QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: wring on January 9, 2009, 09:30

Title: ATKError: inverse(DZMatrix const &) : Could not LU factorize!
Post by: wring on January 9, 2009, 09:30

When I use the 3*3 fcc Au bulk electrode,the calculation of the electrode and equvilatent bulk goes on very well, but the calculation of the two-probe doesn't carry out.It accurs the next error:
raceback (most recent call last):
  File "<string>", line 80, in ?
ATKError: inverse(DZMatrix const &) : Could not LU factorize!
 Why ?
 Thank you for your replies!

Title: Re: ATKError: inverse(DZMatrix const &) : Could not LU factorize!
Post by: Nordland on January 9, 2009, 11:07

Hey Wring!

I have seen this error in two case - perhaps one of them apply for you:

• You are using the newest ATK 2008.10 with the density matrix contraint, and all the charges has left the system. Solution: Tweak the parameters to obtain better convergence!

• The geometry is incorrect at the electrode interfaces, and atoms ends on top of one another in the twoprobe configuration.Solution: Inspect in VNL and adept!

Alternative If you post your script here on the forum or message it to me, I can have a look at it if you need help.

Title: Re: ATKError: inverse(DZMatrix const &) : Could not LU factorize!
Post by: Hasan Sahin on January 9, 2009, 22:56

It was discussed before,

http://quantumwise.com/forum/index.php?topic=38.0