Efficient evaluation of AGP reduced density matrices

Abstract: We propose and implement an algorithm to calculate the norm and reduced density matrices of the antisymmetrized geminal power (AGP) of any rank with polynomial cost. Our method scales quadratically per element of the reduced density matrices. Numerical tests indicate that our method is very fast and capable of treating systems with a few thousand orbitals and hundreds of electrons reliably in double-precision. In addition, we present reconstruction formulae that allows one to decompose higher order reduced density matrices in terms of linear combinations of lower order ones, thereby reducing the computational cost significantly.