Running DFT calculation with other functionals

[jacharrym]

Hello @fsmoncadaa

I got some errors when running LiH.RKS.lowdin test using different functionals by changing the input variable

electronExchangeCorrelationFunctional

I took the name of the fuciontionals by looking at the file DFT/LibxcInterface.f90 and removing the initial "XC_"

By doing the tests I realized that the code assumes that exchange and correlation are called in the same way, right? But this is not the case for some of them, for example M06_2x has two different names for X and C
XC_MGGA_C_M06_2X
XC_HYB_MGGA_X_M06_2X

Then I tried with WB97X_D using the name "HYB_GGA_XC_WB97X_D"

But it gives me the error :
lowdin-DFT.x: hybrids.c:127: xc_hyb_exx_coef: Assertion xc_hyb_type(p) == XC_HYB_HYBRID failed.

However, calling B3LYP with HYB_GGA_XC_B3LYP5 works fine

I'm attaching the files. Please let me know if I'm missing something, or if not all functionals are implemented.

Thanks!

test-dft.zip

Best regards,

[fsmoncadaa]

Hi @jacharrym !

Currenly we support LDA, GGA and HYB_GGA functionals. Meta-GGAs (MGGA) are not supported, because we are not computing yet the laplacian or kinetic energy density elements in the grid that these functionals require.

In the case of a functional that has separated exchange and correlation routines in libxc, we can put in the input:
electronExchangeFunctional="...X..."
electronCorrelationFunctional="...C..."
where as you correctly identified the name of the functional comes from DFT/LibxcInterface.f90 and removing the initial "XC_".

I'm going to check this error, that input for WB97X_D you sent should work fine.

cheers