xc-ladder analogy

[WoutStoffels]

A ladder has discrete runs, which represent complexity of the xc-approximation used. The first run is the LDA, where only the value of the density at each point in space is taken into account. Next is the GGA, which takes into account the direct surrouding of each point. Next is the meta-GGA, which also takes the global kinetic energy density into account. Then come the hybrid functionals and finally the RPA-like approximation, which each take the approximation of the exchange correlation functional one more step closer to the ‘divine’, exact xc-potential. Up the ladder, the accuracy increases but computational efficiency decreases, so a balance should be struck between the two

[Author]

Posts