RI-RPA

[Diganta Dasgupta]

Greetings,

I am attempting to perform a single-point RPA calculation on a system of 128 water molecules using cc-TZ / RI-TZ (primary and auxiliary) basis sets together with GTH-PBE pseudopotentials. However, the calculation consistently crashes with the following error : : CPASSERT failed -   mp2_ri_2c.F:1441. The attched folder has the relevant files for such a run. I have tried altering the MGRID cutoff and the EPS and MEMORY settings — but the error persists. As I am still relatively new to this, I may be overlooking something obvious. Any guidance is greatly appreciated. 

Best regards,

Diganta

[Diganta Dasgupta]

Greetings,

I have been to run single point calculations on a system of 128 waters on cp2k using RPA using the cc-TZ and RI-TZ primary and auxilliary basis sets respectively with a GTH-PBE pseudopotential. However after some period of running the calculations crash with a messege saying : CPASSERT failed - mp2_ri_2c.F:1441. The attached folder contains all the files from a run. I have tried changing MGRID cutoff, the EPS and MEMORY settings but to no avail. I'm new to this and may be making some obvious mistakes. Any guidance in this regard is greatly appreciated.

Best regards,

Diganta

[Frederick Stein]

Dear Diganta,

This CPASSERT is related to a failing Cholesky decomposition. In your case, this may be related to diffuse basis functions making the two-center electron repulsion matrix numerically non-positive definite. You need to add

&RI

DO_SVD

&END

to your &WF_CORRELATION section. Are you later also interested in force calculation?

Best,

Frederick