> Alessandro Landi, Ph.D. Student

> Università degli Studi di Salerno 
> Dipartimento di Chimica e Biologia "Adolfo Zambelli"
> Via Giovanni Paolo II, 132 - 84084 Fisciano (SA), Italy
> Phone number +39 089969390
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20171017/8704a749/attachment.htm>