Magnetic state and electronic structure of plutonium from ``first principles'' calculations

The goal to describe plutonium phases from ``first principles'' calculation methods is complicated by the problem of 5f-electrons localization. While for early actinides (Th, U,Np) standard DFT (Density Functional Theory) description with itinerant 5f-electrons works well for late actinides (Am, Cm) DFT calculations with completely localized (pseudocore) 5f-electrons give satisfactory results. However plutonium presents a border case of partial localization and both limits (itinerant and completely localized) are not valid. We present a review of the methods used to solve this problem and discuss what could be the reasons for their successes and failures.