issues for computation of atomic PDOS (HSE06 level of theory)

[Lorenzo Lagasco]

Good evening,
I am a new user of CP2K. I would like to calculate the atomic PDOS for a coumarin 343 molecule, starting from a RESTART-wfn file of a previous calculation at the PBE level of theory. I have tried various diagonalization methods (CG, DIIS) with different preconditioners (FULL_KINETIC and FULL_SINGLE_INVERSE), but in all cases I obtain an error like this:
“Cholesky decompose failed: the matrix is not positive definite or ill-conditioned.” Here attached the input and output file of my calculation.

How can I solve this problem?

Thanks in advance for the help

Best regards

Lorenzo Lagasco

[Marcella Iannuzzi]

Hi Lorenzo

Try with a smaller number of  NLUMO in PDOS, like 50

Regards

Marcella

[Lorenzo Lagasco]

Good evening,
It worked, thank you for the help!

Best regards,
Lorenzo Lagasco

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/a7483ff0-4d11-4390-a37b-21d20e3a13ebn%40googlegroups.com.

[Lorenzo Lagasco]

Good afternoon,

I’ll take this opportunity to ask for further details regarding a PDOS calculation for a system consisting of a CuAlO₂ slab with an organic dye anchored on its surface.

When I run the calculation with a MEDIUM level of printing (input attached), the job stops before the SCF cycle begins. This makes me suspect there might be a memory issue.

I would like to ask how I could resolve this problem.

Thank you again for your valuable help.

Il giorno lun 20 apr 2026 alle ore 21:38 Marcella Iannuzzi <marci...@gmail.com> ha scritto: