Author Topic: ATKError: inverse(DZMatrix const &) : Could not LU factorize!  (Read 4098 times)

0 Members and 1 Guest are viewing this topic.

Offline namrata

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Reputation: 0
    • View Profile
hello everyone ,

scf calculation of my two probe system does not converge and gives and error 1 in between the scf output and  scf calculations ends giving error 2.

1]# Pulay mixing inversion failed. Using only last step.

2]ATKError: inverse(DZMatrix const &) : Could not LU factorize!
 


« Last Edit: December 23, 2009, 13:01 by namrata »

Offline wring

  • Regular QuantumATK user
  • **
  • Posts: 24
  • Reputation: 0
    • View Profile
Re: ATKError: inverse(DZMatrix const &) : Could not LU factorize!
« Reply #1 on: December 24, 2009, 01:37 »
The structure is not right. Two atoms may be at the same position.

Offline zh

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: ATKError: inverse(DZMatrix const &) : Could not LU factorize!
« Reply #2 on: December 24, 2009, 14:39 »
Yes,  the geometry structure of your system may be not reasonable, as said by 'Wring'. You had better check it and see whether some atoms are too close to each other.

Offline Nordland

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 812
  • Reputation: 18
    • View Profile
Re: ATKError: inverse(DZMatrix const &) : Could not LU factorize!
« Reply #3 on: December 24, 2009, 15:57 »
The error #1 can also occur if you are using the undocumented density matrix constraint from ATK 2008.10 and the system has converged to unphysical zero charge state.