× If you have problems during the execution of MRCC, please attach the output with an adequate description of your case as well as the followings:
  • the way mrcc was invoked
  • the way build.mrcc was invoked
  • the output of build.mrcc
  • compiler version (for example: ifort -V, gfortran -v)
  • blas/lapack versions
  • as well as gcc and glibc versions

This information really helps us during troubleshooting :)

ssr functional for the dft=userd

More
5 months 2 weeks ago #1223 by nesquik91
Dear developer,

Hi, I found the nice userd example with the standard functional in the manual but I wonder if one can also use ssr (local-scaling approximation) for the userd functional.
For example, as shown in example 18 on page 74, doing XYG9[f2]@B44LYP but substituting LDA correlation part to the ssrLDA_C_VWN with the 0.5 for the omega value at the second userd block.

Thank you for the help!
 

Please Log in or Create an account to join the conversation.

More
5 months 2 weeks ago #1224 by mesterdavid
Hi,

thank you for your interest. Indeed, this part of the MINP file is not user-friendly; however, such calculations are feasible.
For this purpose, please complete the first and second block of the userd option with the following line:
0.000000001 lrHFx omega_value

Accordingly, the requested functional should be called as:
dft=userd
5
0.56 B88
0.44 HFx
0.19 VWN5
0.81 LYP
0.000000001 lrHFx 0.5
8
0.0881 ssrLDA_C_VWN
0.000000001 lrHFx 0.5
-0.0644 B88
0.9692 HFx
0.2730 VWN5
0.6447 MP2s
0.6467 MP2t
0.2761 MP3

Unfortunately, as the adjustment of the omega parameter is bounded to a range-separated wave function-based contribution, it causes some overhead in the calculations, and a negligible contribution is added to the density/energy.

Best regards,
Dávid

Please Log in or Create an account to join the conversation.

More
5 months 2 weeks ago #1225 by nesquik91
Dear Dávid,

Thank you for the nice reply. I tried some and it seems like it does well for the LDA and GGA but not for the mGGA?
My test code is like the below. The strange thing for me is that when I change the ssrMGGA_C_SCAN into MGGA_C_SCAN, then it works well. This is quite confusing to me because, if I understand correctly, userd uses the electron density of the first block, so, the convergency does not have to be affected by the second block.

Any clue for this one?



basis    = def2-SVP

dft   = userd
4
0.25 HFx
0.75 PBEx
1.00 PBEc
0.000000001 lrHFx 0.33
2
1.00 ssrMGGA_C_SCAN
0.000000001 lrHFx 0.33

mult     = 1
charge   = 0
mem      = 1GB

unit=angs
geom=xyz
2

na 0 0 0
cl 0 0 2.5

Please Log in or Create an account to join the conversation.

More
5 months 2 weeks ago #1226 by mesterdavid
This part of the code was mainly tested for LDA and GGA functionals. However, it should work for mGGA functionals as well.
You are right, the second block should not affect it. I will check it soon.

Please Log in or Create an account to join the conversation.

More
5 months 2 weeks ago - 5 months 2 weeks ago #1227 by mesterdavid
Hi,
the bug has been fixed. The new integ files (integ.f source code and integ binary file) have been uploaded to the following folder: mrcc.hu/index.php/download-mrcc/category/5-mrcc-patches
Please recompile the code or overwrite the binary with the corresponding new file.

Best regards,
David
Last edit: 5 months 2 weeks ago by mesterdavid.

Please Log in or Create an account to join the conversation.

More
5 months 2 weeks ago #1228 by nesquik91
Thanks for the quick patch! Now it works fine.

Please Log in or Create an account to join the conversation.

Time to create page: 0.018 seconds
Powered by Kunena Forum