Atomic PAW setups
Calculating matrix elements of nabla
This integral is needed for LrTDDFT and response function related quantities:
where \(|\phi_i\rangle = \phi_i(\mathbf r) = \phi_j(r)Y_{\ell m}(\hat{\mathbf r})\), and \(|\tilde\phi_i\rangle = \tilde\phi_i(\mathbf r) = \tilde\phi_j(r)Y_{\ell m}(\hat{\mathbf r})\).
Since we use real-valued spherical harmonics, we have:
Splitting the integral in radial and angular parts, we get:
where \(G_{\ell m,\ell'm'}^{\ell''m''}\) are Gaunt coefficients calculated with
the gaunt() function and the last angular integral has been
calculated with the nabla() function.
More stuff
- class gpaw.setup.Setup(data, xc, lmax=0, basis=None, filter=None, backwards_compatible=True)[source]
Attributes:
Name
Description
ZCharge
typeType-name of setup (eg. ‘paw’)
symbolChemical element label (eg. ‘Mg’)
xcnameName of xc
dataContainer class for information on the the atom, eg. Nc, Nv, n_j, l_j, f_j, eps_j, rcut_j. It defines the radial grid by ng and beta, from which r_g = beta * arange(ng) / (ng - arange(ng)). It stores pt_jg, phit_jg, phi_jg, vbar_g
Attributes for making PAW corrections
Name
Description
Delta0Constant in compensation charge expansion coeff.
Delta_iiLLinear term in compensation charge expansion coeff.
Delta_pLPacked version of
Delta_iiL.dO_iiOverlap coefficients
B_iiProjector function overlaps B_ii = <pt_i | pt_i>
dC_iiInverse overlap coefficients
EReference total energy of atom
MConstant correction to Coulomb energy
M_pLinear correction to Coulomb energy
M_pp2nd order correction to Coulomb energy and Exx energy
M_wpp2nd order correction to erfc screened Coulomb energy and Exx energy for given w.
KcCore kinetic energy
K_pLinear correction to kinetic energy
ExxCCore Exx energy
X_pLinear correction to Exx energy
MBConstant correction due to vbar potential
MB_pLinear correction due to vbar potential
dEH0Constant correction due to average electrostatic potential
dEH_pLinear correction due to average electrostatic potential
I4_iipCorrection to integrals over 4 all electron wave functions
NctAnalytical integral of the pseudo core density
nctIt also has the attribute
xc_correctionwhich is an XCCorrection class instance capable of calculating the corrections due to the xc functional.Splines:
Name
Description
pt_jProjector functions
phit_jPseudo partial waves
vbarvbar potential
nctPseudo core density
ghat_lCompensation charge expansion functions
tauctPseudo core kinetic energy density
- calculate_projector_overlaps(pt_jg)[source]
Compute projector function overlaps B_ii = <pt_i | pt_i>.
- class gpaw.setup.Setups(Z_a, setup_types, basis_sets, xc, *, filter=None, world=None, backwards_compatible=True)[source]
Collection of Setup objects. One for each distinct atom.
Non-distinct atoms are those with the same atomic number, setup, and basis.
Class attributes:
nvalenceNumber of valence electrons.naoNumber of atomic orbitals.ErefReference energy.core_chargeCore hole charge.